The effect of industry leaders’ exploratory innovation on competitor performance

Abstract

This paper examines the effect of an industry leader’s exploratory innovation, defined as the innovation embodying novel knowledge relative to the firm’s extant knowledge, on the performance of its direct competitors. We argue that an industry leader’s exploratory innovation can benefit its competitors, resulting in an average increase in competitors’ sales. The benefit can come from advantageous inter-industry structure, higher perceived status through association, and expanded knowledge pool. The extent of benefit, however, is conditional on the number of competitors in the industry as well as the level of competitors' financial slack and Return on Assets (ROA). Using data on the U.S. computer sector, we find support for our hypotheses. This study suggests that while an industry leader’s exploratory innovation is intended to further its own interests, it also affects the plight of its direct competitors.

Full text

The effect of industry leaders’ exploratory innovation on competitor
performance
Solon Moreira
Entrepreneurship Department
IESE Business School
Av. Pearson, 21, 08034 Barcelona, Spain
E-mail: smoreira@iese.edu
C. Jennifer Tae
Fox School of Business
Temple University
1801 Liacouras Walk, Philadelphia PA 19122
E-mail: jennifer.tae@temple.edu
Abstract. This paper examines the effect of an industry leader’s exploratory innovation, defined
as the innovation embodying knowledge that is novel relative to the firm's extant knowledge, on
the performance of its direct competitors. We argue that an industry leader’s exploratory
innovation can benefit its competitors, resulting in an average increase in competitors’ sales. The
benefit can come from advantageous inter-industry structure, higher perceived status through
association, and expanded knowledge pool. The extent of benefit, however, is conditional on the
number of competitors in the industry as well as the level of competitors’ financial slack and
Return on Assets (ROA). Using data on the U.S. computer sector, we find that competitors with
an industry leader who has a higher proportion of exploratory innovation experiences higher
sales. While the number of competitors in the industry and a competitor’s ROA lessen this effect,
a competitor’s slack resources amplifies it. This study suggests that while an industry leader’s
exploratory innovation is intended to further its own interests, it also affects the plight of its
direct competitors.
1

INTRODUCTION
Innovation plays an important role in firm performance and survival as well as an industry’s
competitive dynamics (e.g. Abernathy and Utterback, 1978; Dosi, 1982; Klepper, 1996; Nelson
and Winter, 1982). Prior studies have looked at various outcomes that innovation affects. They
range from industry demographics such as firm’s entry (Adner and Levinthal, 2001; Adner and
Zemsky, 2005; Agarwal and Gort, 1996; Mitchell and Skrzypacz, 2015), to the way firms
collaborate within an industry (Ahuja, 2000; Powell, Koput, and Smith-Doerr, 1996), and to the
knowledge pool firms can tap into (Katila and Chen, 2008; Yang, Phelps and Steensma, 2010).
These studies have shown that innovation has a significant impact on the competitive position of
both the innovating firm and its direct competitors. Despite the importance of innovation for an
industry competitive landscape, not all firms are able to shape their industry through innovative
activities.
Within an industry, firms vary significantly in their capacity to innovate and to influence
the industry’s overall structure and dynamics (Spithoven, Frantzen, and Clarysse, 2010).
Particularly, due to their market power and superior capabilities, the actions of industry leaders1
have a strong influence on their direct competitors (de Figueiredo and Silverman, 2007;
Jacobides and Tae, 2015; Wang and Shaver, 2014). Accordingly, in order to better understand the
performance of industry participants, it is important to consider not only the strategic actions of
individual firms, but also the influence of actions taken by an industry leader who can shape the
competitive landscape of the industry. Interestingly, the way in which industry leaders’
innovation may impact the performance of other industry participants has received relatively
little attention from prior studies.
Moreover, leaders in different industries are likely to differ in their innovation partly due
to the heterogeneity in their capacity, and to the different competitive environment they face
within their industries. Even when industries are interconnected through a supplier-buyer
relationship (Malhotra, Gosain, and Sawy, 2005), thus making one industry at least partly
dependent on another (Gereffi, Humphrey, and Sturgeon, 2005; Jacobides and Tae, 2015),
competitive environment of each industry is different. For example, in the computer sector, the
empirical context of this study, there are input, or component providing industries such as storage
devices and semiconductors and an output industry, viz. electronic computers. Seagate, an
1 We define an industry leader as a firm with the highest market share.
2

industry leader in storage devices, faces different competitive pressures than Intel, an industry
leader in semiconductors. So does Dell, an industry leader in electronic computers. As a result,
we not only observe heterogeneity in terms of the type and nature of innovation between storage
device, semiconductor, and electronic computer industries, but also among Seagate, Intel, and
Dell. This is in spite of the fact that their performance is partly dependent on that of another:
Computer sales is directly related to how many semiconductors can be sold as well as how many
storage devices may be in demand. The persistence heterogeneity in terms of leader innovation
strategy, even in strongly interdependent industries, makes it particularly relevant to understand
how different industries are shaped by the innovation output of their most influential firms.
Bringing these arguments together, we examine the impact of an industry leader’s
exploratory innovation, defined as innovation that utilizes novel knowledge, on the performance
of its directly competing firms. We focus specifically on this type of innovation because
exploratory innovations are the ones most likely to change the industry competitive landscape
through the creation of new products, reshape networks and create different industry standards
(Katila and Ahuja 2002, Jensen, Van Den Borsch, and Volberda 2006). Our central argument is
that the industry leader’s exploratory innovation has a positive effect on performance of other
firms in the same industry. We start by highlighting that industry leaders undertake exploratory
innovation to maintain their competitive advantage. Next, we propose a number of ways in
which this type of innovation outcome may affect performance of the competitors in the same
industry. Because industries and firms have different characteristics, we also examine important
contingencies related to industry and firm level characteristics affecting the leader exploratory
innovation-performance relationship. In particular, we suggest that the net effect of performance
improvement at the industry level is contingent on the number of its participants, as it determines
the competitive pressure among participants. We then argue that firm-level characteristics such
as financial slack in form of resource availability and return on assets (ROA) moderate the extent
of performance improvement. In doing so we link two different levels of analysis in studying the
effect of industry leader’s innovation on competitor performance: Heterogeneity in innovation
observed across vertically related industries and heterogeneity among firms within each industry.
We empirically test these relationships in the US computer sector. The computer sector
consists of industries that assemble the final product, computers, and supply inputs such as
semiconductors, software, and other peripherals. This setting provides an appropriate context to
3

examine the main relationship of interest in this paper for several reasons. First, this sector is
characterized by constant innovation and dynamic changes in the fate of participants in different
industries. It is a technology-intensive sector wherein innovation and is the fundamental source
of competitive advantage (Stuart, 2000). Firms in the computer sector such as IBM, Microsoft,
and Intel have consistently exhibited high patenting propensity (Cohen, Nelson, and Walsh,
2000). Second, despite many fluctuations in industry demographics, each industry’s leader has
remained stable over the period of observation. This allows us to alleviate any concerns on the
possible ‘changes’ in the influential firms in each industry over time. Third, the potential issue of
different industries being in a different ‘lifecycles,’ which can affect the performance prospects
of firms irrespective of the leader’s exploratory innovation, is weakened given the vertical
relatedness of the industries in this sector and their close co-evolutionary path (Bresnahan and
Greenstein, 1999).
This study complements studies on positive externality (e.g. Agarwal and Bayus, 2002;
Jacobides and Tae, 2015) that one firm’s effort to further its interest can help other firms improve
their performance. We find that increasing levels of exploratory innovation by an industry leader
is positively related to its competitors’ sales growth. Due to competitive pressures, the number of
firms in the industry negatively affects this relationship. Although these conditions influence
opportunities for competitors to improve their plight, they do not equally share the net benefit at
the industry level. We find that firm-level conditions matter in competitors’ ability to appropriate
the opportunities: Whereas financial slack positively moderates the performance improvement,
ROA does so negatively. Focusing on an individual competitor’s performance improvement as
the dependent variable and the industry leaders’ proportion of exploratory innovation as the
independent variable, our analysis complements the existing literature that has long recognized
the positive role innovation plays on both the innovator and entities external to the focal firm. In
so doing, it highlights the role of industry leader to conduct exploratory innovation on industry
growth and competitive dynamics within.
The remainder of this paper is organized as follows. Next, we briefly summarize the
related literatures, followed by the hypotheses. We then explain the empirical setting, data, and
the methods. We present our findings along with a short exposition on the robustness tests we
conducted. The final section concludes.
4

RELATED LITERATURE
The role of innovation in firm performance and industry dynamics has been a decade’s long area
of interest for researchers. And, unsurprisingly, scholars have uncovered and examined many
ways by which innovation affect performance and industry dynamics. In doing so, two related,
but distinct, views have emerged; work on performance and industry dynamics (Hall, Jaffe, and
Trajtenberg, 2005; Klepper, 1996), and work on knowledge spillovers (Griliches, 1992; Yang et
al., 2010). We briefly discuss each, then discuss how literature on industry leadership might
augment this research.
Scholarly work on the firm performance has highlight the importance of innovation for
the innovating firm (Hall et al., 2005; Spithoven et al., 2010; Kapoor and Klueter, 2015).
Research has argued that innovation enable the innovating firm to offer high quality products, be
more efficient, create something substantially novel, or otherwise gain a more advantageous
competitive position against its competitors. Recognizing that innovation sometimes requires
collaboration among multiple firms, some scholars also highlight the importance of networks and
firms’ position within them for successful innovation and subsequent performance (Ahuja, 2000;
Powell et al., 1996). Together, these works note that innovation can be a source of long-term
competitive advantage.
Similarly, work on industry evolution (Abernathy and Utterback, 1978; Dosi, 1982;
Klepper, 1996; Nelson and Winter, 1982) has studied changes in industry demographics resulting
from innovation. Scholars have studied how innovation drives entry and exit (Agarwal and Gort,
1996; Klepper, 1996) for both incumbents and new entrants (Adner and Levinthal, 2001;
Mitchell and Skrzypacz, 2015). These findings are echoed by the literature in submarkets
(Agarwal and Bayus, 2002; Bhaskarabhatla, 2016; Mitchell and Skrzypacz, 2015), arguing that
innovation helps deter potential entry or exploit dynamic capabilities.
In contrast, the literature on knowledge spillover highlights how innovation can benefit
society at large and non-innovating recipient firms in particular (Griliches, 1992; Fritsch and
Franke, 2004; Laursen, Leone, and Torrisi, 2010). They theorize that non-innovating firms
benefit from knowledge used/developed by the innovating firm without having contributed to its
generation because such knowledge or information cannot be kept completely proprietary.
Similarly, Yang et al. (2010) showed how knowledge spillover can benefit the originator
5

(innovator) by expanding the knowledge pool for all participants in an industry and providing
similar knowledge that is accessible.
That what few firms do in general, and not just their innovation, has direct implications
for those outside their boundaries has also long been recognized in the literature. Scholarly work
on dominant firms show that a dominant firm’s behavioral change can result in repositioning or
churns across submarkets in an industry (de Figueiredo and Silverman, 2007; Wang and Shaver,
2014). Other scholars have also stressed the important role played by a few firms such as
systems integrators (Brusoni, Pavitt, and Prencipe, 2001), platform leaders (Gawer and
Cusumano, 2002), and kingpins (Jacobides and Tae, 2015) who exert influence not only on their
direct competitors but also on other participants along a sector’s value chain (Gereffi et al.,
2005).
In the first two of these literatures, the perspective taken is that innovation is necessary
and beneficial to firms. Whereas the work on the effect of innovation on performance and
industry dynamics elucidate that innovating firms can reap the benefits of their efforts by
improved competitive position whilst putting their competitors at a disadvantage, the work that
views innovation as an outcome of knowledge recombination and generation highlights why the
benefits of innovation spills over to other firms. Although few studies have examined the effect
of a firm’s innovation on its rivals’ performance (Kafouros and Buckley, 2008), these studies did
not consider the varying levels of influence innovation exerts conditional on ‘who’ the innovator
is. Yet, the literature on dominant or influential firms emphasize that their actions carry
ramifications beyond their boundaries. For instance, firms are less likely to pay attention to
innovation of a small, specialist firm than to that of an industry leader or a more ‘salient’ firm.
Thus, it is natural to cast innovation of industry leaders as more influential in terms of
performance implications. Yet, our understanding of how innovation of industry leaders impacts
performance of its direct competitors is limited. With the growing interdependence of firms in
different settings in today’s business environment (Gereffi et al., 2005; Jacobides and Tae, 2015;
Malhotra et al., 2005), it is also important to consider how industry leaders in such
interconnected settings can affect the plight of their direct competitors vis-à-vis those in other
industries with their innovation.
THEORY DEVELOPMENT
6

Exploratory innovation and industry leaders
We thus posit that the characteristics of an industry leader’s innovation, specifically the
proportion that embodies novel knowledge, is critical, because not all innovation is the same in
terms of the extent of opportunities it creates for both itself and its competitors to exploit.
Firms recombine and relocate knowledge to innovate (Katila and Chen, 2008; Nelson and
Winter, 1982). When firms explore areas new to the firm during the search process, they run the
risk of lowering expected returns, but can also create more opportunities as novel knowledge
allows for more recombination. In contrast, exploiting the current knowledge base leads to less
uncertain returns at the expense of reaching local optima and exhausting recombination
opportunities. Consequently, firms need to search broadly and incorporate a more diverse set of
knowledge to overcome inertia built into their existing knowledge base (Fleming and Sorenson,
2004; Katila and Ahuja, 2004) and to “translate into more advantageous market positions (Nerkar
and Roberts, 2004: 794).”
Consequently, we can expect exploratory innovation, defined as innovation embodying
knowledge that is novel relative to the firm's extant knowledge (Katila and Ahuja 2002; Phelps,
2010), to improve performance and support future viability (Danneels, 2002). However, not all
firms are capable of successful exploratory innovation. Scholars have observed that in general,
firms with more experience, thus more established and with better capabilities, have more
success (Mitchell and Skrzypacz, 2015; Nerkar and Roberts, 2004). One reason is because of
resource availability; firms need abundant resources to find, assess, and incorporate novel
knowledge. This is also partly because better quality information, which is gathered through
extensive market experience (Adner and Zemsky, 2006; Giarratana, 2004), is important to
increase the chances of success in highly uncertain endeavor such as exploratory search
(Bhaskarabhatla, 2016). Therefore, we theorize that industry leaders, whose leadership is a proxy
of their superior resource endowment, capabilities, and market experience, are better suited to
conduct successful exploratory search.
Just as not all firms are able to successfully generate exploratory innovations, not all
industry leaders will be inclined to engage in exploratory search for their innovation. Industry
leaders in general have little incentive to change the status quo because they already have
competitive advantage and market power (Christensen, 1997; Fleming and Sorenson, 2004).
Even in cases where an industry leader engages in exploratory search, the degree of its activity
7

may be greater or less than exploratory search of leaders in other industries. Consequently, the
degree of exploratory innovation an industry leader carries out may be sufficient to sustain or
further its position within the industry, but not adequate to come out ahead across vertically
related industries in a sector.
Exploratory innovation of industry leaders and their competitors’ performance
In addition to the benefits accruing to the industry leaders, we argue that the higher proportion of
exploratory innovation by industry leaders would help their competitors improve their
performance vis-à-vis firms in other industries of a sector. Following prior literature, we
theoretically articulate a set of different mechanisms that link a leader exploratory innovation to
the changes in the performance of other firms in the same industry.
First, competitors of an industry leader with a high proportion of exploratory innovation
can bandwagon on the architectural advantage (Jacobides and Tae, 2015). Research on industry
architecture argues that firms with technological prowess use this advantage to shape the sector
to their advantage (Jacobides et al., 2006). Having a broad knowledge base enables leaders to
devise new templates of inter-industry interactions. By devising such templates, the industry
leader can enjoy strengthened bargaining power (Moatti, Ren, Anand, and Dussauge, 2015), and
shape consumer expectations and awareness to improve its performance (Agarwal and Bayus,
2002). Such structural advantages for better performance are not limited to the industry leader,
but shared among competitors. Part of this is intentional; a leader usually finds it easier to
promote the good of the industry as a whole than to do so only for itself to avoid potential
antitrust action and circumvent regulatory scrutiny (Kenney and Pon, 2011). That is, competitors
can appropriate the favorable rules of interface set by industry leaders to improve their
performance.
Second, competitors can similarly enjoy the status or reputation benefits. Research on
social networks and alliances explains how firms can benefit from being associated with a firm
with higher status and reputation (Lin, Yang, and Arya, 2009; Stuart, 2000). An industry leader
currently possesses the highest quality of capabilities and resources in the domain (Stuart, 2000).
A high proportion of exploratory innovation by industry leader further signals to the external
stakeholders that it is actively working to improve or at least maintain that quality in the future.
When industry leaders vary in their proportion of exploratory innovation, this heterogeneity
8

eventually leads to the stratification of industries in a sector because leaders of each industry
represent the highest quality within. As a result, firms in an industry with a higher status, due to
the efforts of its leader, will be able to enjoy some degree of ‘halo effects’ by association
regardless of their inferior status or reputation within the industry.
Finally, competitors can leverage the expanded knowledge pool to enhance their
innovative activities. Remaining at the technological frontier tends to be beneficial for firms to
sustain their competitive advantage over time. Doing so requires firms to search broadly and
incorporate a more diverse set of knowledge (Moreira, Markus, and Laursen 2018). However,
because of the uncertain payoff ex ante, firms are often reluctant or financially constrained to
engage in broad search. The expanded knowledge pool at the industry level, due to the industry
leader’s high proportion of exploratory innovation, can mitigate this. An expansion of an industry
leader’s knowledge base can increase the knowledge pool that its competitors can leverage,
which, in turn, benefits the industry leader, culminating in a virtuous cycle for all (Yang et al.,
2010). The salience of an industry leader’s activities would also expedite the speed at which
other firms can leverage the expanded knowledge base compared to the exploratory innovation
by a less visible firm. Therefore,
Hypothesis 1. An industry leader’s proportion of exploratory innovation is positively
associated with its competitors’ performance improvement.
There is a fixed amount of demand for a product in an industry at a given point in time. When the
revenues and profit to be derived from a particular market is fixed, the competitive pressure,
arising from higher levels of competition among many firms, puts a downward pressure on price,
resulting in fragmentation of total revenues among firms rather than the increase in total
revenues (Ethiraj and Zhu, 2008; Moreira, Cabaleiro and Reichstein 2018). The presence of
many firms in an industry can also make it difficult for firms to enjoy the architectural advantage
or the halo effect from the industry leader because each firm will have lower bargaining power
than when there are only a few firms. This suggests that competitors have to ‘share’ a fixed
ability to charge premiums or increase sales, leading to diminished benefit for each competitor.
Therefore,
9

Hypothesis 2: The number of firms in an industry negatively moderates the relationship
between a leader’s exploratory innovation and its competitors’ performance
improvement.
Heterogeneity in performance improvement among competitors
Even when there is net performance improvement in an industry, the improvement is not even
distributed across firms. Firms need to need to be motivated and able to perform better. We argue
that these differences in ability and motivation, determined by endogenous factors, are important
in determining the extent of performance improvement from leader’s exploratory innovation.
Amount of financial slack. Firms “take advantage of opportunities afforded by the
environment (Thompson, 1967: 150)” when they have resources in excess. Because firms often
lack immediate access to external resources to leverage opportunities, firms with internal slack
resources are more likely to exploit the opportunities (Troilo, De Luca, and Atuahene-Gima,
2014). Slack resources are “a pool of resources in an organization in excess of the minimum
necessary to produce a given level of organizational output (Nohria and Gulati, 1996: 1246).”
Among various slack resources, financial slack has been identified as the most fungible and
redeployable (George, 2005). Financial slack does not always correlate to past performance as
firms may have different slack accumulation processes (Dierickx and Cool, 1989) and/or their
managers may have implemented different policies toward slack (Mishina, Pollock, and Porac,
2004).
Studies have shown that how firms manage opportunities and the outcome from
exploiting such opportunities depends on their budget constraints (Kaul, 2012). To leverage a
given opportunity, firms need to acquire more assets that require investments (Nerkar and
Roberts, 2004; Troilo et al., 2014). Firms with more slack resources find it easier to acquire those
assets to improve their performance in light of their industry leader’s exploratory innovation
whereas those with little or no slack may deem the opportunity cost too high. Therefore,
Hypothesis 3: A firm’s financial slack positively moderates the relationship between a
leader’s exploratory innovation and its performance improvement.
Return on Assets (ROA). Firms differ in what they consider the best course of action in a given
situation. One of the factors affecting firms’ motivation to exploit opportunities or exert efforts to
10

change the status quo is their degree of fit with the environment. The fit between firms’ heritage
of routines and the current environment largely determines their fate (Nelson and Winter, 1982).
The differences in routines involving utilization of strategic assets give rise to efficiency rents
(Peteraf and Barney, 2003). A firm’s level of return on its existing assets can thus be an indicator
of how well firms “fit” the competitive environment. Given that firms generating significant
rents from their existing assets are likely to fit well with their current competitive landscape,
there is little incentive for them to deviate from their current course of actions.
Routines that enable profiting from existing resources also impede change (Abernathy
and Utterback, 1978) because they are specialized for a specific task (Teece, 1986) or a specific
use of available assets (Markman, Gianiodis, and Buchholtz, 2009; Nelson and Winter, 1982)
with scale limits (Levinthal and Wu, 2010). Despite the opportunity to improve the performance
further in the wake of the industry leader’s actions, firms with high returns on assets may find it
difficult to do so. Therefore,
Hypothesis 4: A firm’s level of ROA will negatively moderate the relationship between a
leader’s exploratory innovation and its performance improvement.
DATA AND METHODS
Data overview
The empirical context of our study is the computer sector. We identified Standard Industrial
Classification (SIC) codes that correspond to the computer sector using extant research (e.g.
Bresnahan and Greenstein, 1999; Jacobides and Tae, 2015). We then created a panel of firms
operating in the computer sector using three data sources—COMPUSAT (North America),
Harvard Patent Network Dataverse, and the NBER Patent Project.
First, we constructed a firm-year panel (1979-2004 inclusive) by extracting information
from COMPUSTAT on firms reporting their main Standard Industrial Classification (SIC) codes
in the Computer Sector (See Table 1).2 We removed data points with missing information about
our variables of interest. Because we are interested in tracking firms’ performance over time, we
also dropped firms that appear less than four times to increase the overall balance of the panel.
Next, we identified an industry leader for each SIC-year combination. We defined
industry leaders as firms with the highest market capitalization in each SIC-year combination.
2 We chose this time window for consistency as data on patents suffer from right censoring after 2004.
11

We also tried different identification methods by substituting market capitalization with market
share (sales) and amount of R&D investments and obtained identical results. One could be
concerned that firms may operate in multiple SIC codes. We deal with this issue in two ways.
First, we matched both the industry leader and each competitor firm based on the main SIC code.
It means that, even if a firm has sales in different SIC codes, it would be empirically connected
to a leader that operates within its main market, viz. SIC code. Second, we investigated the
dispersion of sales across different SIC codes among firms in our sample. We found that 89% of
the firm-year observations used in our analysis operate within a single SIC. It suggests that the
matching between the industry leader and competitor firms tend to be accurate.
Finally, we used industry leaders’ names to link each firm to the Harvard Patent Network
Dataverse and the NBER Patent Project. This enabled us to track the patents owned by each
industry leader in our sample, including information on the characteristics of each patent
successfully applied for at the USPTO. We used the qualitative information on patents to
calculate industry leaders’ exploratory innovation. We dropped industry leaders from the
COMPUSTAT firm-year data after we created the variable on the industry leaders’ proportion of
exploratory innovation as our study only concerns the performance effect on their competitors in
each industry. Our final sample comprises an unbalanced panel with the firms being observed 11
times on average (min= 4, max= 27) with the total number of firms-year observations 9,986
regarding 904 unique firms.
Dependent variable
Performance improvement. We are interested in the performance improvements that an industry
leader’s exploratory innovation brings to its competitors. Our dependent variable thus moves
away from prior studies that looked at innovation performance (e.g. Laursen et al., 2010). We
computed our dependent variable based on sales information using the following formula:
Performance improvement = ln(sales)it – ln(sales)it-1,
where ln(sales) is the natural logarithm of firm i’s sales, computed for both periods t and t-1. The
final variable captures the change in a firm´s sales based on the difference between the two
periods. By doing so, we tried to address two potential issues arising from the use of sales data—
skewness and small variations that can lead to serial correlation.
12

We have opted to use sales to operationalize performance improvement for several
reasons. First, other performance measures such as net income or Return on Equity (ROE) can be
affected by different factors that are not related to the current underlying operation of the firm.
Thus, in order to capture the changes in performance that are less subject to such issues, we
decided to examine the change in sales.
Explanatory variables
Leader Exploration. We capture the leader´s amount of exploratory innovation using patent
based measure similar to Katila and Ahuja (2002) approach to computing a firm’s search scope.
More precisely, we use as an input for this variable the citations in which the leader patents built
on to be generated. First, we tracked all the USPTO patents that each industry leader successfully
applied for in year t. With this yearly patent portfolio for each industry leader, we traced all
backward citations related to those patents to determine if they were cited by the firm in the past
(Katila and Ahuja, 2002) using a seven-year moving window3. To define the date of a citation,
we use the application year of the leader granted patent that mentions the corresponding citation.
Based on this time-window we tracked each citation to identify if the industry leader had used it
before within this period or if it is a new citation. Based on this criterion we then calculated the
ratio of new citations to total citations (Katila and Ahuja, 2002) to capture the proportion of the
industry leader’s exploratory innovation. The measure ranges from 0 to 1, with higher values
indicating more exploratory innovation with new (unfamiliar) knowledge. We follow prior work
arguing that patent citations are a robust indicator of knowledge flows and of the sources from
which firms draw from to innovate (e.g., Almeida, 1996; Jaffe, Trajtenberg and Henderson,
1993). This applies also to knowledge flows from acquired firms to acquirer inventors. For
example, Nerkar and Paruchuri (2005: 776) note that patent citations are “an excellent indicator
of knowledge flows,” thus serving as reliable evidence of the knowledge a patent builds on.
Number of Firms in the Same Industry. This variable is the number of firms operating
within the same four-digit SIC code as the industry leader in a given year.
Financial Slack. Many studies have operationalized slack based on financial ratio
measures (Cheng and Kesner, 1997; Singh, 1986). We also measured financial slack using the
3 In their paper, Katila and Ahuja (2002) use a five-year moving window. Because we are not looking at the effect
of exploration on the innovator firm, but on other firms in the same industry, we use a seven-year window.
Nevertheless, estimating this variable using the same five-year window they use does not change our results in this
paper.
13

ratio between current assets to current liabilities (Singh, 1986). This measure builds on the idea
that unabsorbed financial resources can be used to take advantage of new submarket.
Return on Assets (ROA). We used ROA as a proxy for the level of efficiency with which
firms leverage their existing assets. It is computed by dividing a firm´s reported net income by
the total assets in a given year t. This ratio is appropriate to account for the relationship of our
interest, as it measures how efficiently a firm deploys its existing assets to generate profits.
Control variables
Fragmentation. In order to control for qualitative dimensions of the industry leader’s innovation,
we compute a version of the fragmentation index proposed by Ziedonis (2004). It controls for the
degree of fragmentation of ownership in each industry leader’s patent portfolio. This measure
captures the extent to which the industry leader builds on a large number of sources to create its
own innovation. To calculate it, we use the following metric:
Fragmentation=1−∑
J=1
(
NBCITESij
NBCITESi
)
2
,i ≠ j ,
where j refers to the unique entities cited by the patents granted to industry leader i in a given
year. Therefore,
NBCITESi
refers to the total number of backward citations present in industry
leader i’s patents in year t, and
NBCITESij
refers to the number of unique entities listed in
those citations.
Industry leader’s Number of Patents. An industry leader’s total innovation output can
affect the fringe firms’ ability to appropriate new submarket opportunities, as the sheer quantity
can influence the number of newly created submarkets. To control for this effect, we used the
industry leader’s total number of patents in year t.
Market Competition. We control for the level of competition within an industry in a given
year using a Herfindahl index based on firm sales data. We use financial information reported in
Compustat to identify all firms operating in the same primary four-digit SIC code in a given year
t. The Herfindahl index is calculated using the sum of an industry’s squared market share,
according to the following formula:
14

where Si j is the market share of firm i in industry j. We perform the above calculations
each year for each SIC code.
Total Assets. Each firms’ size can also affect the extent of performance improvement, as
suggested in the knowledge spillover literature (e.g., Knott, 2003). We controlled for firm size by
including in our models the logarithm of firm i’s total assets in year t.
R&D Intensity. Absorptive capacity is one of the main determinants of a firm’s ability to
benefit from knowledge generated outside its boundaries (Cohen and Levinthal, 1990). As
absorptive capacity is one of the firm-level factors known to affect firm performance, we
consider this construct in our analysis. Following Cohen and Levinthal (1990), we proxied for
absorptive capacity using a firm’s level of R&D intensity computed as the ratio between
investments in R&D to sales at year t. Higher values indicate higher levels of absorptive
capacity.
Share of R&D. We expect competitors with a larger share of the industry’s R&D
investments have better-quality capabilities that can affect their ability to take advantage of
benefits arising from the industry leader’s actions. To isolate this effect, we used the competitor
i’s share of the total industry R&D investments in year t.
Lost Market-Share. Previous performance can affect a firm’s current sales, both relative
to the past and in absolute amounts. To control for the confounding effect of prior performance,
we used a dummy variable that takes the value 1 if a competitor has lost market share relative to
the previous year and 0 otherwise.
Finally, we incorporated year dummies to absorb trend effects shared by all firms in our
sample in the same year.
INSERT TABLES 1 AND 2 ABOUT HERE.
Statistical Method
Timing is one of the empirical challenges in this study. One can expect time lapse between our
dependent variable and independent variables. Given that we hypothesize the existence of
heterogeneous effects across firms in the same industry, different levels of financial slack and
ROA can also play a role in the timing firms are able to benefit from the industry leader’s
innovation. We therefore use two different lags for the dependent variable (Variables t+1 and
15

Variables t+2). This strategy is useful not only to capture both the main and moderation effects
over time, but also to reduce concerns regarding reverse causality (Angrist and Pischket, 2009).
We take advantage of the longitudinal structure of our final dataset and use firm within
fixed effects model to test our hypotheses. The use of firm fixed effects helps to control
unobserved time-invariant characteristics across firms (Wooldridge, 2009). In our setting, we
want to control for some time-invariant firm characteristics, such as geographic location or
industry affiliation, that can correlate unobserved competitors’ characteristics and the outcome
we observe. We also employed robust standard errors to our estimates to address potential
heteroscedasticity concerns.
RESULTS
Table 2 reports descriptive statistics and pairwise correlation coefficients for all independent
variables used. For simplicity, we report the correlation coefficients within the same year as the
dependent variable (t=t). To ensure any potential collinearity issues across the two lags, we
computed the mean uncentered variance inflation factor (VIF) for the full model in each lag
(t+1: 6.67; t+2: 6.90). The results did not raise any concerns that collinearity may bias our
results.
Because we focus our analysis on the industry leaders’ exploratory innovation, we
investigate the distance between the industry leader and the runner-up firm in terms of market
share in a given year. To do so, we compute the highest percentage of the runner-up firm´s
market share relative to the industry leader in each SIC code during the period covered in our
analysis. We find that the runner-up firm’s market share is negligible in most cases ranging from
0.36% (SIC 7372) to 0.89% (SIC 3575)4. The results suggest that there are significant differences
in market power between the industry leader and the runner-up competitor in each industry. This
pattern reinforces our idea to focus on the effect that industry leaders exert on their direct
competitors.
We also investigate how industry leaders and non-leaders differ in terms of innovation
and overall patent output. We conducted t-tests comparing the mean value for the patent count
produced in a given year between both groups. The result indicates that industry leaders (mean:
309.48) have significantly larger patent count than non-leader firms do (mean: 9.42). Looking
4 The full list of the runner-up firm’s market share is as follows: 0.76% (SIC 3571); 0.68% (SIC 3572); 0.89% (SIC
3575); 0.87% (SIC 3577); 0.51% (SIC 3674); 0.36% (SIC 7372); and 0.64% (SIC 7373).
16

specifically at exploratory innovation5, we also found a highly significant difference between
industry leaders (mean: 0.46) and non-leaders (0.18). This finding is also in line with extant
literature that industry leaders are the firms with the highest quality resources and capabilities
who can shape industry architecture, its status, and technological knowledge pool.
Table 3 reports longitudinal models with firm (within) fixed effects for the dependent
variable Performance Improvement with the two different lags (t+1; t+2). We employ a one-
tailed test for all coefficients. Models 1, 2, 3, 4, and 5 indicate the results for all control and
independent variables lagged for one-year relative to the dependent variable. Models 6, 7, 8, 9,
and 10 report the results for the two-year lag. We enter the explanatory variables in the models in
a step-wise manner. Models 1 and 6 only included all the control variables. We introduced the
variable Leader Exploration (Hypothesis 1) in Models 2 and 7. The first interaction Leader
Exploration x Number of Firms in the Same Industry (Hypothesis 2) is introduced in Models 3
and 8. The interaction terms Leader Exploration x Financial Slack (Hypothesis 3) was
introduced in Models 4 and 9. Finally, Models 5 and 10 present the full model with the inclusion
of the interaction term Leader Exploration x Return on Assets (Hypothesis 4). Because the
results in restricted and full models are identical in directions, we only interpret the results in the
full model estimated for each lag.
INSERT TABLE 3 ABOUT HERE.
Hypothesis 1 predicted that the proportion of exploratory innovation by an industry leader is
positively associated with its competitors’ performance improvement, manifest in sales growth.
In line with our expectations, we observe the positive effect of Leader Exploration on its
competitors’ Change in Sales in both time lags. Hypothesis 1 is therefore supported. Hypothesis
2 predicted that the relationship predicted in Hypothesis 1 is negatively moderated by the
Number of Firms in the Same Industry. We find strong support for this hypothesis in period t+1,
but this effect becomes weaker in t+2 although still significant. Concerning Hypothesis 3, we
find mixed results. Whereas the coefficient for Financial Slack is without statistical significance
in one-year lag, the coefficient for two-year lag is in the expected direction and is statistically
significant. Therefore, we find partial support for Hypothesis 3. The reason Financial Slack does
5 Given that our dependent variable concerns change in sales and not innovation, we do not have to restrict our
sample to firms that routinely innovate. It means that for some observations in our sample we will not observe
patenting activity. To perform this test, we replace the exploration value for non-patenting firms as zero.
17

not affect sales growth due to Leader Exploration in a one-year lag model may be because of the
time needed for competitors to learn through trial and error, which is required in addition to the
availability and commitment of necessary resources. In contrast, Hypothesis 4 predicted that the
current level of ROA negatively moderates the relationship in Hypothesis 1. We find support for
this hypothesis in the one-year lag models as ROA significantly and negatively moderates the
relationship between Leader Exploration and Change in Sales. However, this moderating effects
changes sign for the two-year lag. We speculate that this change is due to firms facing inertial
forces at first, but gradually able to leverage the benefits of the industry leader’s action, in line
with H1. Therefore, we found partial support to Hypothesis 4.
To interpret the size effects of the relevant explanatory variables, we estimate the
conditional marginal effects computing the elasticity of Leader Exploration with respect to
Change in Sales of competitors. We first predicted the values for the dependent variable with
Leader Exploration set at its mean. The results indicate that when Leader Exploration increases
by one standard deviation from its mean, its competitors’ sales increase by 36%. Examining the
moderators, we observe that when the Number of Firms in the Same Industry increases by one
standard deviation from its mean, the effect of Leader Exploration on the changes in
competitors’ sales decrease by 21%. When Financial Slack increases one standard deviation
from its mean, the effect of Leader Exploration on competitors’ sales increase by 23%. Finally,
the increase of a one standard deviation in ROA leads to a decrease in competitors’ sales by 16%.
Supplementary Analysis
We performed a series of additional tests to verify if the assumptions leading to our hypotheses
hold against alternative explanations. We ran the supplementary analysis using the dependent
variable Performance Improvement t+2 because two lags are more aligned with our expectations
regarding the timing effect between the industry leader’s investments in exploration and its
effects on competitors’ sales (see Table 4). First, given that our dependent variable concerns
changes in sales, we did not restrict the sample of competitors to the ones that actively innovate.
To verify if competitors that are more innovative are also more likely to improve their
performance from Leader Exploration, we computed the yearly patent stock for all competitors
in our sample and split the sample between firms that had no patent in a focal year (Model 1) and
those that had at least one patent successfully granted (Model 2). The comparison of the
18

coefficients in Model 1 and Model 2 indicate that innovative firms are significantly more likely
to grow their sales due to Leader Exploration. We also examined if the extent of an industry
leader’s dominance within the industry affects performance improvement of its competitors. We
computed the variable industry leader dominance based on the ratio between the industry
leader’s sales and the total sales of all other firms within an industry in a given year and split the
sample into two based on whether the ratio is greater than one. It indicates whether the industry
leader has more than half the market share in the industry. Model 3 reports the observation
within industries in which industry leaders have less than half the market share and Model 4 the
cases in which industry leaders have more than half the market share. The coefficient for Leader
Exploration is statistically significant for both models (p<0.05). However, the comparison of the
coefficients across the different models indicates that in industries with stronger dominance by
the industry leader, the investments in exploration have a stronger positive effect on the change
in sales for competitors. In other words, the benefits of an industry leader’s innovation on its
direct competitors are smaller when the leader is less dominant over close competitors. This
finding is consistent with the findings of Jacobides and Tae (2015), who highlighted that greater
inequality within an industry benefits the industry as a whole, vis-à-vis other industries in the
sector.
INSERT TABLE 4 ABOUT HERE.
We also examined how the degree of scope, measured by the participation in multiple SIC codes
may affect the results we observe as diversified firms may benefit from synergies that are not
available to non-diversified firms. In the analysis not reported here6, we split the sample into two
and re-ran the analysis. Contrary to what one would expect, we find largely consistent and
statistically significant results only for the subset of firms participating in only one SIC code. We
do not find any statistically significant results for the diversified firms. That diversified firms do
not benefit from an industry leader’s innovation may be because the halo effect that they can
enjoy is blurred by the presence in multiple SIC codes.
DISCUSSION
6 The results are available upon request from the authors.
19

We find that an industry leader’s innovative output affects its competitors’ performance,
measured in sales growth, in a sector comprising vertically related industries. By examining the
impact of exploratory innovation of industry leaders on their direct competitors, we highlight the
need to examine not only the innovation characteristics, but also the characteristics of the
innovators in order to gain deeper insights into the relationship between innovation and
competition. The extent of exploratory innovation undertaken by industry leaders and the number
of competitors in an industry influences competitors’ performance. It is, however, each
competitor’s own attributes such as availability of financial slack and level of ROA, that
determine the extent of performance improvement from those exogenous factors. We thus link
two different levels of analysis in this study: Heterogeneity in industry leaders’ innovation across
vertically related industries of a sector and heterogeneity among firms within each industry.
Our finding indicates that exploratory innovation of industry leaders matter: This is
noteworthy as industry leaders generally lack the motivation to actively seek novel knowledge in
their innovation. The positive relationship between Leader Exploration and Change in Sales also
implies that the shared benefits of innovation between the focal firm and its competitors is not
limited to the expanded knowledge pool (Grilliches, 1992; Yang et al., 2010), as has been
extensively studied in the past. The positive relationship also applies to financial performance, an
area that has not received much scrutiny. Consistent with studies that emphasize the role of broad
search (Katila and Chen, 2008; Phelps, 2010), we show that it is not the absolute amount of
innovation, but the proportion of a specific type of innovation, viz. exploratory innovation, that
has a positive effect on competitors’ performance. That is, not all innovation is equally effective
in creating enhancing sales for every participant.
Our findings also complement prior literature on the role of a few large, dominant firms,
i.e. industry leaders (de Figueiredo and Silverman, 2007; Jacobides and Tae, 2015). We echo the
extant literature that dominant firms or industry leaders have the ‘power’ to change the plight of
competitors (Wang and Shaver, 2014). In so doing, our focus was on the technological prowess
and strategic actions of industry leaders, manifest in their innovative output, rather than on the
market power they possess. The benefit competitors derive from their industry leaders, however,
is not out of altruistic motives: It is a by-product of industry leaders’ pursuit of self-interest
(Bhaskarabhatla, 2016; Jacobides et al., 2006). If industry leaders deem that their own gains
from their innovative activities are insufficient, for example, due to the cannibalization of their
20

own sales growth vis-à-vis those of their direct competitors, we can expect them to refrain from
such behavior. Such behavioral change, in turn, will have an effect on how the performance of
their competitors change.
We elucidate that underneath the industry-level performance improvement in aggregate
lies its uneven distribution at the firm level due to firm heterogeneity. The degree of how much
competitors can benefit from the actions of their industry leaders depends partly on the decisions
managers have made in the past or the capabilities firms have developed. A high level of
financial slack represents short-term resource underutilization (Nohria and Gulati, 1996), in lieu
of long-term flexibility. In contrast, high ROA reflects the effectiveness of a firm’s income
generation by efficiently leveraging its assets with little or no waste (Soliman, 2008). Consistent
with the inherent trade-off between flexibility and efficiency, we find that efficiency
maximization can hamper positive externalities while resource flexibility can amplify them. Our
findings show that some degrees of resource flexibility can facilitate firms to benefit from
actions of others, viz. industry leaders, and optimizing current operations can lead to inertia that
prevents firms from leveraging a given opportunity. That past decisions can have a material
impact on whether or not firms can fully take advantage of an opportunity alludes to the
dynamics of persistent firm heterogeneity.
This paper illustrates how this joint analysis can enhance our understanding of industry
dynamics and innovation. However, it is still subject to a number of limitations that also provide
avenues for future research. First, while we argue that industry leaders’ exploratory innovation
creates new opportunities, through which their competitors improve their performance, we do not
directly test the specific mechanisms at work. Our comprehensive data comprise all participants
in multiple SIC codes and their various performance and innovation measures, but we were
unable to gather data on specific opportunities for each industry for the period we cover. The
combined use of industry and patent data was enabling as much as constraining as the data allow
us to link exploratory innovation and performance, but do not provide us with measures to
examine whether and how an industry’s architectural advantage, reputation/status, and
knowledge pool expansion affect performance. We thus chose to articulate the mechanisms that
link industry leaders’ exploratory innovation with sales growth of competitors and then tested
their “reduced-form,” instead of their “structural,” implications and provide empirical
21

regularities. As noted in other sector-level studies, there is no easy way to remedy the problem,
but the shortcoming must be noted.
Second, we treat industry leadership as being static. We did not delve deeper into the
dynamics among industry leaders of vertically related industries. In the current study, we took
the heterogeneity as a given and did not explore why some industry leaders exhibit higher
proportion of exploratory innovation than others do. Examining where this heterogeneity across
industry leaders in a vertically related industries come from can thus yield valuable insights and
complement the current study. Relatedly, we do not consider situations where industry leadership
changes. This is partly because we did not observe such changes in our dataset, thus unable to
empirically examine such dynamics. However, it is likely that industry dethronement will
happen, especially in high-velocity or volatile settings (Ross and Sharapov, 2015; Smith, Ferrier,
and Grimm, 2001). It will be a fruitful avenue of research to consider how dethronement may
affect the relationships discussed in the current study.
Third, we only focused on one dimension of innovation to establish a relationship
between industry leaders and their competitors’ performance. However, other innovation
characteristics of industry leaders can also influence their competitors’ performance. Identifying
those characteristics and examining which characteristic dominates in explaining competitors’
performance change can deepen our understanding of industry leaders’ influence. Extending the
current study, future work can also examine whether the sales growth of competitors come at the
expense of, or in addition to the sales growth of industry leaders, i.e. cannibalization or growth of
the total pie size. Relatedly, it will be interesting to see if the innovation characteristics have any
effect on these outcomes.
Finally, we only include publicly traded firms in our dataset, as we did not have
information on private firms whose inclusion might have changed the results. Because well-
known industry leaders in the computer sector, such as IBM, Microsoft, and Intel, have been
public for a long period of time, we believe omitting private competitors from our data provides
conservative results rather than biased results. Regarding the analysis, we do not hypothesize or
test the joint effect of slack resources and ROA in tandem, although some firms may be able to
accumulate high levels of slack while being efficient with their current operations. Because our
primary interest here was to understand if exploratory innovations of industry leaders leads to
22

performance improvement of their competitors, we opted not to explore the complexity observed
within each competitor or its effect on performance.
This study suggests that one firm’s effort to further its interest can help other firms with
their interests and that this effect is most salient in vertically related industries. The findings of
this study highlight that the joint study of heterogeneity across vertically related industries and
that among firms within each industry can yield valuable insights into which industries perform
better as they evolve and which firms outgrow others. The aggregated industry ramifications
deserve careful scrutiny as they can have important competitive repercussions throughout the
sector, given the inherent interdependence across vertically related industries.
23

REFERENCES
Abernathy, W.J. and J.M. Utterback. 1978. Patterns of industrial innovation. Technology Review
64(7): 254-228.
Adner, R. and D.A. Levinthal. 2001. Demand heterogeneity and technology evolution:
Implications for product and process innovation. Management Science 47(5): 611-628.
Adner, R. and P. Zemsky. 2005. Disruptive technologies and the emergence of competition. The
RAND Journal of Economics 36(2): 229-254.
Adner, R. and P. Zemsky. 2006. A demand‐based perspective on sustainable competitive
advantage. Strategic Management Journal 27(3): 215-239.
Agarwal, R. and B.L. Bayus. 2002. The market evolution and sales takeoff of product
innovations. Management Science 48(8): 1024-1041.
Agarwal, R., and M. Gort. 1996. The evolution of markets and entry, exit and survival of firms.
Review of Economics and Statistics 78(3): 489-98.
Ahuja, G. 2000. Collaboration networks, structural holes, and innovation: A longitudinal study.
Administrative Science Quarterly 45(3): 425-55.
Almeida, P. 1996. Knowledge sourcing by foreign multinationals: Patent citation analysis in the
US semiconductor industry. Strategic Management Journal 17(S2): 155-165.
Angrist, J.D. and J.S. Pischke. 2008. Mostly harmless econometrics: An empiricist's companion.
Princeton, NJ: Princeton University Press.
Bhaskarabhatla, A. 2016. The moderating role of submarket dynamics on the product
customization–firm survival relationship. Organization Science 27(4): 1049-1064.
Bresnahan, T.F. and S. Greenstein. 1999. Technological competition and the structure of the
computer industry. The Journal of Industrial Economics 47(1): 1-40.
Brusoni, S., Prencipe, A. and K. Pavitt. 2001. Knowledge specialization, organizational coupling,
and the boundaries of the firm: why do firms know more than they make? Administrative
Science Quarterly 46(4): 597-621.
Cheng, J.L.C., and I.F. Kesner. 1997. Organizational slack and response to environmental shifts:
The impact of resource allocation patterns. Journal of Management 23(1): 1-18.
Christensen, C.M. 1997. The Innovator's Dilemma: When New Technologies Cause Great Firms
to Fail. Cambridge, MA: Harvard Business Review Press.
Cohen, W.M. and D.A. Levinthal. 1990. Absorptive capacity: A new perspective on learning and
innovation. Administrative Science Quarterly 35(1): 128-152.
Cohen, W.M., Nelson, R.R., and J.P. Walsh. 2000. Protecting their intellectual assets:
Appropriability conditions and why U.S. manufacturing firms patent (or not). NBER
Working Paper No. 7552.
de Figueiredo, J.M. and B.S. Silverman. 2007. Churn, baby, churn: Strategic dynamics among
dominant and fringe firms in a segmented industry. Management Science 53(4): 632-650.
Danneels, E. 2002. The dynamics of product innovation and firm competences. Strategic
Management Journal 23(12): 1095-1121.
Dierickx, I. and K. Cool. 1989. Asset stock accumulation and sustainability of competitive
advantage. Management Science 35(12): 1504-1511.
Dosi, G. 1982. Technological paradigms and technological trajectories: A suggested
interpretation of the determinants and directions of technical change. Research Policy 11(3):
147-162.
Ethiraj, S.K. and D.H. Zhu. 2008. Performance effects of imitative entry. Strategic Management
Journal 29(8): 797-817.
24

Fleming, L. and O. Sorenson. 2004. Science as a map in technological search. Strategic
Management Journal 25(8-9): 909-928.
Gawer, A. and M.A. Cusumano. 2002. Platform Leadership: How Intel, Microsoft, and Cisco
Drive Industry Innovation. Boston, MA: Harvard Business School Press.
George, G. 2005. Slack resources and the performance of privately held firms. Academy of
Management Journal 48(4): 661-676.
Gereffi, G., Humphrey, J. and T. Sturgeon. 2005. The governance of global value chains. Review
of International Political Economy 12(1): 78-104.
Giarratana, M.S. 2004. The birth of a new industry: Entry by start-ups and the drivers of firm
growth: the case of encryption software. Research Policy 33(5): 787-806.
Griliches, Z. 1992. The search for R&D spillovers. Scandinavian Journal of Economics 94(1):
29-47.
Hall, B.H., Jaffe, A. and M. Trajtenberg. 2005. Market value and patent citations. RAND Journal
of Economics 36(1): 16-38.
Jacobides, M.G., Knudsen, T., and M. Augier. 2006. Benefiting from innovation: Value creation,
value appropriation and the role of industry architectures. Research Policy 35(8), 1200-1221.
Jacobides, M.G. and C.J. Tae. 2015. Kingpins, bottlenecks, and value dynamics along a sector.
Organization Science 26(3): 889-907.
Jaffe, A.B., Trajtenberg, M. and R. Henderson. 1993. Geographic localization of knowledge
spillovers as evidenced by patent citations. Quarterly Journal of Economics 108(3): 577-598.
Jansen, J., Van Den Bosch, F., & Volberda, H. 2006. Exploratory Innovation, Exploitative
Innovation, and Performance: Effects of Organizational Antecedents and Environmental
Moderators. Management Science 52(11): 1661-1674.
Kafouros, M.I. and P.J. Buckley. 2008. Under what conditions do firms benefit from the research
efforts of other organizations? Research Policy 37(2): 225-239.
Kapoor, R. and Klueter, T. 2015. Decoding the Adaptability--Rigidity Puzzle: Evidence from
Pharmaceutical Incumbents’ Pursuit of Gene Therapy and Monoclonal Antibodies. Academy
of Management Journal 58(4): 1180–1207.
Katila, R. and G. Ahuja. 2002. Something old, something new: A longitudinal study of search
behavior and new product introduction. Academy of Management Journal 45(6): 1183-1194.
Katila, R. and G. Ahuja. 2004. Where do resources come from? The role of idiosyncratic
situations. Strategic Management Journal 25(8-9): 887-907.
Katila, R. and E.L. Chen. 2008. Effects of search timing on innovation: The value of not being in
sync with rivals. Administrative Science Quarterly 53(4): 593-625.
Kaul, A. 2012. Technology and corporate scope: Firm and rival innovation as antecedents of
corporate transactions. Strategic Management Journal 33(4): 347-367.
Kenney, M. and B. Pon. 2011. Structuring the smartphone industry: Is the mobile internet OS
platform the key? Journal of Industry, Competition and Trade 11(3): 239-261.
Klepper, S. 1996. Entry, exit, growth, and innovation over the product life cycle. American
Economic Review 86(3): 562-83.
Knott, A.M. 2003. Persistent heterogeneity and sustainable innovation. Strategic Management
Journal 24(8): 687-705.
Laursen, K., Leone, M.I., and S. Torrisi. 2010. Technological exploration through licensing: New
insights from the licensee's point of view. Industrial and Corporate Change 19(3): 871-897.
25

Levinthal, D.A. and B. Wu. 2010. Opportunity costs and non-scale free capabilities: profit
maximization, corporate scope, and profit margins. Strategic Management Journal 31(7):
780-801.
Lin, Z., Yang, H. and B. Arya. 2009. Alliance partners and firm performance: Resource
complementarity and status association. Strategic Management Journal 30(9): 921-940.
Malhotra, A., Gosain, S. and O. Sawy. 2005. Absorptive capacity configurations in supply
chains: gearing for partner-enabled market knowledge creation. MIS Quarterly 29(1): 145-
187.
Mishina, Y., Pollock, T.G., and J.F. Porac. 2004. Are more resources always better for growth?
Resource stickiness in market and product expansion. Strategic Management Journal 25(12):
1179-1197.
Mitchell, M. and A. Skrzypacz. 2015. A theory of market pioneers, dynamic capabilities, and
industry evolution. Management Science 61(7): 1598-1614.
Moatti, V., Ren, C.R., Anand, J. and P. Dussauge. 2015. Disentangling the performance effects of
efficiency and bargaining power in horizontal growth strategies: An empirical investigation
in the global retail industry. Strategic Management Journal 36(5): 745-757.
Moreira, S., Markus, A and K. Laursen. 2018. Knowledge diversity and coordination: The effect
of intrafirm inventor task networks on absorption speed. Strategic Management Journal
39(9): 2517-2546.
Moreira, S., Cabaleiro, G. and Reichstein, T., 2018. Licensing decision: a rent dissipation lens
applied to product market competition, openness to external knowledge and exogenous sunk
costs. Industrial and Corporate Change. dty036
Nelson, R.R. and S.G. Winter. 1982. An evolutionary theory of economic change. Cambridge,
MA: Belknap.
Nerkar, A. and S. Paruchuri. 2005. Evolution of R&D capabilities: The role of knowledge
networks within a firm. Management Science 51(5): 771-785.
Nerkar, A. and P.W. Roberts. 2004. Technological and product-market experience and the success
of new product introductions in the pharmaceutical industry. Strategic Management
Journal 25(8-9): 779-799.
Nohria, N. and R. Gulati. 1996. Is slack good or bad for innovation? Academy of Management
Journal 39(5): 1245-1264.
Peteraf, M.A. and J.B. Barney. 2003. Unraveling the resource-based tangle. Managerial
Decision Economics 24(4): 309-323.
Phelps, C.C. 2010. A longitudinal study of the influence of alliance network structure and
composition on firm exploratory innovation. Academy of Management Journal 53(4): 890-
913.
Powell, W.W., Koput, K.W., and L. Smith-Doerr. 1996. Interorganizational collaboration and the
locus of innovation: Networks of learning in biotechnology. Administrative Science
Quarterly 41(1): 116-45.
Singh, J.V. 1986. Performance, slack, and risk taking in organizational decision making.
Academy of Management Journal 29(3): 562-585.
Soliman, M.T. 2008. The use of DuPont analysis by market participants. Accounting Review
83(3):823-853.
Spithoven, A., Frantzen, D., and B. Clarysse. 2010. Heterogeneous firm-level effects of
knowledge exchanges on product innovation: Differences between dynamic and lagging
product innovators. Journal of Product Innovation Management 27(2): 362–381.
26

Stuart T. 2000. Interorganizational alliances and the performance of firms: A study of growth and
innovation rates in a high-technology industry. Strategic Management Journal 21(8): 791-
811.
Teece, D.J. 1986. Profiting from technological innovation: Implications for integration,
collaboration, licensing and public policy. Research Policy 15(6): 285-305.
Thompson, J.D. 1967. Organizations in Action. New York, NY: McGraw-Hill.
Troilo, G., De Luca, L. M. and K. Atuahene-Gima. 2014. More innovation with less? A strategic
contingency view of slack resources, information search, and radical innovation. Journal of
Product Innovation Management 31(2): 259–277.
von Hippel, E. 1988. The Sources of Innovation. Oxford, UK: Oxford University Press.
Wang, R.D. and J.M. Shaver. 2014. Competition-driven repositioning. Strategic Management
Journal 35(11): 1585-1604.
Wooldridge, J. 2009. Introductory Econometrics: A Modern Approach. Boston, MA: South-
Western Cengage Learning.
Yang, H., Phelps, C. and H.K. Steensma. 2010. Learning from what others have learned from
you: The effects of knowledge spillovers on originating firms. Academy of Management
Journal 53(2): 371-389.
Ziedonis, R.H. 2004. Don't fence me in: Fragmented markets for technology and the patent
acquisition strategies of firms. Management Science 50(6): 804-820.
27

TABLES
28

29

30