Technological Diversification in Litigation Lawsuits:
Does Diversification Increase the Risk of Being
Litigated?
Steven Mak
1920596
MSc BA Strategic Innovation Management
Supervisor: Dr. F. Noseleit
Co-assessor: Dr. P. M. M. de Faria
January 20th 2015
Word count: 12.205
Abstract:
Not all firms face similar risk of being litigated in a patent lawsuit. This study examines the effect of technological diversification on the risk of being a defendant in litigation lawsuit. Based on previous research, it is expected that (1) as the diversification of the firm increases the risk of litigation rises, (2) small firms are in a disadvantaged position when diversification increases, and (3) firms diversified in related industries face the highest risk of being involved in a litigation lawsuit. Based on a sample of 126 firms in the Semiconductors, Mobile Communications and Devices, and Consumer Electronics and PC’s industry, results show that the risk of litigation increases as diversification increases, small firms face a higher risk of being involved in a litigation lawsuit when diversification increases, and that firms diversified in unrelated industries face the highest risk of being litigated. Using the result of this study, managers can better assess the risk of being litigated and can take proper action accordingly.
Table of contents
Introduction 3
Literature Review 5
Background 5
Diversification 8
Data and Methodology 11
Introduction
During the 80’s and 90’s the number of patent filings in the US increased almost 10-fold, where the latter decade is accountable for the largest part of this increase (Lanjouw & Schankerman, 2004; Bessen & Meurer, 2005b; Cohen et. al., 2000). With this ever-increasing number of patent applications also comes an increase in patent infringement: the number of patent lawsuits has doubled during the 90’s (Bessen & Meurer, 2005a). More and more companies try to protect their patents and are actively searching for possible infringement of their patents (Sharipo, 2000). What is even more interesting is the fact that in the period from 1979 till 1995, 45% of the patent infringement cases were won by the plaintiff (Sherry & Teece, 2003). So the appearance of the protection of the property in intellectual property is not that certain as it is expected to be and the outcome of a litigation case is difficult to predict ex ante (Bessen & Meurer, 2005).
Costs associated with patent litigations have also risen dramatically during the 80’s and 90’s (Bessen & Meurer, 2008; Kash & Kingston, 2001) but they do not form a barrier for organizations to apply for a patent (Cohen et. al., 2000). This implies a discrepancy between the determination of costs associated with litigation and the risk of encountering these costs.
An important distinction on that matter is needs to be made between plaintiffs and defendants because a lot of the burden of patent disputes falls on defending firms (Bessen & Meurer, 2005a). Not only direct cost associated with the case like attorney fees but also costs in a more indirect manner, for example the response on the valuation of the firm by the market, disruption of the firm’s day-to-day business as resources are needed for the lawsuit, and impediment of the research and development efforts both within and outside the organization are all larger for defending firms (Raghu et. al., 2008; Bessen & Meurer, 2008; Lerner, 1995). Although the rate of litigation per patent among public firms as plaintiffs did not increase much from 1987 to 1999, the rate of litigation per R&D dollar among public firms as defendants increased 70% (Bessen & Meurer, 2005a). Differences in litigation rate exist between firms in terms of the risk they face. Gaining more insight in these differences will be beneficial for mangers, which will give them an indication about the risk they face. Previous research has already identified several of these organizational characteristics such as firm size, portfolio size, and R&D intensity (Lanjouw & Schankerman, 2001; Lanjouw & Schankerman, 2004; Garcia-Vega, 2006).
However, there are limitations to the amount of diversification and the relatedness of the knowledge. Too much diversification into unrelated industries will lead the firm to perform worse (Palich et. al., 2000). So differentiation into related industries create opportunities to use knowledge created in a specific industry and apply it to other industries in which the firm is differentiated (Pavitt, 1991).
The purpose of this study is creating more insight in the organizational characteristics influencing the litigation risk firms face.
The significance of the study lies in the fact that managers will be better able to estimate the risk that their organization faces in being involved in costly litigation cases. Suitable actions can be taken accordingly such as searching more actively for possible infringement, increase R&D activity, and creating a sizable patent portfolio. Also decisions whether to diversify a firm or not and the influence it has on the products produced by the firm, the patents applied for by the firm and the risk of litigation can be made more thoroughly with the results of this research. Gaining insight in litigation costs of the firm and the risk a firm faces to encounter these costs when considering a diversification decision can be very important as these costs keep rising (Bessen & Meurer, 2008). These estimations about the litigation risk can be made based on the firms’ technological knowledge base.
Therefore I developed the following research question:
What are the differences between single industry firms vs. diversified firms in terms of patent litigation?
Secondary research questions:
1. What are the differences between single industry firms vs. diversified firms? 2. Have diversified firms a higher risk of being litigated than single industry firms?
Literature Review Background
The rising patent application originated in the early 80’s and is the result of various patent policy reforms by the US government (Bessen & Meurer, 2005a). Since then patents are better protected by law, which resulted in less invalid patents but also in less infringed patents when disputes appear in court (Lunney, 2004), creating an incentive to apply for more patents in order to better protect a firm’s innovations. As stated by Lanjouw & Schankerman (2001):
“Innovation is at the heart of productivity growth, and granting inventors property rights through the patent system is the most widespread policy to encourage innovation” (p. 130).
However, validity and boundaries of the patents are not always clear, which can lead to a disincentive for a firm to invest in R&D (Bessen &Meurer, 2005a). The increased patent application and patent litigation lawsuits question the incentive to innovate patent protection is assumed to have (Bessen & Meurer, 2005a; Hall, 2007).
As a result, in most industries patents are not considered the main mechanism to protect innovations (Cohen et. al., 2000). Managers indicated other mechanisms like lead-time advantages and secrecy to be more effective in protecting their intellectual property. However, large differences exist across industries in their reliance on patents (Arundel & Kabla, 1998; Schankerman, 1998). These differences exist because the innovation patterns are idiosyncratic across industries and path dependent and therefore unstable and dynamic (Park, Yoon & Lee, 2005). As a consequence, the number of patent applications also differs across industries (Schankerman, 1998; Lanjouw & Schankerman, 1997). Patents were originally used in growing industries to incentivize firms on R&D spending but in many industries this role has changed dramatically (Kash & Kingston, 2001). Industries that do rely on patents and those that do not share some common characteristics.
One of these characteristics that is highly useful for the purpose of this research, is the distinction on the basis of the underlying knowledge that is required by the product in order to function properly (Kash & Kingston, 2001). Two types of technologies can be distinguished: Discrete technologies that require only a few patents to function, which is the case in the pharmaceutical industry and complex technologies that are characterized by products that comply numerous patentable elements1.
Complex technologies, such as computers, semiconductors and telecommunications, often need to rely on multiple patents, owned by multiple actors both within and outside the industry in which the firm is active in order to function properly (Sharipo, 2000; Hall & Ziedonis, 2001; Hall & Ham, 1999; Grindley & Teece, 1997; Kash & Kingston, 2001). These industries patent very aggressively and rely much more on patents than other manufacturing industries, both in terms of utility and design patents (Hall & Ham, 1999; Cohen et. al., 2000; Census Bureau, 2013). They also have different motives to patent: the cumulative nature of the
patents in these industries make firms rely on other patents more heavily (Hall & Ham, 1999; Cohen et. al., 2000).
Firms choose to patent but also not to patent for different reasons that differ across industries and technological areas (Cohen et. al., 2000). The complex industries use patents for (1) the prevention of copying, (2) patent blocking, (3) the prevention of lawsuits, and (4) generating revenue through patent licensing. The first three can be considered defensive in nature, which are different motivations than the assumptions on which the patent system is based (Kash & Kingston, 2001). Today, patents are both used to threaten possible infringers when the patent portfolio comprises essential parts of a technology, and as a defense mechanism when companies sometimes file countersuits, where they claim the plaintiff is the one infringing the defendants patents (Cohen et. al., 2000; Somaya, 2003). When identifying the reasons not to use patents to protect their intellectual property, managers indicated that the (1) ease of inventing around, (2) information disclosure, and (3) lack of novelty were their biggest concern (Cohen et. al., 2000).
Costs associated with patent litigations do not form a barrier for organizations in applying for patents but have also raised dramatically during the 80’s and 90’s (Cohen et. al., 2000; Bessen & Meurer, 2008). This implies a discrepancy between the determination of costs associated with litigation and the risk of encountering these costs. In light of this assumption, it is important to make the distinction between plaintiffs and defendants, because a lot of these costs are borne by the defending firms (Bessen & Meurer, 2005a). Not only direct cost associated with the case like attorney fees but also costs in a more indirect manner, for example the response on the valuation of the firm by the market, disruption of the firm’s day-to-day business as resources are needed for the lawsuit, impediment of the research and development efforts both within and outside the organization are larger for defending firms (Raghu et. al., 2008; Bessen & Meurer, 2008; Lerner, 1995). Although the rate of litigation per patent among public firms as plaintiffs did not increase much from 1987 to 1999, the rate of litigation per R&D dollar among public firms as defendants increased by 70% (Bessen & Meurer, 2005a). Plaintiffs suffer losses as well which are, however, not as substantial as costs encountered by the defendants. These costs are even more substantial for smaller firms as they have a lower ability in protecting their patents.
size of the portfolio. The value of a portfolio of patents from a single industry is more suitable to settlement than a diversified patent portfolio.
As a result, accelerating patent application takes place in order to gain a large patent portfolio, leading toward patent races, which is defined as the competition between firms in trying to gain as much patents as possible (Hall & Ziedonis, 2001). Also generating revenue out of patent licensing through large numbers of patents increases the rate of patent applications and is called patent mining (Sharipo, 2000): firms applying for many patents and trying to create value out of them by asserting them aggressively by searching actively for infringement and defending them in court, even if the alleged infringer is not a direct competitor. The first firm using patents in this distinctive manner was Texas Instruments, the US electronics company, in the 80’s and increased its patent application ever since (Sharipo, 2000). This hinges to a more strategic use of patents instead of just protecting a firm’s intellectual property, and previous research indicated that this is indeed a reason to apply for patents by the most active patenting firms (Dolfsma, 2011; Cohen et. al., 2000). The effect of defendants portfolio size is found to lower the risk of litigation and is stronger for smaller firms, making patent portfolio a viable way to strengthen position of a smaller firm in settling disputes and avoiding litigation (Bessen & Meurer, 2005a, Lanjouw & Schankerman, 1997/2004). The problem that arises with the increase in portfolio size is the firms’ inability to check for infringements of individual patents. One way of reducing this problem is portfolio licensing, where all patents in a specific field are licensed to a firm at once (Grindley & Teece, 1997). This reinforcing structure also partially explains the increase in patenting activity during the last three centuries (Hall & Ziedonis, 2001). The large amount of patents and the cumulative nature of the industries technologies are not only difficult for the patent owners but also for other firms within the industry (Bessen & Meurer, 2005a). The image exists is that they are pirates and just copy the technologies developed by others, however, most of the time they are innovators as well, investing as much in R&D as the plaintiffs. This indicates again the difficulty in searching through the patent thicket, resulting in the fact that often firms unknowingly infringe another’s patents, creating a potential for infringement lawsuits involving large costs, both direct and indirect (Sharipo, 2000). Other reasons why firms are unaware of the patents they infringe are that patents have not been issued yet, inadequate investigation of possible infringement, and an unclear scope of the patent right (Bessen & Meurer, 2005a).
The possible consequences of patent disputes indicate the difficulty in solving them. One of these is the damaged relationship with the other party that can harm future collaborative R&D efforts. Also the delay in access to indispensable components can incur substantial cost, especially for technologies with a short life-cycle (Kash & Kingston, 2001). Private costs associated with firms listed on the stock exchange are substantial and have a much larger effect on the plaintiff as it has on the defendant (Bessen & Meurer, 2008). They can reaching up to 19% of a firms R&D spending and is estimated to exceed the value of the patents granted.
industries (Bessen & Meurer, 2008). However, the relative costs associated with the defense of patents are disproportionately accountable for the smaller firms as they bare similar cost as large firms but can only spread their costs over a small output (Cohen, et al 2000; Kash & Kingston, 2001).
Another factor influencing the cost and risk of litigation is the country of origin. Domestic firms will be better able to protect their patents than foreign firms (Lanjouw & Schankerman, 1997/2004). The reason is twofold. Firstly, it is harder for foreign companies to detect infringement. They are less familiar with the country, its regulations, and industry characteristics than domestic firms, which put them in a disadvantageous position. Secondly, if infringement is found, the costs of pursuing litigation are higher for domestic firms. Cost of, among others, translating of documents, and communication, are added to the fixed costs of the lawsuit, also encountered by the domestic counterpart (Lanjouw & Schankerman, 2004). This result in a lower likelihood of involvement of foreign owned companies in litigation lawsuits.
Diversification
Important in the distinction between a diversified firm, defined as a firm active in multiple industries and a single, specialized firm, defined as a firm active in a single industry, is the knowledge that is generated and used in the firms.
specialized firms as they rely solely on a single industry, which makes settlement and strong relationship important (Lanjouw & Schankerman, 2004; Bessen & Meurer, 2008). Specialized firms make up around 70% of the innovating firms but only represent 10% of all the patent applications (Breschi et. al., 2003). Therefore I expect that as the number of industries increases, the necessity for a firm to settle becomes less relevant, as they are less dependent on relationships in a single industries, which results in a higher likelihood of being litigated:
H1: Increase in the number of industries in which firms are active, will increase the risk of being involvement in litigation lawsuits as a defendant.
A factor that would interact with the number of industries is firm size. For small firms it is more difficult to protect their intellectual property rights than for their bigger counterparts do due to, among others, their limited financial capacity, use of external legal counseling, and lower output (Lanjouw & Schankerman, 2004). The latter is important because the fixed legal cost, encountered by both parties, are spread over the lower output, which put them in a disadvantaged position and will lower the incentive to invest in R&D and to use patent protection as a mechanism to protect the innovations (Cohen et. al., 2000; Bessen & Meurer, 2008). On the other hand, small firms use patents as a mean to enhance the firm’s reputation in order to acquire financing or alliance partners, creating an incentive to patent (Cohen et. al., 2000). Larger firms do not indicate this as a reason to use patents, implying that they are in a stronger position for other reasons when negotiating settlement. Both the lower output and a strong reputation increase the willingness to settle the dispute without a lawsuit, which put the large firms in a natural advanced position. Combining these mechanisms with their higher litigation costs, results in a lower probability for small firms to pursue patent infringement (Cohen et. al., 2000; Bessen & Meurer, 2008; Lanjouw & Schankerman, 2004). These disadvantages in litigation costs for small firms, both as plaintiff and defendant, are also found to be a reason not to patent (Cohen et. al., 2000). Small firms are in an even more disadvantaged position when diversification increases. Within a single industry, firms are more likely to interact with each other, which increases the likelihood of settlement (Lanjouw & Schankerman, 2004). Although small firms are in a disadvantaged position in a single industry, as the number of industries grow there are even more drawbacks. Larger firms have a better ability to interact with more firms in different industries due to their size. Small firms will have lower interaction within a single industry as well as in multiple industries but these interactions become more difficult when diversification increases. Other reasons are increasing coordination costs and difficulties in detecting infringement (both as the one infringing another firms patents as well as infringement of their own patents) that make it even more inconvenient for small firms as opposed to larger ones (Miller, 2006). Therefore I expect that as the number of industries in which the firm is active increases, the difference in litigation risk between small and larger firms is expected to get bigger.
H2: As diversification increases, the difference in the risk of being involved in a litigation lawsuit as a defendant between small and large firms gets bigger.
Benavides-Velasco, 2008; Ettlie et. al., 1984). Distant knowledge may lead to new, path-breaking inventions that disrupt the market (Miller, Fern & Cardinal, 2007) and will have a higher risk of litigation due to their novelty and uncertainty (Lanjouw & Schankerman, 2004). Radical innovation takes place in the beginning of a new technology’s development, forming the basis of the technology, facilitating subsequent more incremental innovations (Haubt et. al., 2007; Dolfsma, 2011). These base technologies have a tendency to have a higher likelihood of being involved in a litigation lawsuit (Lanjouw & Schankerman, 2001).
Empirical findings also indicate that firms are more likely to diversify into related industries than unrelated ones due to the switching cost involved in acquiring unrelated knowledge (Breschi et. al., 2003). Related industries share some knowledge and/or scientific principles that create opportunities to use knowledge created in a specific industry and apply it to other industries in which the firm is differentiated (Pavitt, 1991). Firms that are not differentiated or differentiated into unrelated industries do not have the ability to use related knowledge within the organizational borders, and as a consequence have to search outside their own organization to be able to access new knowledge.
However, there are limits to the effect diversification has on firm performance. Too much diversification into unrelated industries will lead the firm to perform worse as it becomes difficult to learn from the partner (Palich et. al., 2000). This result was only supported by financial measures but also innovative performance tends to behave in similar direction for strategic alliances (Sampson, 2007). On the other hand, when the knowledge is highly similar to the existing knowledge base, there are no opportunities to combine them into new innovations.
Bessen & Meurer (2005a) found that around a quarter of the patent litigations lawsuits involved two firms in distant, unrelated industries, indicating that most of the disputes are resolved between direct competitors or at least firms that are active in related industries. In this finding lies an important distinction between a single industry firm and a multi-industry firm as the multi-industry firm has, in terms of the increased connection between the firms, such as direct competition between the firms, blocking the other firm as a possible competitor, possibilities cross-licensing agreements with complementary patents. Of all patent lawsuits, 29% of the litigation takes place between firms in the same industry (Bessen & Meurer, 2005a). 28% of the lawsuits take place between unrelated firms, which leaves 43% for the related firms. Being technologically close to more firms in related industries increases the risk of being litigated. Comparing single industry firms, related and unrelated diversified firm in their risk of litigation results in a third hypothesis:
H3: Related diversified firms will be involved in most litigation lawsuits as a defendant.
Data and Methodology
The data was collected using the RPX Corporation Client Portal on patent litigation and provided a list of 360 companies in the Semiconductors, Mobile Communications and Devices, and Consumer Electronics and PC’s industry, selected by involvement as a defendant in three or more litigation cases during the period 01-01-2010 till 31-12-2012 (RPX Corporation, 2014). Selection was based on publicly listed companies, as they require to provide company information publicly which was needed to analyze organizational characteristics. I took 2012 as a basis and gathered several organizational figures through Orbis Company Info (Bureau van Dijk Electronic Publishing, 2014a). Data on the number of patents granted by the USPTO was gathered through and USPTO database and statistics website (uspto.gov). Further selection was based on the availability of the data and resulted in a list of 126 organizations.
The NAICS classification codes were generated by analyzing the IP licensing agreements between companies and were provided through Orbis (however, the actual analysis was done by another party), the company information website (Bureau van Dijk Electronic Publishing, 2014a). As this classification is generated through analyzing IP licensing agreements, it provides information into the intellectual property that is owned by the firm. This makes it a useful tool for analyzing a firm’s underlying technological knowledge base.
One additional remark regarding the data needs to be placed on the existence of subsidiaries. These separate entities are part of a corporate group where some activities take place at the subsidiary level and others at the corporate level. The data used in this research was based on the corporate level but the revenues of the subsidiaries are part of the revenue measure of the corporation. However, patents are not measured at the corporate level, which may result in biased data on this measure. Several other measures and the availability of technological knowledge throughout the corporation may be difficult to capture as the question whether or not to include the subsidiary figures in the corporate group differs across firms. For this research the problem is expected to be of minimal influence, as the technological knowledge is not expected to flow throughout the corporation to subsidiaries very easily as it is a separate business unit.
Independent Variables
Diversification. This variable measures the number of industries in which the firm is active. The
Share of related diversification. As a measuring diversification of a firm, the share of
diversification indicates the relatedness of the industries in which the firms is diversified. The primary NAICS code functions as the basis from where the level of diversification is measured. If the secondary NAICS code, independent of the number of industries, falls within the three digits of the primary NAICS code, it is considered related. If it falls outside of the three-digit range, it is considered unrelated. The level of diversification is computed as the number of related industries divided by the total number of industries (primary excluded as it does not relate to diversification). This generates a decimal number between 0 and 1, where 0 is completely unrelated diversification and 1 is completely related diversification. The single industry firms, without any secondary NAICS code, were granted a 0, as they are not diversified at all. One additional remark needs to be placed on the coding of the variable. Most of the firms are active within the range of the NAICS 334 code (Computer and Electronic Product Manufacturing). However there is another group within which the knowledge is expected to be highly similar to the knowledge in that industry, which is NAICS code 5415 (Computer Systems Design and Related Services). Based on its code description2 I control for this group by identifying this industry as being a related industry. For example, if a firm has primary code 3341 and a secondary code 5415, they still were identified as being related. By controlling for this industry, firms as if the firm diversified in a related industry, although it cannot be identified as such based on the codes, as it does not fall within the three-digit range. For all other industries that did not have similar secondary industry codes (three digit range) as the primary industry code, firms were identified as being unrelated diversified.
Dependent Variable
Number of defendant cases. As the dependent variable and the measure of litigation, the number of
cases the firm is involved in a litigation case will indicate the risk of litigation. This was obtained by searching the database in the earlier mentioned industries (Semiconductors, Mobile Communications and Devices, and Consumer Electronics and PC’s) and selecting firms that where involved in patent litigation as a defendant three times or more.
The literature makes a distinction between two types of defendant cases: an infringement case and a validity case, where the plaintiff in the former claims the defendant infringes one or more patents owned by the plaintiff and the latter that the patents owned by the defendant are invalid (Sherry & Teece, 2003). So the difference between the two lies within the ownership of the patents. There is however no significant qualitative difference between the two types of cases (Lanjouw & Schankerman, 2001). For the purpose of this research no distinction is made between the two types of cases. The underlying theory of knowledge differences affect both the patents owned by the defendant as well as possible plaintiffs.
2
Control Variables
For all variables marked with a * I use the logged value in order to reduce the influence of outlier values and thereby increase reliability of the regression analysis. For the merged variables I first computed the original value and then created the logged value. Most of the control variables have been discussed extensively in the literature review so I will only elaborate on them briefly.
Firm size*. Small firms have more difficulty in protecting their intellectual property right than do
their bigger counterparts (Lanjouw & Schankerman, 2004). Costs associated with the dispute are higher for the smaller firm because they usually also have a lower output (Cohen et. al., 2000; Bessen & Meurer, 2008). In this research I will use revenue as a measure for firm size as larger firms can spread their litigation cost over a higher output in comparison to a smaller counterpart (Cohen et. al., 2000). Firm size is expected to lower the risk of involvement in litigation lawsuits as a defendant.
Plaintiff cases*. The image exists is that defendants in litigation cases are pirates and just copy the
technologies developed by others, however, most of the time they are innovators as well, investing as much in R&D as the plaintiffs (Bessen & Meurer, 2005a). Patents are both used to threaten possible infringers, as the other party infringes the intellectual property of the firm, and as a defense mechanism as companies sometimes file countersuits where they claim the plaintiff is the one infringing the defendants patents (Cohen, et al, 2000; Somaya, 2003). Controlling for the number of times the firm is active as a plaintiff in litigation cases will check for the firms’ activity in filing countersuits or being the one that is more active in protecting their intellectual property instead of being the one that infringes some other firms’ patents. So it is expected that the higher the number of plaintiff cases will also increase the number of defendant cases.
Patent portfolio size*. The influence of the defendant’s portfolio size is found to lower the risk of
litigation (Bessen & Meurer, 2005a, Lanjouw & Schankerman, 1997, 2004). Firms that use patents the most are trying to protect them against suits and use these patents in negotiations (Cohen et. al., 2000).
Patents are used to negotiate in patent disputes and can result in the exchange of licenses to settle in these cases, making a larger patent portfolio a mean to better cope with patent infringement (Hall & Ham, 1999; Kash & Kingston, 2000). Patent portfolio size is measured by the total patents the USPTO has granted the firm throughout the years, starting in 1976 when the recording of patents started. Patent portfolio size is expected to lower the involvement in patent litigation lawsuits.
R&D intensity. Technologically diversified firms tend to spend more on R&D and the increase in
diversity in technology portfolio tends to increase a firms’ innovativeness (Garcia-Vega, 2006). Further results of that research suggest that R&D expenditures and sales are highly related. Also, as diversity increases, R&D costs rise, as well as sales (Granstrand & Oskarsson, 1994). So in order to control for differences in innovativeness, I created a variable R&D intensity, which is measured by R&D expenditures over sales.
Domestic ownership. Domestic firms will be better able to protect their patents than foreign firms
variable. Domestic owned firms where assigned a 1 and foreign owned firms a 0. Increase in this variable will imply a higher risk for domestic owned firms in litigation risk.
Empirical Approach
The dependent variable, number of defendant cases, indicates a high number of firms that are present in a litigation lawsuit only a limited number of times (Appendix 7 - Table 6). The suitable analysis method for this type of data is a negative binomial regression analysis. Using SPSS for the analysis of the data several steps have led to reliable results in order to test the hypotheses:
First, I developed a model to test hypothesis 1 and 3, regarding the number of industries (diversification) and related diversification (share of related diversification). This model shows the relationship between diversification and the relatedness of the industries in which the firm is diversified as the independent and number of defendant cases as dependent variable.
Second, another model similar to the first model but with the inclusion of the interaction effect of firm size as measured by revenue and will answer hypothesis 2.
Results Descriptive analysis
First of all, the descriptive analysis of the data gives insight into the main industry in which the firms in the sample focus their activities. 92 out of the 126 firms (approximately 72%) of the firms are active in the Computer and Electronic Product Manufacturing industry (NAICS code 334; Appendix 1 - Graph 1). Two firms are active in the related technologically related Computer Systems Design and Related Services industry.
The average number of defendant cases in which the firm is involved in is 19.37 with a standard deviation of 33.99 (Appendix 7 - Table 6). Almost 50% of the firms are involved in six cases as a defendant and 75% only in 16 cases. So the high mean is a result of a few outliers with the number of cases as high as 179 cases as indicated by a relatively high standard deviation.
The sample consists out of 71 (56%) diversified firms and 55 (44%) single industry firms (Appendix 5 - Table 4). However, 20 firms have a diversification level of 1, indicating diversification within a single industry. This tells us that 75 firms (60%) are operating in a single industry firms, which results in a base of similar knowledge within the boundaries of the firm. Another finding of this model is that out of the 71 firms that are diversified, 53 firms where either completely unrelated diversified (33 firms) or completely related diversified (20).
On average, a firm in the sample is active in 1.39 industries with a standard deviation of 1.968 (Table 1). A diversified firm is, on average, active in 2.46 industries with a standard deviation of 2.055 (Table 2).
The number of industries in which the firm is diversified are small. Results show that 83% of the firms are active in 2 industries or less. This might be an indication that firms have difficulties in managing activities in numerous industries, lowering coordination costs and creating synergies between the industries (Miller, 2006).
Independent variables N Minimum Maximum Mean Standard
Deviation
Diversification 126 0 10 1.39 1.968
Share of related diversification
126 0,00 1,00 .2197 .3766
Table 1 - Descriptive statistics of total sample
Independent variables N Minimum Maximum Mean Standard
Deviation
Diversification 71 1 10 2.46 2.055
Share of related diversification
71 0 1 .3899 .4312
Table 2 - Descriptive statistics of diversified firms (single industry firms excluded)
Hypothesis testing
The results of the negative binomial regression analysis are depicted in table 3. For all models I checked for multicollinearity and does not seem to be a problem for either of the models.
Model 1 Model 2 Revenue (ln) .339** (.0209) (.0245).387** Plaintiff cases (ln) .551** (.0275) (.0276).541** R&D intensity 2.024** (.3386) 2.066** (.3410) Patent portfolio size (ln) .043**
(.0102) .044** (.0475) Domestic ownership .309** (0595) .308** (.0601) Diversification -.114 ** (.0173) (.1416).448**
Share of related diversification -.078 (.0666)
-.142*
(.0683)
Diversification * Revenue -.033**
(.0083)
Model fit AIC = 1609,890 AIC = 1597,275
Hypothesis 1. This hypothesis is based on the number of industries in which a diversified firm is
active, where 0 is a single industry firm and going up to 10, which was to most industries in which a firm was active in this sample. It was expected that diversification would result in a higher risk of being litigated. The results show of the first model show a different direction of the relation: being active in more industries lowers the risk of being litigated. When the interaction effect of firm size is included into the equation (model 2), the relation is changed in the direction as it has been described in the hypothesis. The AIC value indicates a better fit for the model so therefore this model will be guideline for the conclusion for this hypothesis.
So on the basis of model 2 it can be said that the number of industries is positively related to the amount of defendant cases, supporting hypothesis 1. This indicates that as the number of industries grows, the risk of being a defendant in a litigation suit also rises (p < .01). Based on this finding, hypothesis 1 is supported. This finding can be interpreted accordingly. First, as the number of industries grows, the knowledge breadth within the firm gets wider3. Probably this result leads to different knowledge, as the NAICS codification was based on the intellectual property owned by the organization (Bureau van Dijk Electronic Publishing, 2014b). This can be interpreted as a growth in knowledge breadth within the firm and lays in the line of reasoning of the hypothesis.
Second, if a firm is active in more industries it will be close to more firms. Nearly 72% of the patent litigation lawsuits take place between firms that are technologically close (Bessen & Meurer, 2005a). Firms will be seen as threat for a firm’s position within the market, which results in a higher litigation risk when they are closer. Possible market entry, and the use of its patents are, among others, possible indications for a patent owner that their position within the market might not be sustainable over time and will be more likely to pursue infringement of their patents, increasing the risk for the diversified firm.
Hypothesis 2. Model 2 forms the basis for answering hypothesis 2. Looking at the interaction effect
of firm size on the relationship between diversification and the number of defendant cases shows a negative result (p < .01). This indicates that as diversification increases, smaller firms have an increased risk of litigation as opposed to larger firms, supporting hypothesis 2.
The underlying reasoning is as follows. First, increased coordination costs in order to manage the diversity of activities in the various industries make that small firms have a disadvantage as opposed to larger firms. Second, small firms interact with fewer firms. This put them in a disadvantaged position: more frequent interaction lowers the risk of litigation (Lanjouw & Schankerman, 2004). Third, also the inability to check for possible infringement due to their size make them less able to settle disputes and avoid litigation suits when diversification increases (Lanjouw & Schankerman, 2004). Larger firms also tend to be more diversified than smaller firms (Garcia-Vega, 2006).
Hypothesis 3. For hypothesis 3 I used the measure share of related diversification, which is
conducted by checking the secondary 4-digit NAICS codes for similarity in the 3-digit range, taking the primary NAICS code as reference point.
Results indicate a tendency of the share of related diversification to have a negative influence on the number of defendant cases a firm is involved in, which was not expected in the hypothesis (p <.05). This indicates that as the share of related diversification increases, the risk of being litigated drops. Based on this finding
hypothesis 3 is not supported. The conclusion that can be drawn from this finding is that it is not just the
number of industries that determines the number of defendant cases but also the origin of the industry in which the firm diversifies.
Several explanations may underlie this result. First, the assumption that related diversified firms are more involved in litigation lawsuits due to low interaction with other industries may be a false assumption. As the industries are related, it is possible that they interact with related industry firms as they still have useful patents that firms can use. Interaction could even be required if related industries have patents that are required for products to function, assuming that related industries are better searchable for patented technologies as opposed to unrelated industries.
Second, the reasoning behind the results is that the useful patents that are increasing the litigation risk might increase as opposed to decrease the ability to settle. It may be in fact this usefulness of the patents that make them better able to settle as it brings opportunities for firms to diversify into other industries. Settling a case with a firm in a related industry could give the firm access to knowledge that would otherwise be unattainable, opening options for diversification or recombining knowledge without the costs of R&D. Third, the higher innovative performance of related diversified firms leads to higher patenting activity (Palich et. al., 2000; Garcia-Vega, 2006). They build strong patent portfolios, which bring them in an advanced position in settlement negotiations. As a result of this strong patent portfolio they have a lower risk of litigation due to this advanced position.
Combining the findings of the three hypotheses leads to the conclusion that unrelated firms face the highest risk of being litigated. Hypothesis 1 confirmed that as the number of industries increases, the risk of litigation rises. Hypotheses 3 results in the conclusion that related diversification lowers the risk of litigation. As a result, unrelated diversified firms face the highest risk of litigation because of their inability to settle due to the lower amount of relevant patents combined with lower interaction with the plaintiff.
Control variables
Firm size. The result regarding firm revenue is, contrary of what was expected, possessively related
to risk of litigation. This means that as a firm grows, it faces higher risk of being litigated. A possible explanation is that the measure of firm size is not related to a firm’s revenue. This is highly unlikely as earlier research found it to be an indicator of firm size (Garcia-Vega, 2006). Another explanation is that the timespan of the measure. Revenue was only based on the year 2012 and thus captures only one year. This figure can be biased through the influence of onetime events like takeovers, sale of a business unit or government size. Longitudinal measurements of a firm’s revenue will lower this bias and generate a more reliable measure.
Plaintiff cases. The number of times a firm acted as a plaintiff affected the number of defendant
Patent portfolio size. Patent portfolio size is positively related to the risk of litigation, implying that
when a patent portfolio gets bigger, the risk of litigation also rises. Previous research showed a negative relationship between the number of patent held by the firm and the risk of litigation (Bessen & Meurer, 2005a, Lanjouw & Schankerman, 2004). The base year for data that measures patent portfolio size was 1976. It could be that the patents from the earlier years are not relevant anymore and do not bring an advantage to the negotiation position as they may have expired or are replaced by other patents.
R&D intensity. For both models the effect of R&D intensity on the risk of litigation is positive. This
indicates that when R&D intensity raises the risk of litigation is higher, as was expected.
Domestic ownership. In this sample, US firms have, contrary to what is expected based on the
literature a higher risk of litigation (Lanjouw & Schankerman, 1997/2004). One possible explanation is that foreign firms in the sample are large and have substantial activities in the US, which eliminates the disadvantages I expected foreign firms to encounter in protecting their intellectual property4. The disadvantages where found to have a larger effect on smaller firms, which supports this explanation (Lanjouw & Schankerman, 2004). If this is true, foreign firms in the sample have similar abilities to protect their intellectual property as they are bigger, are likely to have an internal legal department, specialized in the US market and have a equal likelihood of filing a suits.
Discussion
The protection of intellectual property through patents is not as straightforward as one might assume when one encounters the word property (Bessen & Meurer, 2005). The stronger patent protection policy introduced in the early 80’s has increased the validity of many patents and resulted in a rapid increase in patent application (Bessen & Meurer, 2005a). The enforcement of the patent rights has been rising since, causing holdup problems in complex industries (Sharipo, 2000). Still, only 45% of the patent infringement cases are won by the plaintiff (Sherry & Teece, 2003). This uncertainty affects almost every firm within a complex industry as fundamental technologies are protected by patents, making them indispensable for many products (Sharipo, 2000; Hall & Ziedonis, 2001; Hall & Ham, 1999; Grindley & Teece, 1997; Kash & Kingston, 2001). Also costs associated with patent litigation suits have risen dramatically during the 80’s and 90’s but do not form a barrier for organizations to apply for a patent (Cohen et. al., 2000; Bessen & Meurer, 2008). The acquisition of the rights to make use of the patents is an absolute necessity in order for products to function properly. The rise in patenting activity increases the difficulty in searching through the patent thicket, resulting in the fact that often firms unknowingly infringe another’s patents, creating a potential for infringement lawsuits (Sharipo, 2000). Thus avoiding patent litigation is not always possible, even if one strives not to infringe patent rights of others.
4
However, not every company faces the same rate of litigation and similarities between firms seem to uncover characteristics that increase the risk of litigation. Previous research has already identified several organizational characteristics such as firm size, portfolio size, and R&D intensity (Lanjouw & Schankerman, 2001; Lanjouw & Schankerman, 2004; Garcia-Vega, 2006).
The purpose of this research is to uncover if differences in knowledge base, more specifically the diversification of the knowledge base and its influence on the risk of being litigated.
Theoretical Implications
The results of this study have several implications regarding the risk of litigation. First, results indicate that the number of industries in which the firm is active increases the risk of being litigated. As the number of industries increases, the knowledge base will get wider and will result in more radical innovations that form the foundation for new technologies (Quintana-Garcia & Benavides-Velasco, 2008; Ettlie et. al., 1984). These types of innovations have a higher probability of being involved in a litigation lawsuit (Lanjouw & Schankerman, 2001).
Second, as the number of industries in which the firm is active grows, so does the number of close firms, which increases the risk of litigation. As the diversification into different industries rises, the knowledge base gets wider. By extension, a firm gets technologically closer to more firms. Previous research found that 72% of the patent litigations takes place between firms that are in the same or at least to related industries (Bessen & Meurer, 2005a).
Third, as the number of industries in which the firm is active grows, small firms will have a larger risk of being litigated than larger firms. Larger firms will be better able to (1) check for patent infringement (of their own patents and if they infringe patents of others), (2) interact with more firms, which result in a lower risk of litigation, and (3) have lower coordination costs. This will increase their ability to avert costly litigation lawsuits.
Fourth, diversification into unrelated industries has the largest effect on the risk of litigation. Diversification into related industries has the biggest effect on innovativeness, as opposed to unrelated diversification and single industry focus (Sampson, 2007). Unrelated diversified firms have lower amount of relevant patents to the plaintiff, which makes the likelihood of settlement lower, and based on the results of this study increases the risk of being litigated. Firms active in related industries give the firm an ability to combine in related industries as firms diversify in related industries most often (Breschi et. al., 2003). Single industry firms will be able to settle the dispute due to repeated interaction and reliance on future collaboration within the industry (Lanjouw & Schankerman, 2004; Bessen & Meurer, 2008).
Managerial Implications
industries only and 51 have at least some activity in unrelated industries. This indicates that firms in the sample either do not worry about the risk of litigation and the associated costs or are not aware of them. Now that they can better assess the risks, firms can take proper action accordingly. Changing their patenting activity in order to gain a sizable patent portfolio, increasing R&D activity are, among others, actions that will lower this risk.
Also the decision whether or not to diversify into (un)related industries can be made more thoroughly. Not only the effects on for example knowledge base, financial investment and R&D activities but also the effects on the patent portfolio and the litigation risk can be included into the equation.
Ideally, as a firm diversifies into more industries, it is, based on the results of this research, better to grow the firm completely as it lowers the risk of being litigated. Previous research already found a positive link between diversification and firm size (Garcia-Vega, 2006).
Limitations
This research is bound by several limitations. Among them are several limitations in the methodology of the research. First, within the variable of the industry codes, no value is assigned to the activities of the firm within a specific industry. The results indicated that 61% of the activities took place in technologically different industries and was not in line with the literature. By making no distinction between the activities performed in each industry, the results of this study are possibly affected. A possible consequence is that knowledge can be far more similar than the data would suggest which could have resulted in a higher or lower influence of the number of litigations.
Second, there was no qualitative assessment of patent portfolio included into the analysis, which results in the same limitation as the first one. Including a qualitative assessment would have gained insight into the type of innovations introduced by the firm as well (Ettlie & Rubenstein, 1987). Based on the literature, I assume that diversified firm introduce more radical innovation. However, including qualitative patent analysis would have increased the reliability. Another quality measure that will increase the reliability is the inclusion of backward citation of patents, which is found the decrease the risk of litigation due to maturity of the industry and results in less uncertainty about property rights (Lanjouw & Schankerman, 2004).
Third, inclusion of a variable measuring the litigation cost encountered by each firm will benefit the reliability as well. This variable controls for avoidance of litigation if a firm is confronted with high litigation costs (Lerner, 1995). This would explain the differences in litigation risk even better.
The revenue variable was only measured at a single point in time, making it more vulnerable to onetime events like takeovers, sale of a business unit or government decisions. Longitudinal measurements of a firm’s revenue will lower this bias and generate a more reliable measure.
eight to ten years, which may be more useful for the purpose of this research (Bessen & Meurer, 2005a; Lanjouw & Schankerman, 2004).
The result of this research is generalizable to other complex industries as they rely on patents in similar fashion as the Semiconductors, Mobile Communications and Devices, and Consumer Electronics and PC’s industries. Discrete industries use patents in very distinct ways and will not face the same litigation risk as the industries investigated in this research.
Also the limited sample size of 126 firms limits the results of this research. For the measure of diversification only 71 firms remained and results are likely to be influences by some outlier firms.
Future research
The revenue variable resulted in unexpected behavior and inclusion of different measures for firm size might lead in more reliable results. A longitudinal measurement of this variable will increase its reliability as it was only recorded for one single year in this study. Also including variables that measure the quality of the patent portfolio for each firm will give insight into the patent portfolio breadth of the firm, which this study does not account for. This will gain more insight into the value of activities performed within each industry. Lastly, including the types of innovations introduced by the firm will gain more insight into the characteristics of the patent portfolio. Radical innovations have a higher risk of being litigated than incremental innovations (Lanjouw & Schankerman, 2001).
Conclusion
The purpose of this study is creating more insight in the organizational characteristics involved in the litigation risk firm encounter when applying for patents. Managers can, based on the results of this study, better assess the risk of being involved in a litigation lawsuit. Being specialized and focus their R&D activities and subsequently their patenting activities in a single industry will lower this risk. Increase in diversification puts smaller firms at a disadvantage in comparison to larger firms. Being active in unrelated industries will incline the firms with the highest risk of being litigated.
References
Abernathy, W. J., Clark, K. B. (1985). Innovation: Mapping the Winds of Creative Destruction. Research Policy Research Policy, 14(1), 3-22.
Arundel, A., & Kabla, I. (1998). What Percentage of Innovations are Patented? Empirical Estimates for European Firms. Research Policy, 27(2), 127-141.
Bessen, J. E., & Meurer, M. J. (2005a). The Patent Litigation Explosion. Boston University School of Law Working Paper, (05-18).
Bessen, J. E., & Meurer, M. J. (2005b). Lessons for Patent Policy From Empirical Research on Patent Litigation. Lewis & Clark Law Review, 9, 1-39.
Bessen, J. E., & Meurer, M. J. (2008). The Private Costs of Patent Litigation. Boston University School of Law Working Paper, (07-08).
Breschi, S., Lissoni, F., Malerba, F. (2003). Knowledge-Relatedness in Firm Technological Diversification. Research Policy, 32(1), 69-87.
Brouwer, E., & Kleinknecht, A. (1999). Innovative Output, and a Firm's Propensity to Patent.: An Exploration of CIS Micro Data. Research Policy, 28(6), 615-624.
Bureau van Dijk Electronic Publishing. (2014a). Orbis Company Info. Retrieved from: orbis.bvd.com Bureau van Dijk Electronic Publishing. (2014b). Orbis User Guide (KtMine data FAQ). Retrieved from:
webhelp.bvdep.com
Burk, D. L., & Lemley, M. A. (2003). Policy Levers in Patent Law. Virginia Law Review, 89(7), 1575-1696.
Census Bureau. (2013). Business Research and Development and Innovation: 2008–10 (Report no.: NSF13332). Retrieved from: http://www.nsf.gov/statistics/nsf13332/pdf/nsf13332.pdf
Cohen, W. M., Nelson, R. R., & Walsh, J. P. (2000). Protecting Their Intellectual Assets: Appropriability Conditions and Why US Manufacturing Firms Patent (Or Not). NBER working paper series (No. 7552).
Dolfsma, W. (2011). Patent Strategizing. Journal of Intellectual Capital, 12(2), 168-178.
Ettlie, J. E., Bridges, W. P., & O'Keefe, R. D. (1984). Organization Strategy and Structural Differences for Radical Versus Incremental Innovation. Management Science, 30(6), 682-695.
Garcia-Vega, M. (2006). Does Technological Diversification Promote Innovation? Research Policy, 35(2), 230-246.
Gollop, F. M., Monahan, J. L. (1991). A Generalized Index of Diversification: Trends in U.S. Manufacturing. The Review of Economics and Statistics, 73(2), 318-330.
Granstrand, O., Oskarsson, C. (1994). Technology Diversification in "MUL-TECH" Corporations. IEEE Transactions on Engineering Management, 41(4), 355-364.
Grindley, P. C., Teece,D. J. (1997). Managing Intellectual Capital: Licensing and Cross-Licensing in Semiconductors and Electronics. California Management Review California Management Review, 39(2), 8-41.
Gulati, R., Nohria, N., Zaheer, A. (2000). Strategic Networks. Strategic Management Journal, 21(3), 203-215.
Hall, B. H. (2007). Patents and Patent Policy. Oxford Review of Economic Policy, 23(4), 568-587.
Hall, B. H., & Ham, R. M. (1999). The Patent Paradox Revisited: Determinants of Patenting in the US Semiconductor Industry, 1980-94. NBER working paper series (No. 7062).
Hall, B. H., & Ziedonis, R. H. (2001). The Patent Paradox Revisited: an Empirical Study of Patenting in the US Semiconductor Industry, 1979-1995. RAND Journal of Economics, 32(1), 101-128.
Haupt, R., Kloyer, M., & Lange, M. (2007). Patent Indicators for the Technology Life Cycle Development. Research Policy, 36(3), 387-398. doi:http://dx.doi.org.proxy-ub.rug.nl/10.1016/j.respol.2006.12.004 Kash, D. E., & Kingston, W. (2001). Patents in a World of Complex Technologies. Science and Public
Policy, 28(1), 11-22.
Kortum, S., & Lerner, J. (1998). Stronger Protection or Technological Revolution: What is Behind the Recent Surge in Patenting? NBER working paper series (No.6204).
Lanjouw, J. O., & Schankerman, M. (1997). Stylized Facts of Patent Litigation: Value, Scope and Ownership. NBER working paper series (No. 6297).
Lanjouw, J. O., & Schankerman, M. (2001). Characteristics of Patent Litigation: a Window on Competition. RAND Journal of Economics, 32(1), 129-151.
Lanjouw, J. O., & Schankerman, M. (2004). Protecting Intellectual Property Rights: Are Small Firms Handicapped?. Journal of Law and Economics, 47(1), 45-74.
Lerner, J. (1994). The Importance of Patent Scope: an Empirical Analysis. The RAND Journal of Economics, 25(2), 319-333.
Luca, L. M. De,Atuahene-Gima, K. (2007). Market Knowledge Dimensions and Cross-Functional Collaboration: Examining the Different Routes to Product Innovation Performance. Journal of Marketing Journal of Marketing, 71(1), 95-112.
Lunney Jr, G. S. (2003). Patent Law, the Federal Circuit, and the Supreme Court, A Quiet Revolution. Supreme Court Economic Review, 11, 1-80.
Miller, D. J., Fern, M. J., Cardinal, L. B. (2007). The Use of Knowledge for Technological Innovation Within Diversified Firms. Academy of Management Journal, 50(2), 307-325.
Miller, D. J. (2006). Technological Diversity, Related Diversification, and Firm Performance. Strategic Management Journal, 27(7), 601-619.
Palich, L. E., Cardinal, L. B., Miller, C. C. (2000). Curvilinearity in the Diversification-Performance Linkage: an Examination of Over Three Decades of Research. Strategic Management Journal, 21(2), 155-174.
Park, Y., Yoon, B., & Lee, S. (2005). The Idiosyncrasy and Dynamism of Technological Innovation Across Industries: Patent Citation Analysis. Technology in Society. 27(4), 471-485.
RPX Corporation. (2014). RPX Client Portal. Retrieved from. search.rpxcorp.com
Quintana-García, C., Benavides-Velasco,C. A. (2008). Innovative Competence, Exploration and
Exploitation: The Influence of Technological Diversification. Research Policy Research Policy, 37(3), 492-507.
Raghu, T., Woo, W., Mohan, S., & Rao, H. R. (2008). Market Reaction to Patent Infringement Litigations in the Information Technology Industry. Information Systems Frontiers, 10(1), 61-75.
Sampson, R. C. (2007). R&D Alliance and Firm Performance: the Impact of Technological Diversity and Alliance Organization on Innovation. Academy of Management Journal, 50(2), 364-386.
Schankerman, M. (1998). How Valuable is Patent Protection? Estimates by Technology Field. The Rand Journal of Economics, 29(1), 77-107.
Scherer, F. M. (1984). Innovation and growth: Schumpeterian perspectives. Cambridge, Mass.: MIT Press.
Shapiro, C. (2001). Navigating the Patent Thicket: Cross Licenses, Patent Pools, and Standard Setting. Innovation Policy and the Economy, 1 ,119-150.
Sherry, E. F., & Teece, D. J. (2004). Royalties, Evolving Patent Rights, and the Value of Innovation. Research Policy, 33(2), 179-191.
Somaya, D. (2003). Strategic Determinants of Decisions Not to Settle Patent Litigation. Strategic Management Journal, 24(1), 17-38.
Streb, J. (2003). Shaping the National System of Inter-Industry Knowledge Exchange:Vertical Integration, Licensing and Repeated Knowledge Transfer in the German Plastics Industry. RESPOL Research Policy, 32(6), 1125-1140.
Teece, D. J. (1982). Towards an Economic Theory of the Multiproduct Firm. Journal of Economic Behavior & Organization, 3(1), 39-63.
Appendices Appendix 1
Graph 1 - Primary and diversification industry codes of the sample
1 1 1 1 4 2 20 12 6 46 5 3 2 2 2 1 1 1 2 3 1 1 1 1 1 1 2 1 1 1 1 2 2 3 1 1 1 5 1 1 15 16 2 10 5 6 3 1 1 1 1 2 3 1 1 1 2 5 5 1 2 1 1 11 2 2 2 1 4 6 2 1 1 2 1 1 1 2 18 1 2 5 2 1 1 2 0 15 30 45 60 2382 3222 3231 3252 3259 3279 3262 3314 3329 3331 3332 3333 3334 3336 3339 3341 3342 3343 3344 3345 3346 3352 3353 3359 3361 3362 3363 3364 3365 3366 3391 3399 4234 4236 4238 4431 4539 4881 4889 5112 5152 5171 5172 5179 5182 5191 5222 5223 5239 5241 5259 5311 5321 5322 5324 5415 5416 5417 5511 5614 5615 5619 6114 8112
Appendix 2
Table 3 - Industry sector names with codes and number of firms active
Main sector Secondary sector Tertiary sector Code
(Descent)
Number of firms
Construction Specialty trade contractors Building Equipment contractors 2382 1 Manufacturing Paper Manufacturing Converted Paper Product
Manufacturing
3222 1
Printing and Related Support Activities
Printing and Related Support Activities
3231 2
Chemical Manufacturing Resin, Synthetic Rubber, and Artificial Synthetic Fibers and Filaments Manufacturing
3252 2
Other Chemical Product and Preparation Manufacturing
3259 4
Plastics and Rubber Products Manufacturing
Rubber Product Manufacturing 3262 1 Nonmetallic Mineral Product
Manufacturing
Other Nonmetallic Mineral Product Manufacturing
3279 1
Primary Metal Manufacturing Nonferrous Metal (except Aluminum) Production and Processing
3314 1
Fabricated Metal Product
Manufacturing Other Fabricated Metal Product Manufacturing 3329 1 Machinery Manufacturing Agriculture, Construction, and
Mining Machinery Manufacturing 3331 1 Industrial Machinery Manufacturing 3332 1
Commercial and Service Industry
Machinery Manufacturing 3333 9
Ventilation, Heating, Air-Conditioning, and Commercial Refrigeration Equipment Manufacturing
3334 1
Engine, Turbine, and Power Transmission Equipment Manufacturing
3336 2
Other General Purpose Machinery Manufacturing
3339 1
Computer and Electronic Product Manufacturing
Computer and Peripheral Equipment Manufacturing
3341 35
Communications Equipment Manufacturing
3342 28
Audio and Video Equipment Manufacturing
3343 8
Semiconductor and Other Electronic Component Manufacturing
3344 56
Navigational, Measuring, Electromedical, and Control Instruments Manufacturing
3345 10
Manufacturing and Reproducing Magnetic and Optical Media
3346 9
Electrical Equipment, Appliance, and Component Manufacturing
Household Appliance Manufacturing
3352 5
Manufacturing
Other Electrical Equipment and Component Manufacturing
3359 3
Transportation Equipment Manufacturing
Motor Vehicle Manufacturing 3361 3 Motor Vehicle Body and Trailer
Manufacturing
3362 1
Motor Vehicle Parts Manufacturing
3363 2
Aerospace Product and Parts Manufacturing
3364 4
Railroad Rolling Stock Manufacturing
3365 1
Ship and Boat Building 3366 1
Miscellaneous Manufacturing Medical Equipment and Supplies Manufacturing 3391 2 Other Miscellaneous Manufacturing 3399 3 Wholesale Trade Merchant Wholesalers, Durable Goods
Professional and Commercial Equipment and Supplies Merchant Wholesalers
4234 7
Electrical and Electronic Goods Merchant Wholesalers
4236 8
Machinery, Equipment, and
Supplies Merchant Wholesalers 4238 1 Retail Trade Electronics and Appliance
Stores
Electronics and Appliance Stores 4431 2 Miscellaneous Store Retailers Other Miscellaneous Store
Retailers
4539 1
Transportation and
Warehousing
Support Activities for Transportation
Support Activities for Air Transportation
4881 1
Other Support Activities for Transportation
4889 1
Information Publishing Industries (except Internet)
Software Publishers 5112 11
Broadcasting (except Internet) Cable and Other Subscription Programming
5152 1
Telecommunications Wired Telecommunications Carriers
5171 2
Wireless Telecommunications Carriers (except Satellite)
5172 3
Other Telecommunications 5179 3
Other Telecommunications Data Processing, Hosting, and Related Services
5182 2
Other Information Services Other Information Services 5191 5 Finance and
Insurance
Credit Intermediation and Related Activities
Nondepository Credit Intermediation
5222 6
Activities Related to Credit Intermediation
5223 2
Securities, Commodity
Contracts, and Other Financial Investments and Related Activities
Other Financial Investment Activities
5239 1
Insurance Carriers and Related Activities
Insurance Carriers 5241 1
