ResearchPDF Available

Alternative responses to anti-commons: state financing of latecomer patenting

Research

Alternative responses to anti-commons: state financing of latecomer patenting

Abstract

Governments in developing countries, especially large middle-income ones such as China, often face a dilemma: How can they govern intellectual property rights (IPR) to alleviate “anti-commons” problems, such as patent thickets, without also restraining technological development? Existing scholarship has not yet explored the role of extra-legal institutions such as state financing for patenting (SFP) (e.g., subsidies and financial awards for patenting) as a potential response to this challenge. Drawing on hierarchical regression analyses of original survey data from Chinese firms, I find that SFP in China appears to have temporarily alleviated some anti-commons problems. It has accomplished this by providing additional resources to recipient firms in industries where technology is highly cumulative to accumulate patents for cross-licensing, which enables them to bargain to reduce patent litigation risks and, in turn, reduce royalties owed to others. However, China has simultaneously experienced negative policy externalities from sub-optimally designed SFP. I conclude with recommendations about how governments can learn from China’s experiences.
1
Alternative responses to anti-commons:
state financing of latecomer patenting
Dan Prud’homme
EMLV Business School
Paris, France
Duke Kunshan University
Kunshan, China
GLORAD, Tongji University
Shanghai, China
GLORAD Center for Global R&D and Innovation
Working Paper
August 2020
2
Alternative responses to anti-commons: state financing of latecomer patenting
Dan Prud’homme
Abstract: Governments in developing countries, especially large middle-income ones such as China,
often face a dilemma: How can they govern intellectual property rights (IPR) to alleviate “anti-
commons” problems, such as patent thickets, without also restraining technological development?
Existing scholarship has not yet explored the role of extra-legal institutions such as state financing for
patenting (SFP) (e.g., subsidies and financial awards for patenting) as a potential response to this
challenge. Drawing on hierarchical regression analyses of original survey data from Chinese firms, I
find that SFP in China appears to have temporarily alleviated some anti-commons problems. It has
accomplished this by providing additional resources to recipient firms in industries where technology
is highly cumulative to accumulate patents for cross-licensing, which enables them to bargain to reduce
patent litigation risks and, in turn, reduce royalties owed to others. However, China has simultaneously
experienced negative policy externalities from sub-optimally designed SFP. I conclude with
recommendations about how governments can learn from China’s experiences.
Keywords: patent anti-commons, intellectual property rights, IPR policy, technological development,
state financing for patenting (SFP)
3
1. Introduction
Many latecomer firms in developing countries face unprecedented intellectual property rights (IPR)
resource disadvantages vis-à-vis foreign incumbents, the most significant type of which results from
patent anti-commons (e.g., Song, 2013; Xiao, 2013). “IPR resource disadvantages” here mean
comparatively limited control of knowledge protected by IPRs and ownership of few or no intellectual
property legal rights (Hall, 1992). “Anti-commons”, a situation where rational individuals acting
separately consume less than the social optimum of a resource because too many individuals have the
property rights to exclude others from exploiting technological knowledge (Heller, 1998), occurring
in the context of patent portfolio (PP) ownership are called “patent anti-commons” (Heller and
Eisenberg, 1998). Given the increase in patent anti-commons in recent years, governments in
developing countries face a pressing question: How can they most strategically govern IPR to alleviate anti-
commons problems? Available scholarship on IPR policy and technological development strategy does
not yet appear to offer a clear answer to this question.
Conventional wisdom suggests that there are two main ways that governments can try to reduce
latecomers’ IPR resource disadvantages vis-à-vis incumbents, including those resulting from patent
anti-commons. First, development economists have argued that latecomer economies can maintain
relatively “weak” IPR institutions, meaning those offering limited statutory legal protections related
to appropriability for IPRs and/or limited enforcement of IPRs, in the early stage of their catch-up
process (e.g., Ginarte and Park, 1997; Park, 2008). This policy approach allows the latecomers to avoid
being constrained by incumbents’ IPR power that would otherwise prevent imitation and associated
technological learning (e.g., Kim, 1997). Without such an approach to IPR institutions, latecomer
firms cannot efficiently engage in the learning needed to accumulate fundamental knowledge
capabilities and eventually indigenously innovate, which is essential to sustain economic growth in the
long-run (Odagiri, 2010a; Kim et al., 2012). This policy approach to catching up is well documented
in the context of both East Asian countries (e.g., Kumar, 2003; Odagiri, 2010b) and Western ones
(e.g., Dutfield and Suthersanen, 2005). The second somewhat overlapping policy prescription, derived
from the law & economics literature in response to patent anti-commons in particular, is that the state
should consolidate patent ownership in a single or limited number of entities, or preclude patent
protection altogether for certain upstream research inputs (Burk and Lemley, 2003). Patent pools,
4
restrictions on patentable subject matter, and compulsory licensing are the most often recommended
instruments for operationalizing this advice (e.g., Merges, 1996; OECD, 1989; Geiger and Gross,
1997). I call this second group of policy tools “other restrictions on IPRs” for simplicity.
Meanwhile, both of the aforementioned approaches to IPR governance can restrain technological
development. Inadequate IPR institutions, in both developing and developed economies, have been
shown to sometimes discourage technology transfer, R&D spending, and other types of investment
necessary for technological development (e.g., Groizard, 2009; Hall, 2014). Other restrictions on IPRs,
in both developed and developing economies, can sometimes also discourage critical investments
needed for innovation (e.g., Choi and Gerlach, 2015; Lampe and Moser, 2016). Policymakers in
developing countries therefore face a fundamental dilemma when considering currently available policy
advice regarding how to respond to patent anti-commons: commonly employed approaches may
indeed reduce latecomers’ IPR resource disadvantages but at the same time restrain technological
development. Among developing countries, this dilemma may be especially acute for policymakers in
large (Grossman and Lai, 2004, p. 1652) upper-middle-income economies because IPR institutions
therein appear to matter more to technological development compared with smaller, lower-income
economies (e.g., World Bank, 2001; Gentile, 2020). How can these policymakers govern IPR to alleviate patent
anti-commons without also restraining technological development?
There appears to be an underexplored policy tool available that may offer a way around this dilemma:
state financing for patenting (SFP), meaning subsidies and financial awards for patenting. Because SFP, a
unique extra-legal policy instrument existing in parallel to formal legal institutions, is distinct from the
aforementioned two policy instruments, policymakers deploying SFP as a response to patent anti-
commons do not face the dilemma associated with choosing from those instruments. Further, as
discussed in this paper, a government may be able to alleviate patent anti-commons problems by
relying on SFP instead of those two policy instruments.
China is an especially relevant empirical context in which to examine the ability of SFP to reduce
patent anti-commons problems. China is a large, upper middle-income country at present. In addition,
the Chinese government, although not always successful, has concertedly attempted over the last
decade or so to institute relatively robust legal IPR protections to facilitate the domestic and foreign
5
investments necessary to advance the country technologically (e.g., Lee, 2015; Huang, 2017;
Prud’homme and Zhang, 2019). Meanwhile, provincial governments throughout China have devoted
a significant amount of their resources to fund SFP in recent years (e.g., Song et al., 2016). Further,
although not yet explored as a potential response to patent anti-commons problems, valuable recent
research does suggest that SFP in China has had a major impact on Chinese firms’ patenting, enabling
China to become the world leader in domestically-filed patents as of 2011 (e.g., Liefner et al., 2016;
Cheng and Drahos, 2017). Considering all this, further research seems warranted about how SFP in
China may or may not have helped to reduce problems associated with patent anti-commons in the
country. Research on this topic is useful not only to scholars and policymakers studying the successes
or failures of China’s IPR-related policies, but also for those looking to potentially apply those
experiences in other developing nations facing similar development challenges.
This paper aims to bridge the aforementioned gap in the literature. I start by discussing how additional
resources provided by SFP can help latecomer firms facing the most significant patent anti-commons
patent thickets, meaning dense webs of overlapping patents that firms must navigate to actually
commercialize new technology (e.g., Shapiro, 2000, 2002; Hegde et al., 2009) expand their patent
portfolios for cross-licensing. I then look at how this bargaining mechanism can help recipient firms
reduce patent resource dependencies on external organizations by deterring patent litigation and
reducing royalties owed. I collect original survey data about the extent to which receipt of financial
support from SFP in China is reported by managers at Chinese firms in different industries to help
them accumulate patents for various purposes. I analyze this data with mediated hierarchical
regressions and other multivariate regressions and discuss the theoretical and policy implications of
the analysis.
This paper makes several contributions to the study of China’s technological development, IPR policy,
and the strategic role of the state therein. In terms of theory, my main contribution is that the Chinese
government appears to have temporarily alleviated some anti-commons problems associated with
patent thickets without also substantially weakening the appropriability offered by the country’s IPR
legal institutions. SFP has accomplished this by providing additional resources to recipient firms in
industries where technology is highly cumulative to accumulate patents for cross-licensing, which
6
enables them to bargain to reduce patent litigation risks and, in turn, reduce royalties owed to others.
At the same time, however, and very importantly, I acknowledge the negative policy externalities
simultaneously experienced from sub-optimally designed SFP in China. In terms of policy implications,
I discuss how developing countries looking to China for technological development policy leadership
must not only consider China’s successes with SFP but also better manage the negative policy
externalities it experienced from sub-optimally designed SFP.
2. Conceptual framework and hypotheses
In this section, I discuss how the additionality of state financial support can empower firms to reduce
patent anti-commons. I start by briefly discussing how SFP can help recipient firms expand their
patent portfolios/PPs (Section 2.1). I then discuss why these additionality effects will help recipient
firms facing the most significant patent anti-commons patent thickets expand their PPs for cross-
licensing in particular (Section 2.2). I then look at how this bargaining mechanism, in turn, can help
recipient firms reduce patent resource dependencies on external organizations by deterring patent
litigation and reducing royalties owed to others (Section 2.3).
2.1 How SFP can help firms expand patent portfolios: the role of additionality effects
SFP can take a variety of forms, including subsidies provided by government departments to cover
the official costs of filing and examining patents, and sometimes their renewal fees, as well as financial
awards provided to firms when their patents are granted (e.g., Prud’homme, 2016; Xu and Munari,
2016). State subsidies and grants can positively affect firm-level production of patents via additionality
meaning the additional resources, relative to a baseline without such support, that government
financial support can offer to recipient firms (e.g., Hsu et al., 2009; Czarnitzki and Hussinger, 2017).
In China in particular, SFP is unlikely to simultaneously crowd-out private investment because many
Chinese firms have limited resources and do not prioritize spending the non-state-provided resources
that they do have on PPs (e.g., Zhu, 2013; Song, 2013; Song et al., 2016). In other words, SFP in China
will likely serve as a source of fresh IPR resources for recipient latecomer firms that they otherwise
likely would not have timely accumulated.
7
Although valuable recent research has already shown that SFP has been integral to a prolonged
patenting surge in China (e.g., Song et al., 2016; Long and Wang, 2016, 2019; Liefner et al., 2016;
Cheng and Drahos, 2017),
1
there is sparse if any literature describing the exact mechanisms by which
SFP’s additionality effects can help firms accumulate patents. Interviews were therefore conducted
with several Chinese firms to better understand these mechanisms. By “accumulate” patents I mean
to self-create (rather than in-license or purchase assignment rights for) and obtain more patents from
the patent office: in other words, to independently expand one’s PP.
2
Based on the interviews, SFP in
the form of state subsidies for the costs of invention patenting and financial rewards for granted
invention patents appears to contribute to recipient firms’ patent accumulation through three main
mechanisms: (1) by making it more likely that firms, if they were previously reluctant to file patents
primarily because of the official costs of doing so, now seek patents; (2) by allowing firms to spend
the money they saved on official patenting costs for prosecuting (i.e., writing and filing) other patent
applications and/or conducting R&D needed for other patents; and (3) by allowing firms to spend
the award money they receive from granted patents on prosecuting other patent applications and/or
conducting R&D needed for other patents (Prud’homme and Long, 2019).
2.2 How SFP can help firms bargain in patent anti-commons: the role of cross-licensing
An important clarification is warranted before I move forward with the remainder of the conceptual
framework: hereafter I do not attempt to design a framework in which receipt of SFP per se
determines which patent strategies firms choose. Firms first and foremost choose patent portfolio/PP
strategies according to the competitive landscape, technological fields included, in which they operate
(e.g., Somaya, 2012; Hall and Ziedonis, 2001; Hall, 2005). Instead, in the remainder of this framework
I merely intend to look how useful SFP might be to offering additional resources to firms to accumulate
patents for the specific PP strategies that they have already selected.
1
At the same time, the domestic patent surge in China is also owed to a number of other factors not directly attributable to SFP, for
example rising R&D intensity by Chinese firms, competition and learning, foreign firms seeking to ensure freedom to operate, and
changes in Chinese patent law and the institutions governing its administration and enforcement (e.g., Hu et al., 2016; Keupp et al.,
2012).
2
I use the concept of “self-create” here because it is in line with the qualification criteria in SFP programs in China reviewed.
8
As mentioned in the introduction, patent anti-commons occur because too many individuals have the
property rights to exclude others from exploiting technological knowledge (Heller and Eisenberg,
1998). Anti-commons occur horizontally if patents cover different components that must be
integrated into a product, and vertically if patents cover different steps in a cumulative innovation
process (Burk and Lemley, 2003). Highly fragmented ownership of patents owned by upstream firms
effectively creates significant transaction costs in the form of bargaining required with potential
downstream users of patented technology (Burk and Lemley, 2003; Lemley and Shapiro, 2007). As
also mentioned in the introduction, anti-commons problems are particularly pronounced in
technologies prone to patent thickets (Shapiro, 2000, 2002; Ziedonis, 2004; Murray and Stern, 2007;
Huang and Murray, 2009).
3
I therefore focus the remainder of my patent anti-commons framework
on patent thickets.
Patent thickets are more often found in technology-based industries where inventions are highly
“cumulative” (i.e., closely technologically related) and “complex” (i.e., reliant on multiple inter-related
components), such as information & communication technologies (Hall and Ziedonis, 2001; Hall,
2005), other electronics, machinery, and related industries (von Graevenitz et al., 2013). In this paper,
I refer to these as industries where technology is highly cumulative.” Industries less prone to patent
thickets can be labelled as being “discrete” (i.e., comprised of fields in which inventions are less
cumulative and therefore more distinct in terms of underlying technology and/or its application) (e.g.,
Besson, 2003; Hall, 2005). Discrete technologies are concentrated in chemicals, materials, agriculture,
services, and related industries (von Graevenitz et al., 2013).
Firms generally pursue differing strategies to overcome PP resource constraints based upon the
dispersion of patents in the industry in which they operate (Mansfield, 1984; Levin et al., 1987). Taking
this conceptualization a step further, von Graevenitz et al. (2007, 2013) divides firms’ patenting
3
This being said, some law scholarship distinguishes anti-commons from patent thickets. For example, Burk and Lemley (2003, p. 1613-
1614) find that although both anti-commons and patent thickets potentially prevent parties from making a final product incorporating
multiple inventions, anti-commons do this by creating multiple rights owned by different owners whereas patent thickets do this via
patents whose scope is too wide. Further, one could suggest that there are at least three major types of sub-optimalities related to this
discussion: when there are too many owners of IP rights, when the concentration of ownership of rights is in too few hands, and when
the breadth and/or duration of granted IP rights is too wide/long. While appreciating these dynamics, patent thickets have been treated
in the economics and management literature as essentially as a type of anti-commons problem (e.g., Ziedonis, 2004; Murray and Stern,
2007; Huang and Murray, 2009), and therefore, to be consistent with that literature, I follow such an approach.
9
strategies into two broad groups: “patent portfolio maximization” and “patent portfolio optimization”.
I adopt these two conceptualizations here but rename them the (1) “bargaining” patent portfolio (BPP)
and (2) “traditional” patent portfolio (TPP) strategies, respectively, because these labels are arguably
easier to understand.
4
The BPP strategy, which spans both offensive and defensive objectives (Arundel,
2001), aims to accumulate patents as bargaining tools to cut through patent thickets; in contrast, the TPP
strategy aims to accumulate fewer numbers of patents primarily to protect products and processes
against free-riding. Of course, firms may use a combination of these two strategies and the dichotomy
presented is not the only way to think about patenting strategies, rather is just one useful approach.
At the same time, firms that face significant patent thickets will, typically, heavily employ BPPs
whereas firms not facing significant patent thickets are much less likely to use the BPP strategy as
often (e.g., Shapiro, 2000, 2002; Blind et al., 2006, 2009; Galasso and Schankerman, 2010; von
Graevenitz et al., 2007, 2013; Cockburn et al., 2010; Torrisi et al., 2016).
The most frequent tactic employed as part of the BPP strategy is to use patents to negotiate with
external organizations via cross-licensing meaning an agreement where a party agrees to license his/her
patents in exchange for licenses from another party, often without the payment of royalties. In this
sense, firms engaging in cross-licensing exchange legal rights via bargaining so that all parties are better
off. Although patents serving a “blocking” function, meaning those preventing other patents from
being exploited because the latter group relies on technology claimed by the former, may also help
their owners navigated patent thickets (e.g., Merges, 1994), cross-licensing is still the most
foundational component of the BPP strategy (e.g., Shapiro, 2000, 2002; Galasso, 2012).
In sum, we know that firms facing significant patent thickets will often use patents for cross-licensing
to bargain their way through anti-commons. And, to be sure, this tactic is more common than
employing blocking patents or patents primarily protecting against free-riding. When also considering
the concept of additionality effects discussed in the prior section, it seems intuitive that SFP could
offer latecomer firms seeking to bargain their way through patent thickets via cross-licensing useful
4
This is because the words “maximization” and “optimization” can be synonyms. In terms of related conceptualizations, some have
used the term “strategic patenting” (e.g., Blind et al., 2009; Torrisi et al., 2016) to refer to what von Graevenitz et al. (2007) calls the
“patent maximization” strategy. However, there is no universally agreed upon definition of “strategic patenting” to date (Blind et al.,
2009, p. 429). Moreover, the portfolio optimization approach, like the maximization approach, involves careful allocation of scarce
resources in an attempt to achieve competitive advantage, therefore also broadly fits the definition of a “strategy”. Additionally, I call
the second strategy the “traditional” portfolio strategy because it follows the fundamental motivation that early scholars argued should
drive patenting (e.g., Arrow, 1962).
10
additional resources (financing) to accumulate patents for this purpose. Considering these concepts, I
propose that:
H1: SFP is commonly used by latecomer firms facing higher incidence of anti-commons (proxied by patent
thickets) to accumulate patents for cross-licensing rather than for other purposes.
2.3 How SFP can help firms reduce problems associated with patent anti-commons: the role
of cross-licensing to reduce litigation risks and royalties
Patent litigation, or just the threat of litigation, is a significant resource dependency problem associated
with patent anti-commons and one especially pronounced amidst patent thickets (e.g., Shapiro, 2002).
Patent litigation can create significant holdup problems (i.e., when a patent owner sues an
implementer firm after it has used a focal technology and can claim a large settlement because it is too
late for the implementing firm to change course) and/or reluctance or delay in fully exploiting new
products due to the threat that they will inadvertently infringe on patents granted after the products
were designed and expensive litigation will then ensue (Shapiro, 2000; Chien, 2014). Patent litigation
risks can present themselves to latecomer firms in other ways as well.
5
Patents that are most useful for cross-licensing will likely also be most useful to reduce patent litigation
risks associated with patent thickets. The main strategy for avoiding patent litigation is to in-license
greater numbers of patents in order to maintain freedom-to-operate (FTO), i.e., the ability to utilize a
product, process, or service without infringing on patents owned by other entities (Cohen et al., 2002;
Galasso and Schankerman, 2010). Another method is for a focal firm to leverage its own PP to sue or
threaten to countersue firms that otherwise might sue it for alleged patent infringement (e.g., Arundel,
2001; von Graevenitz et al., 2007, 2013). Patents for cross-licensing are likely to be an effective tool
for latecomers to pursue either of these BPP-based strategies because, more so than patents for other
purposes, they are explicitly meant to be used as bargaining chips in negotiations with other firms. It
5
Incumbents acting as plaintiffs and courts hearing patent disputes have the collective power to: (1) dictate the scope of the latecomer
defendant’s research and commercial application thereof, including but not limited to clearly requiring certain terms for patent licensing
agreements; (2) otherwise legally restricting the future direction of the defendant’s research by more clearly cordoning off applications
of technology with patent-related court judgments; and (3) requiring the defendant to devote scarce resources to the patent legal battle,
thereby distracting them from more productive tasks.
11
therefore follows, when also considering H1, that firms receiving SFP and facing anti-commons may
accumulate patents for cross-licensing with the follow-on objective of leveraging these patents to
reduce litigation risks. Considering these concepts, I hypothesize that:
H2: SFP’s usefulness to accumulating patents for cross-licensing serves as a positive mediator between higher
incidence of anti-commons (proxied by patent thickets) and the ability of SFP to help latecomer firms prevent
patent litigation against themselves.
Owing significant patent royalties and licensing fees to external organizations is yet another significant
IPR resource dependency problem associated with patent anti-commons, especially patent thickets,
that latecomers face (e.g., Blind et al., 2006, 2009; Torrisi et al., 2016). Yet, as firms’ PPs for cross-
licensing expand with the help of SFP, they may also help reduce royalties that they would otherwise
need to pay to external organizations. This seems intuitive when considering the literature showing
that cross-licensing patents, used for bargaining exchanges, offer firms a direct way to in-license
patents while simultaneously avoiding the royalty and other fees associated with typical in-licensing
(e.g., Gallini and Winter, 1984; Shapiro, 2000).
In addition, building on H2, the usefulness of SFP to cross-licensing PPs may determine how helpful
SFP is to preventing litigation and, in turn, this may shape the ability of SFP to ultimately reduce
royalties. As mentioned, cross-licensing offers latecomer firms relatively more resource-based power
in negotiations to maintain FTO. Yet this FTO bargaining power not only reduces the immediate risk
of patent litigation. When a latecomer expands its cross-licensing portfolio to reduce the threat of
litigation that external organizations can leverage over it, the latecomer simultaneously reduces the
leverage of those external organizations to use litigation threats to drive up patent royalty rates they
charge. Therefore, if latecomer firms draw on SFP to accumulate patents for cross-licensing and then
to reduce litigation threats, as H2 states, this behavior should, in turn indirectly also enable them to
reduce royalties otherwise owed to external organizations. Drawing on these concepts, I hypothesize:
H3: SFP’s usefulness to accumulate patents for cross-licensing is positively related to its contribution to reducing
royalties owed to external organizations, and this relationship is positively mediated by SFP’s contribution to
preventing litigation.
12
3. Method and materials
In this section, I discuss the methods materials used to test my hypotheses. I start out discussing the
empirical context for my research (Section 3.1). I then discuss my survey instrument, variable
construction, and econometric estimation methods (Section 3.2).
3.1 Overview of empirical context
In order to make the research for this paper manageable and focus it on a meaningful unit of analysis,
I chose Jiangsu province in China as my empirical context. Jiangsu, one of the richest provinces in
China, has a population of 74 million, which is larger than the majority of countries. Like other
provinces in China, Jiangsu has some flexibility from the central government to design its own IPR
policies, including SFP. Jiangsu has strong technological capabilities compared to the other 30
provinces in China: from 2006-2015, firms in Jiangsu owned the third highest number of invention
patents granted in China (after Guangdong and Beijing).
6
Also, the growth rate of invention patents
granted to firms in Jiangsu was above average in China during that time.
7
In this paper, I focus on two main types of SFP offered in Jiangsu: subsidies for costs of invention
patenting and financial awards for invention patenting.
8
At the time of my data collection, Jiangsu’s
provincial-level invention patent subsidies provided 1,000 RMB for the costs of filing and
substantively examining an invention patent. This covers a significant portion of the cost of an average
invention patent filing in China. Also, Jiangsu’s provincial government provided a 3,000 RMB award
per invention patent upon its granting. Both of these types of SFP are only intended for Chinese firms.
Taken together, this funding per patent (4,000 RMB) constituted around 5% of the average annual
per capita income in Jiangsu at the time of research (87,995 RMB). Moreover, at the time of research,
there were not stringent restrictions in practice on applicants for such programs, including on the
numbers of patents per firm that can receive these subsidies and awards. As a result, an individual
6
Calculated based on data from China’s National Intellectual Property Administration
7
Ibid
8
Policy details available from the Jiangsu Ministry of Finance and Jiangsu IP Office
13
Chinese firm can receive the equivalent of thousands of US dollars in SFP funds from the Jiangsu
government.
3.2 Survey and econometric analyses
Survey method
A survey was formulated to collect information from wholly-owned Chinese firms in Jiangsu province.
The survey first gathered background information on the respondent firms. It then collected data on
the perceived usefulness of the financial support from Jiangsu’s SFP to help respondent firms
accumulate patents for different purposes. All firms surveyed that provided complete responses
received both types of SFP considered (subsidies and awards) for each of their patents. I used a survey
as my research instrument due to insufficient public data in China about how exactly Chinese firms
use their patents, and because there is no comprehensive publicly available information about what
firms actually receive SFP in China. In this regard, my study appears to be the first to collect firm-
level data about actual receipt of SFP.
9
This being said, despite the novelty of my dataset, there are
limitations of it which are discussed further in Section 5.2.
The survey was written in Chinese and distributed in paper form. Based upon discussions with R&D
managers, IPR managers/counsels, and CEOs prior to distributing the survey, IPR
managers/counsels were identified as the most knowledgeable firm representatives to answer the
survey. The surveys were administered by officials running the Jiangsu National IPR Training Base,
which is an official designation provided by China’s National Intellectual Property Administration
(CNIPA) to institutions accredited to provide IPR management trainings to Chinese firms. The
Jiangsu National IPR Training Base invites firms to participate in their IPR management trainings by
issuing an electronic advertisement to the public and emailing the individuals in their database of local
firm representatives. Firms are invited to participate in the base’s trainings if they operate in any high-
9
The closest existing datasets gauge the existence of SFP at the provincial or city levels which are deductively assumed to apply to all
firms (e.g., Li, 2012; Lei et al., 2013; Dang and Motohashi, 2015). The next closest datasets are based on survey data but attempt to
gauge the impact of “government support”, broadly defined and not specific to patenting, on patenting and innovation in China (e.g.,
Shu et al., 2015, p. 305).
14
tech industry, are publicly listed or in the process of being listed, or are larger than a microenterprise
and operate in any traditional (non-high-tech) industry. The surveys used for my research were
distributed to the IPR managers/counsels of all the Chinese firms participating in one of the trainings
provided by the Jiangsu National IPR Training Base. The survey was optional to answer. 370 firms
received the survey and responses from 200 IPR managers/counsels, each representing a different
Chinese firm operating in Jiangsu (a response rate of 54%), were received. All responses were received
on the same day that the survey was distributed.
The survey responses were manually transferred from paper to an electronic spreadsheet. They were
then coded and constructed into variables. In line with standard statistical analysis practice, survey
responses lacking data on all dependent and independent variables were deleted in their entirety. There
were 47 “responses” that fell into this category (the only data provided was related to my control
variables) and therefore they were deleted, resulting in 153 usable responses. Then, the Multiple
Imputation by Chained Equation (MICE) method (see Su et al., 2011) was used to impute missing
data points for several partially incomplete responses.
Variables
Table 1 provides the full details of the variables I constructed from my survey questions. Table 2
provides the descriptive statistics for these variables.
15
Table 1: Construction method for variables
Concept/construct
Proxy/observed variable
(abbreviated name)
Reduction of resource
dependency
Patents to prevent litigation
(SFP to prev. lit.)
Patents to reduce royalties and
licensing fees
(SFP to red. roy.)
BPP strategy
Cross-licensing (SFP to cross-
lic.)
Other patenting strategies
Patents protecting
products/processes, which
drives the TPP strategy (SFP
to protect)
Blocking patents
(SFP to block)
Patent anti-commons
Patent thickets rating
(Patent thick. rating)
Industries where technology is
highly cumulative
16
(Patent thick. industry)
Technological capabilities
R&D intensity index
(R&D)
Other capabilities
Firm size index
(Firm size)
Industry dimensions
Industries where technology is
highly cumulative
(Patent thick. industry)
Industry dummies
GVC length (GVC length)
10
See definitions at: http://eur-lex.europa.eu/legal-content/EN/TXT/?uri=URISERV%3An26026
11
Thanks to Sebastian Miroudot of the OECD for providing the raw data necessary to make these calculations. For the few respondent firms operating in multiple GVCs, all GVC
length values per each respondent were identified and then divided by the total number of GVCs in which the firms operated.
17
Number of GVCs in which
operate (Diversification)
Internationalization
level
Foreign operations (Foreign
op.)
Diversification
Number of GVCs in which
operate (Diversification)
Strategic alliances
JV with a foreign partner
(Foreign JV)
Other alliances (Other
alliances)
BPP = bargaining patent portfolio, TPP = traditional patent portfolio. 5-point Likert scale used for patent thickets question: 1 = none at all, 2 = minor magnitude, 3 =
moderate magnitude, 4 = high magnitude, 5 = very high magnitude; 7-point Likert scale used: 0 = N/A my firm has not applied for any patenting support, 1= not useful
at all, 2 = low usefulness, 3 = slightly useful, 4 = neutral, 5= moderately useful, 6 = very useful, 7 = extremely useful
18
Table 2: Descriptive statistics from survey responses
Variable
Mean
Std. dev.
Min.
Max
SFP to prev. lit.
3.93
1.89
1
7
SFP to red. roy.
3.69
1.74
1
7
SFP to cross-lic.
3.55
1.80
1
7
SFP to block
3.78
1.75
1
7
SFP to protect
4.29
1.74
1
7
Patent thick. rating
2.88
0.85
1
5
R&D
2.25
0.69
1
3
Firm size
2.97
0.83
1
4
GVC length
4.88
0.79
2.79
5.67
Diversification
1.46
1.49
1
12
N = 153
Dependent variables
My dependent variables for H1 are the reported usefulness of SFP to help recipient firms accumulate
patents for cross-licensing, blocking, or traditional protective purposes. My dependent variable for H2
is the reported usefulness of SFP to help recipient firms accumulate patents to prevent patent litigation
against themselves. My dependent variable for H3 is the reported usefulness of SFP to help recipient
firms accumulate patents for reducing royalties and licensing fees owed to others.
Independent and mediator variables
My independent variable for H1 and H2 is the magnitude of patent thickets that my survey
respondents faced in China. My H2 mediator variable and H3 independent variable is the reported
usefulness of SFP to help recipient firms accumulate patents for cross-licensing. My H3 mediator
variable is the reported usefulness of SFP to help recipient firms accumulate patents to prevent patent
litigation against themselves.
19
Controls
Several control variables were constructed from the survey data. First, a variable for R&D intensity
was included to proxy technological capabilities. R&D provides absorptive capacity needed to
transform external resources into internal resources (Cohen and Levinthal, 1990) and patenting
requires at least some R&D capabilities (Pakes and Griliches, 1980). Second, a variable for firm size
was included as a rough proxy for firm capabilities generally. Third, the “length” of global value chains
(GVCs) (i.e., the number of intermediate inputs in the value chain, whereby more intermediate inputs
indicate a longer GVC and fewer intermediate inputs indicate a shorter GVC) running through China
was included as a rough proxy of resource interdependencies among firms in a given industry. The
increasingly global nature of value chains, at least at the time that this research was conducted, while
providing opportunities for many Chinese firms has often also relegated them to being low-value-
added suppliers to foreign multinational corporations (MNCs) (e.g., Brandt and Thun, 2010; Fuller,
2013). Fourth, individual industry dummies were included. Fifth, a proxy for the level of respondent
firms’ level of diversification (number of GVCs in which they operated) was included. Sixth, a variable
measuring joint ventures (JVs) that latecomer firms entered into with foreign firms was included as
was, sixth, another variable measuring other types of alliances that latecomer firms entered into with
foreign firms.
Analysis of reliability and validity of data
Several methods were used to explore the reliability and validity of the data and help ensure a basic
level of robustness of the results. First, sensitivity tests and variance inflation factor (VIF) checks, to
ensure VIFs did not exceed 3.3, were conducted for all variables to avoid introducing problematic
multicollinearity into the estimates. Some persistent cases of multicollinearity involving control
variables were identified (for example, when R&D intensity and firm size were included in the same
estimates) and so in these cases problematic controls were not included in the same estimates. Second,
a Kaiser-Meyer-Olkin (KMO) and Bartlett’s Test of Sphericity were conducted, the results of which
(0.631 and p<0.01, respectively) confirm the adequacy of the sampling.
20
Several steps were taken to test the data for common methods bias (CMB). CMB can arise in self-
reported survey responses, often taking the form of responses that are overly positive (“telling the
researcher what he/she wants to hear”) and/or sets of questions being interpreted by respondents as
very similar to one another and therefore answered in very similar ways even though the researcher
intended the questions to measure distinct constructs (Doty and Glick, 1998; Fuller et al., 2016). To
address this, first, the descriptive statistics (see Table 2) were reviewed, which reflect notable variance
and therefore somewhat help alleviate concerns that the survey data might be overwhelmingly biased
towards only the most positive responses or otherwise. Second, Cronbach’s α (alpha) was estimated
for all variables that might be said to form the same latent constructs to test if they should in fact be
aggregated (because they “measure the same thing”) rather than treated distinctly in the estimates. All
possible combinations of the following variables generated a Cronbach’s alpha well below 0.90, the
ideal cut-off for internal consistency, therefore indicating that they should not be aggregated, meaning
they should be treated as distinct observed variables in the regressions: reduction of resource
dependency (SFP to prev. lit, SFP to red. roy.), BPP strategy (SFP to cross. lic., SFP to block), and
TPP strategy (SFP to protect). Third, the confirmatory factor analysis (CFA) technique mentioned in
Fuller et al. (2016) and Doty and Glick (1998) was conducted and none of the corresponding fitness
indexes indicated that CMB significantly undermined the dataset.
Estimation methods
Several econometric techniques were chosen to analyze my survey data based upon the hypothesized
relationships to be tested. Multivariate regressions (Little, 2013) were used to test H1. Mediated
hierarchical regressions (Little, 2013) and non-parametric indirect-effect tests (Preacher and Hayes,
2008) were used to test H2 and H3. Sections 4.1-4.2 depict the results of regressions that include the
most control variables and highest R2 without creating any serious multicollinearity. As explained in
Section 4.4, several steps were taken to further check the robustness of my regressions.
21
4. Results
4.1 SFP’s contribution to cross-licensing amidst patent anti-commons
As illustrated in Table 3, my regression results support H1 that SFP is commonly used by latecomer firms
facing higher incidence of anti-commons (proxied by patent thickets) to accumulate patents for cross-licensing rather than
for other purposes. Model 1a shows that greater incidence of patent thickets faced by latecomer firms is
positively and significantly correlated (0.57, p<0.01) with the usefulness of SFP to accumulate patents
for cross-licensing. Further, Model 1b shows that operation in industries where patent thickets are
most prominent is positively and significantly correlated (1.20, p<0.01) with the usefulness of SFP to
accumulate patents for cross-licensing. Meanwhile, Models 2a and 2b show that neither greater
incidence of patent thickets faced by latecomer firms nor operation in industries in which patent
thickets are most prominent is significantly correlated with the usefulness of SFP to help firms
accumulate patents for blocking. Models 3a and 3b show that neither greater incidence of patent
thickets faced by latecomer firms nor operation in industries in which patent thickets are most
prominent is significantly correlated with the usefulness of SFP to help firms accumulate patents to
protect against free-riding.
22
Table 3: SFP’s contribution to cross-licensing amidst patent anti-commons (H1)
SFP to cross-lic.
SFP to block
SFP to protect
Model #
1a
1b
2a
2b
3a
3b
Patent thick. rating
0.57***
(0.18)
-
0.26
(0.17)
-
0.23
(0.17)
-
Patent thick. industry
-
1.20***
(0.40)
-
0.70
(0.40)
-
0.42
(0.40)
R&D
0.76
(0.40)
0.64
(0.41)
-0.17
(0.40)
-0.24
(0.40)
0.41
(0.39)
0.36
(0.39)
Foreign op.
-0.26
(0.37)
-0.39
(0.36)
-0.19
(0.36)
-0.23
(0.35)
0.37
(0.35)
0.30
(0.34)
Foreign JV
0.40
(0.49)
0.14
(0.49)
0.77
(0.48)
0.65
(0.47)
0.10
(0.47)
-
Other alliances
-0.20
(0.30)
-0.31
(0.30)
-0.15
(0.30)
-0.18
(0.30)
-0.35
(0.29)
-0.37
(0.29)
Diversification
-
-
0.68**
(0.30)
0.58
(0.31)
0.97***
(0.29)
0.92***
(0.30)
GVC length
-0.48
(0.77)
-0.86
(0.80)
-1.08
(0.76)
-1.35
(0.79)
-1.71**
(0.75)
-1.85**
(0.78)
Constant
2.34
(1.29)
4.63***
(1.21)
4.84***
(1.28)
6.04***
(1.20)
5.67***
(1.25)
6.59***
(1.19)
Durbin Watson test
1.83
1.88
1.93
1.97
1.96
1.95
R2
0.11
0.10
0.10
0.10
0.13
0.12
N
153
153
153
153
153
153
Vertically-listed variables are dependent, ones listed horizontally are independent or controls. ***Significant at 1%,
**Significant at 5%. Some controls dropped due to collinearity issues and separately evaluated in robustness checks.
4.2 SFP’s contribution to preventing patent litigation risks
I adopt two methods to test the mediation effects forecasted in H2. Mediation arises when an
independent variable is highly correlated with the dependent variable, mediator is highly correlated
with the dependent variable, and independent and mediator variables are correlated (Little, 2013). If
23
the significance of the correlation between the independent and dependent variable decreases but
remains significant when including the mediator, this constitutes “partial” mediation; if the
significance disappears completely, meaning the direct effect of the independent variable on the
dependent variable is equal to zero, this constitutes “full” mediation (Little, 2013). The Preacher-
Hayes test (Preacher and Hayes, 2008) provides an additional estimate of mediation effects, also
known as “indirect effects”, based on bootstrapped confidence intervals which, to be significant,
should not include zero.
Based on these estimation methods, the results in Table 4 support H2 that SFP’s usefulness to accumulating
patents for cross-licensing serves as a positive mediator between higher incidence of anti-commons (proxied by patent
thickets) and the ability of SFP to help latecomer firms prevent patent litigation against themselves. Incidence of
patent thickets is significantly positively correlated with the usefulness of SFP to prevent litigation
(Model 2a: 0.41, p<0.05; Model 2b: 0.87, p<0.05), SFP’s usefulness to accumulating patents for cross-
licensing is significantly positively correlated with the usefulness of SFP to ultimately prevent patent
litigation (Model 3a/b: 0.61, p<0.01), and incidence of patent thickets is significantly positively
correlated with SFP’s usefulness to accumulate patents for cross-licensing (Model 1a: 0.56, p<0.01;
Model 1b: 1.20, p<0.01). In Models 4a and 4b, the correlation between incidence of patent thickets
and the usefulness of SFP to prevent litigation disappears when also considering the usefulness of
SFP to accumulate patents for cross-licensing, meanwhile the usefulness of SFP to accumulate patents
for cross-licensing remains significantly positively correlated with the usefulness of SFP to prevent
litigation (0.60, p<0.01). These results indicate a full mediation effect is present. This is further
confirmed by the statistically significantly indirect effects identified by the Preacher-Hayes tests (0.32,
p<0.05; 0.74, p<0.05).
24
Table 4: SFP’s contribution to preventing litigation via cross-licensing (H2)
SFP to cross-lic.
SFP to prev. lit.
SFP to cross-lic.
SFP to prev. lit.
Model #
1a
2a
3a
4a
1b
2b
3a/b
4b
Patent thick. rating
0.56***
(0.18)
0.41**
(0.19)
-
0.89
(0.16)
-
-
-
-
Patent thick. industry
-
-
-
-
1.20***
(0.40)
0.87**
(0.43)
-
0.14
(0.37)
SFP to cross-lic. (M)
1
-
0.61***
(0.07)
0.60***
(0.07)
1
-
0.61***
(0.07)
0.60***
(0.07)
R&D
0.77
(0.41)
-
-
-
0.64
(0.41)
-
-
-
Firm size
-
0.53
(0.48)
0.43
(0.40)
0.42
(0.40)
-
0.63
(0.49)
0.43
(0.40)
0.45
(0.40)
Foreign op.
-0.26
(0.37)
-0.58
(0.39)
-0.47
(0.31)
-0.42
(0.32)
-0.39
(0.36)
-0.71
(0.39)
-0.47
(0.31)
-0.47
(0.32)
Foreign JV
0.40
(0.49)
-0.07
(0.52)
-0.44
(0.42)
-0.39
(0.43)
0.14
(0.49)
-0.28
(0.52)
-0.44
(0.42)
-0.46
(0.43)
Other alliances
-0.22
(0.30)
-0.28
(0.32)
-0.15
(0.27)
-0.15
(0.27)
-0.31
(0.30)
-0.25
(0.32)
-0.15
(0.27)
-0.06
(0.26)
Diversification
0.09
(0.31)
0.48
(0.32)
0.46
(0.27)
0.45
(0.27)
-
-
0.46
(0.27)
-
GVC length
-0.45
(0.78)
-0.44
(0.86)
-0.27
(0.71)
-0.30
(0.71)
-0.90
(0.80)
-0.97
(0.89)
-0.27
(0.71)
-0.49
(0.74)
Constant
2.29
(1.31)
3.37**
(1.38)
2.13
(1.09)
1.91
(1.15)
4.63***
(1.21)
5.34***
(1.29)
2.13
(1.09)
2.51**
(1.12)
Durbin Watson test
1.83
2.17
2.31
2.32
1.88
2.19
2.31
2.29
Preacher-Hayes test
-
-
-
0.32**
(0.12)
-
-
-
0.74**
(0.25)
R2
0.11
0.10
0.39
0.39
0.10
0.07
0.39
0.37
N
153
153
153
153
153
153
153
153
Vertically-listed variables are dependent, ones listed horizontally are independent or controls. (M) = hypothesized mediator.
***Significant at 1%, **Significant at 5%. Some controls dropped due to collinearity issues and separately evaluated in
robustness checks.
25
4.3 SFP’s contribution to reducing royalties and licensing fees owed to others
The results in Table 5 support H3 that SFP’s usefulness to accumulate patents for cross-licensing is positively
related to its contribution to reducing royalties owed to external organizations, and this relationship is positively mediated
by SFP’s contribution to preventing litigation. The usefulness of SFP to help recipient firms accumulate
patents for cross-licensing PPs is significantly positively correlated with the usefulness of SFP to
reduce royalties owed to external organizations (Model 1b: 0.61, p<0.01), the usefulness of SFP to
accumulate patents for cross-licensing is significantly positively correlated with the usefulness of SFP
to prevent litigation (Model 1a: 0.60, p<0.01), and SFP’s usefulness to prevent litigation is significantly
positively correlated with SFP’s usefulness to reduce royalties (Model 1c: 0.45, p<0.01). In Model 1d,
there is a positive and significant correlation between SFP’s usefulness to accumulating patents for
cross-licensing and SFP’s usefulness to reduce royalties (0.50, p<0.01) when also considering the
usefulness of SFP to preventing patent litigation. Meanwhile, the usefulness of SFP to preventing
patent litigation remains significantly positively correlated, although slightly less so than in Model 1b,
with SFP’s usefulness to reduce royalties (0.18, p<0.01). These results indicate the presence of a partial
mediation effect. This is further confirmed by the statistically significantly indirect effects identified
by the Preacher-Hayes test in Model 1d (0.11, p<0.05).
26
Table 5: SFP’s contribution to preventing litigation and reducing royalties (H3)
SFP to prev. lit.
SFP to red. roy.
Model #
1a
1b
1c
1d
SFP to cross-lic.
0.60***
(0.07)
0.61***
(0.06)
-
0.50***
(0.07)
SFP to prev. lit (M)
1
-
0.45***
(0.06)
0.18***
(0.07)
SFP to protect
-
-
-
-
R&D
-
-0.51
(0.29)
-0.21
(0.33)
-0.49
(0.29)
Firm size
0.42
(0.40)
-
-
-
Foreign op.
-0.42
(0.32)
-
-
-
Foreign JV
-0.39
(0.43)
1.14***
(0.34)
1.43***
(0.39)
1.20***
(0.34)
Other alliances
-0.15
(0.27)
-0.30
(0.25)
-0.30
(0.25)
-0.26
(0.21)
Diversification
0.45
(0.27)
0.21
(0.22)
0.05
(0.25)
0.13
(0.22)
Patent thick. industry
-
-
-
-
Patent thick. rating
0.09
(0.16)
-
0.10
(0.14)
-
GVC length
-0.30
(0.71)
-
-1.03
(0.64)
-0.90
(0.55)
Constant
1.91
(1.15)
3.29
(0.88)
3.39***
(1.06)
2.96***
(0.87)
Durbin Watson test
2.3
1.94
1.95
1.97
Preacher-Hayes test
-
-
-
0.11**
(0.05)
R2
0.39
0.50
0.36
0.50
N
153
153
153
153
Vertically-listed variables are dependent, ones listed horizontally are independent or controls. (M) = hypothesized mediator.
***Significant at 1%, **Significant at 5%. Some controls dropped due to collinearity issues and separately evaluated in robustness
checks.
27
4.4 Robustness checks
Several steps were taken to check the robustness of my analysis. First, a key robustness check already
illustrated in Tables 3-5 was to run the regressions with the self-reported patent thickets rating as well
as the objective measure of patent thickets (the industries where technology is highly cumulative
dummy). Second, all control variables not depicted that were removed from the models presented
hereto due to the problematic multicollinearity were tested in separated regressions individually with
the independent and dependent variables, and alongside as many other control variables as possible
without creating new multicollinearity. This included twelve different industry dummies and other
controls. Third, as illustrated in Tables 4-5, the Preacher and Hayes (2008) non-parametric indirect-
effect process macro was used to confirm the existence/non-existence of mediation effects. Fourth,
the natural log was taken of the dependent and independent variables and they were then analyzed in
separate regressions to ensure all distributions were as smooth as possible. Fifth, all of the
econometrics were run with and without bootstrapping. Sixth, path-analysis structural equation
modelling (SEM) was used for the sake of testing the hypotheses with a different statistical technique.
None of these checks generated results that altered the statistical significance of the hypothesis-
relevant correlations presented in the regressions in Tables 3-5 nor generated significantly higher R2
values. Therefore, the results presented are thought to be reasonably robust. Still, the results do not
necessarily represent causal relationships, rather reflect correlations, and I interpret them accordingly throughout this
paper.
5. Discussion
5.1 Contributions
Theoretical implications
The main theoretical contribution I offer with this paper is that the Chinese government appears to
have leveraged an unconventional policy tool state financing for patenting/SFP to temporarily
alleviate some anti-commons problems associated with patent thickets without also substantially
28
weakening the appropriability offered by the country’s IPR legal institutions. The most common IPR
policy approaches to reducing patent anti-commons can simultaneously restrain technological
development. However, rather than reducing anti-commons by substantially weakening IPR legal
institutions, SFP appears to have provided additional resources to recipient firms to accumulate
patents for cross-licensing, which enables them to bargain to reduce patent litigation risks and, in turn,
reduce royalties owed to others. These benefits appear most useful to firms in industries where
technology is highly cumulative, as they tend to face more significant patent thickets. In other words,
SFP affords recipient latecomer firms greater bargaining power in the resource-based sense (e.g., Hall,
1992; Casciaro and Piskorski, 2005) which they can leverage to reduce prominent IPR resource
dependencies on incumbents amidst anti-commons. At the same time, however, and very importantly, I
acknowledge the negative policy externalities simultaneously experienced from sub-optimally designed
SFP in China which I discuss further in the next (policy implications) section.
Policy implications
Considering the aforementioned theoretical contributions, it may be economically rational for
governments from developing economies looking to China for policy leadership to consider
temporarily alleviating anti-commons by leveraging SFP to enable domestic firms to accumulate more
property rights. SFP may be most useful to policymakers in large upper-middle-income economies in
particular because IPR institutions have been shown to matter more to their technological
development compared with smaller, lower-income economies (e.g., World Bank, 2001; Grossman
and Lai, 2004; Hall, 2014; Kim et al., 2012; Gentile, 2020).
However, and very importantly, policymakers must be cautious about how they go about rewarding
SFP. In order to avoid potential claims from foreign stakeholders that SFP is a discriminatory form
of state aid, it should only be offered in limited amounts and only to needy latecomer firms rather
than to all domestic firms. Further, governments must carefully deploy SFP or it, like its alternative
IPR policy approaches, may restrain technological development. SFP, unlike patentability restrictions
and compulsory licensing (e.g., OECD, 1989; Merges, 1996; Burk and Lemley, 2003) or patent pools
(e.g., Merges, 1996; Geiger and Gross, 2017), does not attempt to reduce the fragmentation of patent
ownership that creates patent anti-commons in the first place. As such, although SFP may temporarily
29
alleviate patent anti-commons, it by nature cannot “correct” them. Moreover, long-term reliance on
SFP inevitably will seriously exacerbate anti-commons by further fragmenting patent
ownership/making patent thickets even denser. It will also perversely distort recipient firms’ survival
incentives, cost significant public resources, and potentially lock-in the state’s dependency on a costly
policy tool. Finally, and importantly, if SFP programs are not accompanied by at least basic
qualification requirements, applicants may file low-quality/“junk” patents which can create path
dependency on imitation and otherwise stifle genuine innovation something that has often afflicted
SFP in China (e.g., Prud’homme, 2012, 2016; Dang and Motohashi, 2015; Long and Wang, 2016, 2019;
Prud’homme and Zhang, 2019).
Considering these negative policy externalities, it appears that several safeguards should be built into
SFP policies. First, SFP policies could benefit from stricter qualification requirements than those
imposed by the Chinese state. Secondly, if adopted, SFP should only be instituted for a few years in
order to avoid seriously exacerbating anti-commons problems. Third, governments could require that
in order to receive financial support from SFP for expanding bargaining portfolios, local firms must
demonstrate, via a patent map or FTO analysis, that they indeed face significant patent thickets from
incumbent firms. At the same time, I caution policymakers in developing countries against being overly
preoccupied with fostering only high “quality” patents, i.e., those with high social value. Just because
patents do not meet some traditional definitions of patent “quality” used by academics does not
necessarily mean they do not have strategic private value (e.g., Lerner, 2015), especially to latecomer firms from
developing nations. SFP can help expand latecomers’ portfolios of patents with strategic private value.
5.2 Limitations
There are a number of limitations of my survey used for this paper. First, some of the survey questions
are rather subjective. (This being said, as mentioned in Section 3.3, the checks for CMB provide
reasonable confidence in the reliability and validity of the data received.) Second, there is likely some
selection bias in my sample given that the firm representatives invited to and involved in state-
organized trainings that I sampled from might be more serious about properly managing their IPR
than some firms that did not participate in the trainings. As such, it is not certain that the firms sampled
are representative of “average” Chinese firms in Jiangsu. Nor do I attempt to measure the impact of
SFP on firms in other subnational regions in China. Third, since all firms in my sample indicated that
30
they received both types of the SFP I analyzed, a control group could not be formed.
12
I therefore
focused on variances among firms receiving SFP. Fourth given that most of my survey respondents
were only willing to support my research on the basis of anonymity, I could not match their responses
with patent office data.
Fifth, my survey does not capture the full range of firm patenting strategies. Even though, as
mentioned in the literature reviewed in Section 2 and confirmed by my results, firms facing significant
patent thickets most often follow a BPP strategy rather than a TPP strategy, this does not necessarily
mean that firms following the BPP strategy or TPP strategy do not also patent for other reasons. The
relationship between receipt of SFP and these alternative motives including patenting as signals to
raise capital, defensive patenting, patenting to build reputation, and so on deserves to be explored
in future research. Sixth, my survey respondents do not distinguish which incumbents they are
bargaining with amidst patent thickets: it could be forerunner Chinese firms from Jiangsu or different
provinces in China, foreign incumbents, or both. However, judging from the literature indicating that
Chinese firms’ FTO in China is significantly limited by foreign MNCs (e.g., Zhang et al., 2014; Song,
2013; Xiao, 2013; Keupp et al., 2012), it seems likely that at least some of this bargaining is conducted
with foreign incumbents. Seventh, no controls for the social value of patents were available to be
included in my statistical analysis.
My research also faces several other limitations. Judging from the relatively small coefficients in many
of my results, the contribution of SFP to latecomer firms’ larger ability to actually become more
competitive should not be over-emphasized. SFP is likely a relatively small contributor, although the
exact extent of its contribution deserves to be validated by additional research. Also, it is unclear
exactly how much better, if it all, SFP is than undiscussed strategies that latecomer firms may adopt
related to business model innovation, for example, to address anti-commons problems. I consider
each of these limitations when cautiously wording my findings. However, none of these limitations
should undermine the strong descriptive value of this paper as a contribution to an otherwise highly
under-researched yet important topic involving the Chinese economy.
12
Four firms indicated that they did not receive both types of incentives being analysed; however, those respondents left most questions
incomplete and therefore such responses were unusable.
31
5.3 Future research
Future research could address the aforementioned limitations and extend my research in several ways.
It would be helpful to investigate the issues discussed in this paper with patent data for firms proven
to have received SFP compared to a control group of firms that did not receive SFP. Future studies
could also collect more objective data, alongside survey data, about how exactly SFP shapes latecomers’
competitive positions vis-a-visa incumbents. Further research is also warranted into the type of
dynamic capabilities that firms must possess in order to actually efficiently churn SFP resources into
patent portfolios and strategically use these portfolios to reduce problems associated with anti-
commons.
6. Conclusions
In this paper, I show how the Chinese government appears to have leveraged an unconventional
policy tool state financing for patenting/SFP to temporarily alleviate some anti-commons
problems associated with patent thickets without also substantially weakening the appropriability
offered by the country’s IPR legal institutions. In a sense, SFP has allowed the state to avoid some of
the traditional IPR policy dilemmas when responding to patent anti-commons. I go on to explain,
however, that the design of SFP in many places in China has also unwittingly created negative policy
externalities. I therefore recommend that governments from developing countries turning to China
for leadership in technological development policy must more carefully design and implement SFP at
home.
Acknowledgements
Thanks to Anthony Howell, Geng Xuesong, John Mathews, and Mark Cohen, who provided helpful
comments on earlier drafts of this paper or directly related research. Thanks also to Long Xingle, Song
Hefa, and Chen Xiangdong for providing helpful comments on the design of the survey instrument
used in the research. And thanks to Nick Papageorgiadis for inviting me to present a version of this
paper at the University of Liverpool Management School in 2019.
32
References
Arrow, K., 1962. Economic Welfare and the Allocation of Resources for Invention, in: Nelson, R.
(Ed.), The Rate and Direction of Inventive Activity, Economic and Social factors. Princeton, NJ.
Arundel, A., 2001. The relative effectiveness of patents and secrecy for appropriation. Research Policy
30, 611-624.
Blind, K., Edler, J., Frietsch, R., Schmoch, U., 2006. Motives to patent: Empirical evidence from
Germany. Research Policy 35, 655-672.
Blind, K., Cremers, K., Mueller, E., 2009. The influence of strategic patenting on companies’ patent
portfolios. Research Policy 38, 428-436.
Brandt, L., Thun, E., 2010. The fight for the middle: upgrading, competition, and industrial
development in China. World Development 38, 1555-1574.
Burk, D., Lemley, M., 2003. Policy levers in patent law. Virginia Law Review 89, 1575-1696.
Casciaro, T., Piskorski, M., 2005. Power imbalance, mutual dependence, and constraint absorption: A
closer look at resource dependence theory. Administrative Science Quarterly 50, 167-199.
Cheng, W., Drahos, P., 2017. How China built the world’s biggest patent office – the pressure driving
mechanism. IIC International Review of Intellectual Property and Competition Law 49, 5-40.
Chien, C., 2014. Holding up and holding out. Michigan Telecommunications & Technology Law Review 21,
1-41.
Cohen, W., Goto, A., Nagata, A., Nelson, R., Walsh, J., 2002. R&D spillovers, patents, and the
incentives to innovate in Japan and the United States. Research Policy 31, 1349-1367.
Choi, J.P., Gerlach, H., 2015. Patent pools, litigation, and innovation. RAND Journal of Economics 46,
499-425.
Cockburn, I., MacGarvie, M., Muller, E., 2010. Patent thickets, licensing and innovative performance.
Industrial and Corporate Change 19, 899-925.
Cohen, W., Levinthal, D., 1990. Absorptive capacity: A new perspective on learning and innovation.
Administrative Science Quarterly 35, 128-152.
Czarnitzki, D., Hussinger, K., 2017. Input and output additionality of R&D subsidies. Applied
Economics 50, 1324-1341.
Dang, J., Motohashi, K., 2015. Patent statistics: A good indicator for innovation in China? Patent
subsidy program impacts on patent quality. China Economic Review 35, 137-155.
De Backer, K., Miroudot, S., 2013. Mapping global value chains. OECD Trade Policy Paper No. 159.
33
Deng, P., 2013. Chinese outward direct investment research: Theoretical integration and
recommendations. Management and Organization Review 9, 513-539.
Dutfield, G., Suthersanen, U. 2005. Harmonisation or differentiation in intellectual property
protection? The lessons of history. Prometheus, 23: 131-147.
Fuller, C., Simmering, M., Atinc, G., Atinc, Y., Babin, B., 2016. Common methods variance detection
in business research. Journal of Business Research 69, 3192-3198.
Fuller, D., 2013. Building ladders out of chains: China’s hybrid-led technological development in
disaggregated value chains. Journal of Development Studies 49, 547-563.
Galasso, A., 2012. Broad cross-license bargainings. Journal of Economics & Management Strategy 21, 873-
911.
Galasso, A., Schankerman, M., 2010. Patent thickets, courts, and the market for innovation. RAND
Journal of Economics 41, 472-503.
Gallini, N., Winter, R., 1985. Licensing in the theory of innovation. The RAND Journal of Economics 16,
237-252.
Geiger, S., Gross, N., 2017. Market failures and market framings: Can a market be transformed from
the inside? Organization Studies 1-20.
Gentile, E., 2020. Intellectual property rights and foreign technology licensing in developing countries:
An empirical investigation. Economic Development and Cultural Change 68.
Ginarte, J., Park, W.G., 1997. Determinants of Patent Rights: A Cross-National Study. Research Policy
26, 283-301.
Groizard, J., 2009. Technology trade. Journal of Development Studies 9, 1526-1544.
Grossman, G., Lai, E., 2004. International protection of intellectual property. American Economic Review
94, 1635-1653.
Hall, B., 2005. Exploring the patent explosion. Journal of Technology Transfer 30, 35-48.
Hall, B., 2014. Does patent protection help or hinder technology transfer?, in: Ahn, S., Hall, B.H., Lee,
K. (Eds), Intellectual Property for Economic Development. Edward Elgar and Korean Development
Institute, Northhampton, Mass., pp 11-32.
Hall, B., Ziedonis, R, 2001. The patent paradox revisited: an empirical study of patenting in the US
semiconductor industry, 1975-1995. RAND Journal of Economics 32, 101-128.
Hall, R., 1992. The strategic analysis of intangible resources. Strategic Management Journal 13, 135-144.
Hegde, D., Mowery, D., Graham, S., 2009. Pioneering inventors or thicket builders: which U.S. firms
use continuations in patenting? Management Science 55, 1214-1226.
34
Heller, M., 1998. The tragedy of the anticommons: Property in the transition from Marx to
markets. Harvard Law Review 111, 62188.
Heller, M., Eisenberg, R., 1998. Can patents deter innovation? The anticommons in biomedical
research. Science 280, 698-701.
Hu, A.G., Zhang, P., and L. Zhao. 2016. China as number one? Evidence from China’s most recent
patenting surge. Journal of Development Economics 124, 107119.
Huang, C., 2017. Recent development of the intellectual property rights system in China and
challenges ahead. Management and Organization Review 13, 39-48.
Huang, K., Murray, F., 2009. Does patent strategy shape the long-run supply of public knowledge?
Evidence form human genetics. Academy of Management Journal 52, 1193-1221.
Hsu, F., Horng, D.J., Hsueh, C.C., 2009. The effect of government-sponsored R&D programmes on
additionality in recipient firms in Taiwan. Technovation 29, 204-2017.
Keupp, M., Friesike, S., von Zedtwitz, M., 2012. How Do Foreign Firms Patent in Emerging
Economies with Weak Appropriability Regimes? Archetypes and Motives. Research Policy 41, 1422-
1439.
Kim, L., 1997. Imitation to Innovation: The Dynamics of Korea’s Technology Learning. Harvard Business School
Press, Cambridge, Mass.
Kim, Y. K., Lee, K., Park, W.G., Choo, K., 2012. Appropriate intellectual property protection and
economic growth in countries at different levels of development. Research Policy 41, 35875.
Kumar, N., 2003. Intellectual Property Rights, Technology and Economic Development: Experiences
of Asian Countries. Economic and Political Weekly 38, 209-226.
Lampe, R., Moser, P., 2016. Patent pools, competition, and innovation Evidence from 20 industries
under the New Deal. Journal of Law, Economics, and Organization 32, 1-36.
Lee, N. 2015. Intellectual property law in Chinafrom legal transplant to governance, in: Lee, N.,
Bruun, N., Li, M. (Eds), Governance of Intellectual Property Rights in China and Europe. Edward Elgar,
UK, pp 519.
Lei, Z., Sun, Z., Wright, B., 2013. Patent subsidy and patent filing in China. Fung Institute Working Paper.
Lerner, J., Baker, M., Speen, A., Leamon, A., 2015. Financial patent quality: Finance patents after State
Street. Harvard Business School Working Paper 16-068.
Li, X., 2012. Behind the recent surge of Chinese patenting: An institutional view. Research Policy, 41,
236-249.
35
Liefner, I., Kroll, H., Peighambari, A., 2016. Research-driven or party-promoted? Factors affecting
patent applications of private small and medium-sized enterprises in China’s Pearl River Delta.
Science and Public Policy 43, 849-858.
Little, T., 2013. The Oxford Handbook of Quantitative Methods, Vol. 2 Statistical Analysis. Oxford University
Press, New York.
Lemley, M. A., and C. Shapiro. 2007. Patent holdup and royalty stacking. Texas Law Review 85, 1991
2049.
Levin, R., Klevorick, A., Nelson, R., Winter, S., 1987. Appropriating the Returns from Industrial
Research and Development. Brookings Papers on Economic Activity.
Long, C.X., Wang, J., 2016. Evaluating patent promotion policies in China: consequences on patent
quantity and quality, in: Prud’homme, D., Song, H. (Eds) Economic impacts of intellectual property-
conditioned government incentives, Springer, pp 235-258.
Long, X., Wang, J., 2019. China’s patent promotion policies and its quality implications. Science and
Public Policy 46, 91-104.
Mansfield, E., 1984. R&D and Innovation: Some Empirical Findings, in: R&D, Patents, and Productivity,
Griliches, Z. (Ed). University of Chicago Press, Chicago, IL. pp 127-154.
Merges, R., 1994. Intellectual property rights and bargaining breakdown: The case of blocking patents.
Tennessee Law Review 62, 75-106.
Merges, R., 1996. Contracting into liability rules: Intellectual property rights and collective rights
organizations. California Law Review 84, 1293-1393.
Murray, F., Stern, S., 2007. Do formal intellectual property rights hinder the free flow of scientific
knowledge? An empirical test of the anti-commons hypothesis. Journal of Economic Behavior &
Organization 63, 648-687.
Odagiri, H., Goto, A., Sunami, A., Nelson, R. (2010a). Introduction, in: Odagiri, H., Goto, A., Sunami,
A., Nelson, R (Eds), Intellectual Property Rights, Development, and Catch-Up. Oxford University Press,
Oxford, UK, pp 1-28.
Odagiri, H., Goto, A., Sunami, A. (2010b). IPR and the catch-up process in Japan, in: Odagiri, H.,
Goto, A., Sunami, A., Nelson, R., (Eds). Intellectual Property Rights, Development, and Catch-Up. Oxford
University Press, Oxford, UK, pp 95-132.
OECD, 1989. Competition Policy and Intellectual Property Rights. OECD Publishing.
OECD, 2012. Mapping global value chains. OECD Publishing.
Park, W., 2008. International patent protection 1960-2005. Research Policy 37, 761-766.
36
Preacher, K., Hayes, A., 2008. Asymptotic and resampling strategies for assessing and comparing
indirect effects in multiple mediator models. Behavior Research Methods 40, 879-891.
Prud’homme, D., 2012. Dulling the Cutting Edge: How Patent-Related Policies and Practices Hamper Innovation
in China. European Union Chamber of Commerce in China.
Prud’homme, D., 2016. IP-conditioned government incentives in China and the EU: A comparative
analysis of strategies and impacts on patent quality, in: Prud’homme, D., Song, H. (Eds) Economic
Impacts of Intellectual Property-Conditioned Government Incentives, Springer, pp 13-74.
Prud’homme, D., Long, X., 2019. Case studies of Chinese firms' experiences with intellectual
property-conditioned state incentives. International Journal of Technology, Policy and Management 19,
213-225.
Prud’homme, D., Zhang, T., 2020. China’s Intellectual Property Regime for Innovation: Risks to Business and
National Development. Springer.
Shapiro, C., 2000. Navigating the patent thicket: Cross licenses, patent pools, and standard setting.
Innovation policy and the economy 1, 119-150.
Shapiro, C., 2002. Competition Policy and Innovation, OECD STI Working Papers.
Shu, C., Wang, Q., Gao, S., Liu, C. 2014. Firm patenting, innovations, and government institutional
support as a double-edged sword. Journal of Product Innovation Management 32, 290-305.
Somaya, D., 2012. Patent strategy and management: An integrative review and research agenda. Journal
of Management 38, 1084-1114.
Song, H., 2013. Indigenous Innovation Capability-Building and the Development of Intellectual Property Rights (in
Chinese). Chinese Academy of Sciences Intellectual Property Rights Press, Beijing.
Song, H., Li, Z., Xu, D., 2016. Study on policies to stimulate patenting in China: what led to the
upsurge of China’s domestic patent applications? in: Prud’homme, D., Song, H. (Eds) Economic
Impacts of Intellectual Property-Conditioned Government Incentives, Springer, pp 179-204.
Su, Y.S., Gelman, A., Hill, J., Yajima, M., 2011. Multiple imputation with diagnostics (mi) in R:
Opening windows into the black box. Journal of Statistical Software 45, 1-31.
Torrisi, S., Gambardella, A., Giuri, P., Harhoff, D., Hoisl, K., Mariani, M., 2016. Used, blocking and
sleeping patents: Empirical evidence from a large-scale inventor survey. Research Policy 45, 1374-
1385.
von Graevenitz, G., Wagner, S., Hoisl, K., Hall, B.H., Harhoff, D., Giuri, P., Gambardella, A., 2007.
The Strategic Use of Patents and its Implication for Enterprise and Competition Policies. Report for the
European Commission.
37
von Graevenitz, G., Wagner, S., Harhoff, D., 2013. Incidence and growth of patent thickets: the
impact of technological opportunities and complexity. Journal of Industrial Economics 62, 521-563.
World Bank, 2001. Intellectual property: Balancing incentives with competitive access, in: Global
Economic Perspectives. World Bank, pp. 129150.
Xiao, Y., Tylecote, A., Liu, J., 2013. Why not greater catch-up by Chinese firms? The impact of IPR,
corporate governance and technology intensity on late-comer strategies. Research Policy 42, 749-764.
Xu, L., Munari, F., 2016. The impact of public support for SMEs’ patenting activity: Empirical
evidence from Italy, in: Prud’homme, D., Song, H. (Eds.) Economic Impacts of Intellectual Property-
Conditioned Government Incentives, Springer, pp 101-132.
Zhang, G., Lv, X., Zhou, J., 2014. Private value of patent right and patent infringement: An empirical
study based on patent renewal data of China. China Economic Review 28, 37-54.
Zhu, X. 2013. Dialectical perspective of patent quantity and quality in China (in Chinese). Bulletin of
Chinese Academy of Sciences 4, 003.
Ziedonis, R., 2004. Don’t fence me in: Fragmented markets for technology and patent acquisition
strategies of firms. Management Science 50, 804-820.
ResearchGate has not been able to resolve any citations for this publication.
Article
Full-text available
Based on China's patent data from 1985 to 2010, we find that behind the country's patent number explosion, the overall innovation content of Chinese patents has not achieved proportional growth. An important explanation for this phenomenon is the patent promotion polices (PPPs), which have prompted the quantitative expansion of patent applications and approvals, but have had negative effects on average patent quality, as proxied by patent application withdrawal rate and patent renewal rate. These patterns are observed both at the provincial level and at the patent level, and they are especially pertinent to patent applications filed by firms, which are the main targets of the PPPs. © The Author(s) 2018. Published by Oxford University Press. All rights reserved.
Article
Full-text available
In 2011 China’s patent office received more patent applications than any other patent office in the world. While explanations for this patent surge focus on some relevant factors such as the use of subsidies for application fees, what is missing from the literature is an analysis of how China turns its fragmented levels of government into an efficient system for obtaining compliance with patent targets set by the higher levels of government. Drawing on the governance literature from China we introduce the concept of the pressure driving mechanism and show how this mechanism enables China to reach the goals and targets that it sets for its patent system. We discuss goal and target-setting at the higher levels of government and show how these are then transmitted to lower levels of government. We explain how performance evaluation and peer transparency are used to generate pressure on Chinese officials to comply with patent targets. Provincial-level patent data are presented to show how well compliance works, as well as to demonstrate how the mechanism operates to make the patent system responsive to changes in targets. Through the application of the pressure driving mechanism, China is developing a unique tool of patent regulation.
Book
This book evaluates the risks that China’s intellectual property (IP) regime poses to innovation. China's IP regime has been heavily criticized as potentially stifling innovation. However, the country’s innovation capabilities have risen significantly and major reforms have recently been made to its IP regime. How risky, really, is China's IP regime for innovation? This book investigates this question at different units of analysis based on a multidisciplinary assessment involving law, management, economics, and political science. Specifically, it critically appraises China's substantive IP laws, measures for boosting patent quantity and quality, measures for transmitting and exploiting technological knowledge, new experimental IP measures, and China's systems for administering and enforcing IP. Practitioners and scholars from various backgrounds can benefit from the up-to-date analysis as well as the practical managerial tools provided, including risk assessment matrices for businesses and recommendations for institutional reform.
Article
How do actors innovate markets in cases of perceived market failures? This paper’s aim is to examine what happens when a market is innovated or, as we call it, ‘redevised’ in situations where public and commercial interests significantly diverge. Market devices can serve an important function in such attempts to innovate markets: they are material and/or social arrangements that are put into place to shape the market in question in certain ways. But can such devices really transform a market from within? To examine this question we trace the history of the Geneva Medicines Patent Pool, a civil society initiative introduced to change pharmaceutical firms’ licensing and collaboration practices in the market for HIV/AIDS medicines. Our empirical results indicate that redevising a market in response to market failures can shift the market’s frames and contribute to altering its practices, but that this is a pragmatic and often lengthy process that is never fully predictable in advance. By attending to the intended and unintended consequences - or misfires - of redevising a market, our study raises important questions around acting in and on the market, market innovation’s’ ontological impact, zooming in and zooming out when studying redevising, and attending to the temporality of market innovation.
Article
This article analyses the effects of public R&D subsidies on R&D input and output of German firms. We distinguish between the direct impact of subsidies on R&D investment and the indirect effect on innovation output measured by patent applications. We disentangle the productivity of purely privately financed R&D and additional R&D investment induced by the public incentive scheme. For this, a treatment-effect analysis is conducted in a first step. The results are implemented into the estimation of a patent production function in a second step. It turns out that both purely privately financed R&D and publicly induced R&D show a positive effect on patent outcome.
Article
As Peng, Ahlstrom, Carraher, and Shi (2017) rightly noted, Intellectual Property Rights (IPR) protection in a country is not static. It evolves over time. Peng et al. (this issue) revealed through their historical analysis that during the 19 th century, the US was not a leading IPR advocate but a leading IPR violator. It was only when indigenous inventors, authors, and organizations of the US emerged and demanded protection of their IPR in foreign countries in the late 19 th century that the US passed the International Copyright Act (the Chace Act) in 1891 to extend IPR protection to foreign works. The US case illustrated that a country's IPR system as an institution evolves as its economy and society develop. If we examine this evolution over a relatively long time span, the change can be quite dramatic. Therefore, when reviewing a country's IPR system, an important question to be asked is in which direction the country's IPR system evolves.