Research and development strategy for environmental technology in [621317]

Research and development strategy for environmental technology in
Japan: A comparative study of the private and public sectors
Hidemichi Fujii, Shunsuke Managi ⁎
aGraduate School of Fisheries and Environmental Sciences, Nagasaki University, 1-14 Bunkyo-machi, Nagasaki 852-8521, Japan
bUrban Research Canter, Department of Urban and Environmental Engineering, School of Engineering, Kyushu University, 744 Motooka, Nishiku, Fukuok a, 819-0395, Japan
cQUT Business School, Queensland University of Technology, Level 8, Z Block, Gardens Point, 2 George St, Brisbane, QLD 4000, Australia
abstract article info
Article history:
Received 25 November 2015Received in revised form 17 February 2016Accepted 18 February 2016Available online 2 March 2016Environmental protection technology plays an important role in a sustainable society, simultaneously promoting
economic development and pollution control. This study examines the determinants of technology inventions re-lated to environmental protection in Japan. We use patent application data in a decomposition analysis frame-
work. We find that environmental patent applications increase according to the prioritization of
environmental patents by private companies and according to ef ficiency improvements in patent applications
in the public sector. Additionally, patent applications related to emission trading increased rapidly among privatecompanies, mainly due to their increased priority after 2005. The different determinants of environmental tech-
nologies between the private and public sectors are useful for formulating effective policies to promote environ-
mental innovation.
© 2016 Elsevier Inc. All rights reserved.Keywords:
Green inventionDecomposition analysisResearch and development strategyPatent dataLog mean Divisia index
1. Introduction
Environmental protection technology (hereafter, environmental
technology) plays an important role in effectively and economically
controlling pollutant emissions. In this way, such technology contrib-
utes to the creation of a sustainable society, that is, one balancing eco-
nomic development and environmental protection ( Sun et al., 2008 ).
The global importance of environmental technology has been increas-
ing. Environmental technology is listed as a high priority in Japan's 5th
science and technology basic plan, which covers 2016 to 2020. Similarly,
the U.S. government budgeted approximately 7.4 billion U.S. dollars for
clean energy technology programs in 2016. Additionally, Germany's
high-tech innovation strategy, introduced in 2014, includes creating a
sustainable economy and energy supply as one of its six priority tasks.
These research and development strategies focus on the expansion of
the environmental business market and are intended to improve inter-
national market competitiveness.
However, not all environmental technologies contribute equally to
improved pollution control and resource conservation. Technology for
environmental protection is diverse and comes in many forms. Someof these are pollution control technologies applicable to waste manage-
ment, air pollution control, and wastewater treatment. The resource
conservation area includes renewable energy, energy ef ficiencyimprovements, and energy-saving products. It is clear that the market
demand for and cost of inventions differs depending on the type of en-
vironmental technology. Therefore, it is important to consider the char-
acteristics of each environmental technology when suggesting an
economical and effective environmental technology invention system.
To understand the characteristics of environmental technologies, clari fi-
cation is imperative.
The clari fication of environmental technology was introduced by the
Organisation for Economic Co-operation and Development (OECD,
2009) and the World Intellectual Property Organization (WIPO, see
http://www.wipo.int/classi fications/ipc/en/est/). Meanwhile, previous
literature focusing on the characteristics of environmental technology
patents (hereafter, environmental patents) is limited, and most studies
focus on the U.S. and European countries ( Fujii, 2016 ). In recent years,
several academic studies have focused on speci fic environmental tech-
nologies, such as wind energy technology in Europe ( Lindman and
Söderholm, 2015 ) and green chemistry in Japan ( Fujii, 2016 ).
Fujii (2016) applied a factor decomposition analysis to identify the
determinants of patent applications related to green chemistry in
Japan. This study addressed green chemistry but not the other environ-mental technologies. Therefore, pollution control and alternative energy
technologies, which have different characteristics from green chemis-
try, are not discussed in Fujii (2016) . Additionally, few previous studies
have used Japanese environmental patent data. Therefore, there is no
previous research providing a factor decomposition analysis of the de-
terminants of environmental patents in Japan that focuses on the char-
acteristics of each technology. However, the results of such a factorTechnological Forecasting & Social Change 112 (2016) 293 –302
⁎Corresponding author.
E-mail addresses: hidemichifujii@nagasaki-u.ac.jp (H. Fujii), managi.s@gmail.com
(S. Managi).
http://dx.doi.org/10.1016/j.techfore.2016.02.012
0040-1625/© 2016 Elsevier Inc. All rights reserved.
Contents lists available at ScienceDirect
Technological Forecasting & Social Change

analysis, which considers the characteristics of speci fic environmental
technologies, are important for creating effective research and develop-
ment policy.
This study tries to clarify the determinants of Japanese environmen-
tal patents from 2001 to 2010. During this period, the so-called “lost de-
cades, ”the Japanese economy experienced slow growth due to high
appreciation of the yen and reduced consumer spending in the domesticmarket ( Hamada and Okada, 2009; Lise et al., 2014 ). Studies of how re-
search and development advanced in Japan during the lost decades are
limited, especially concerning environmental technologies. However,
according to the Organisation for Economic Co-operation and Develop-
ment (OECD) (2014) , Japan led the world in high-value inventions in
environmental technology between 2009 and 2011. This pattern
shows that Japan was highly competitive in research and development
technology in 2011. Interestingly, Japan invented environmental tech-
nologies during an economic depression, during which reductions in
R&D expenditures would normally be expected.
Fig. 1 presents the number and share of patent applications for envi-
ronmental technologies from 1990 to 2010. The bars illustrate the num-
ber of patent applications for environmental technologies by type of
technology, following the WIPO environmental patent classi fication.
Fig. 1 also shows the GDP growth rate. During the 1990s and 2000s,
the GDP growth rate in Japan stagnated around 0%; it dramatically de-
clined in 1998 and 2009 due to the Asian financial crisis in 1997 and
the global financial crisis triggered by the collapse of Lehman Brothers
in 2008, respectively. Fig. 1 shows that the share of environmental pat-
ent applications, out of total patent applications, gradually increased
from 3% to 9% over the 1990 –2010 period. One interpretation of this
change is that market and social demands for environmental protection
were increased by worsening environmental problems, such as climate
change ( Jin, 2015 ).
Table 1 summarizes both Japanese policies and international events
related to environmental technology invention. Table 1 lists the pollu-
tion control policies focusing on the early 1990s. Environmental man-
agement and climate change mitigation were required beginning in
the late 1990s. Since 2000, both climate change mitigation and appro-
priate handling of toxic chemical substances have been subject to strong
international demand ( Ermoliev et al., 2015 ). These policy trends are
reflected in increasing environmental patent application shares. Asseen in Fig. 1 , the number of patent applications for waste management,
including pollution control technologies, increased during the early
1990s. Since the late 1990s, energy conservation and alternative energy
production have increased each year. In addition, the administrative,
regulatory, and design aspects of technology, including emission trading
technologies, increased dramatically after the Kyoto Protocol, which en-
tered into force in 2005.
Fig. 1. Trends in Japanese environmental patent applications by type of technology. Note: The share is calculated as the number of environmental patent appl ications/number of total
patent applications.Source: Patent data are from patent database published by the Institute of Intellectual Property. GDP growth is from World Development Indicators pu blished by the World Bank.Table 1
Policies and international events related to environmental technology inventions.
Year Japanese policies and international events
1990 –1994Business council for sustainable development created (1990)
United Nations conference on environment and development (1992)Law concerning special measures for total emissions of nitrogenoxides and particulate matter reductions in Japan (1992)
Convention on biological diversity (1993)
Basic environmental law in Japan (1993)Framework convention on climate change (1994)The basic environmental plan in Japan (de fined concept of
environmental risk) (1994)
1995 –2000ISO14001 certi ficate started (1996)
Environmental impact assessment law in Japan (1997)Kyoto Protocol adopted (1997)
Home appliance recycling law in Japan (1998)
Law concerning the promotion of measures to cope with globalwarming in Japan (1998)Act on special measure for industrial revitalization (JapaneseBayh –Dole Act) (1999)
2000 –2004Cartagena Protocol on biosafety (2000)
Basic law for establishing a recycling-based society in Japan (2000)Act on promoting green purchasing in Japan (2000)
Law concerning special measures against dioxins in Japan (2000)
World summit on sustainable development (2002)Amendment of chemical substances control law in Japan [introducedconcept of environmental risk impact into ecological system] (2003)
2005 –2010Kyoto Protocol entered into force (2005)
Effect on RoHS directive (2006)Amendment of air pollution control law in Japan [emission restric-
tions on VOCs] (2006)
Eco-point system for energy-saving electric products in Japan (2009)Government subsidy program for eco-friendly cars (2009)Carbon emission trading system is started in Tokyo, Japan (2010)294 H. Fujii, S. Managi / Technological Forecasting & Social Change 112 (2016) 293 –302

As mentioned above, decision making for research and development
strategy differs by the type of environmental technology because of dif-
ferences in the costs, expected pro fits, and market competitiveness
of each technology.1Additionally, R&D strategies for environmental
technology differ between private companies and the public sector.
Table 2 shows the breakdown of private and public sector R&D expendi-
tures and numbers of patent applications. Most patent applicationswere filed by private companies. However, the public sector did in-
crease its patent applications gradually from 2000 to 2010.
In the private sector, most R&D spending is for technology and prod-
uct development, while in the public sector, approximately 75% of
spending is for basic and applied research. It is clear that private compa-
nies focus on development because they face the threat of bankruptcy if
their products lose competitiveness in the market. Thus, private compa-
nies concentrate their R&D resources on product development because
this directly contributes to corporate pro fits over the short term. Addi-
tionally, the R&D expenditures of private companies are financed by
corporate pro fits. According to the 2014 Survey of Research and Devel-
opment in Japan, 91% of R&D expenditures in private companies come
from self- financed budgets. Meanwhile, 97% of R&D expenditures in
the public sector come from central and local government budgets.
This is one reason for the low share of R&D expenditures allocated to
basic research, which contributes to corporate pro fits only over the
long term. In other words, investing pro fits in basic research technology
is quite risky for private firms. Recently, several types of fund raising in-
struments have been applied by private companies in Japan. For exam-
ple, the Toyota Motor Corporation (TMC) created a unique fundraising
system to support its R&D activities.
2
On the contrary, the R&D expenditures of public research institutes
are mainly financed by government budget allocations. Much of the re-
search budget from the government is directed toward increasing the
international technological competitiveness of Japanese basic research
or applied research over the long term, thus promoting social welfare
rather than short-term pro fits. Therefore, R&D strategies differ between
the private and public sectors because of differences in the objectives
andfinancing of R&D activities.
As discussed above, many previous studies have focused on environ-
mental technologies. However, most of them analyzed patent data for
Europe and the U.S.; few studies have used Japanese patent data. Addi-
tionally, most previous studies have not discussed the characteristics ofthe inventors, even though R&D strategies clearly differ in the private
and public sectors. Based on this background, the objective of this
study is to clarify the determinants of environmental patents among
Japanese manufacturing firms and public sector actors. To consider the
characteristics of each environmental technology, we divided the data
into seven environmental patent groups, following the WIPO
3:( 1 )a l t e r –
native energy production, (2) transportation, (3) energy conservation,(4) waste management, (5) agriculture/forestry, (6) administrative,
regulatory or design aspects, and (7) nuclear power generation. This
study analyzes why the share of Japanese environmental patent appli-
cations increased during Japan's lost decades by focusing on the R&D
strategies of private companies and public research institutes and con-
sidering the characteristics of each environmental technology.
The remainder of this paper is organized as follows. Section 2 de-
scribes the decomposition analysis methodology. Section 3 describes
the dataset. The results of the decomposition analysis using Japanese
environmental patent data are discussed in Section 4 ,a n d Section 5
concludes.
2. Methodology
We apply a decomposition analysis framework to identify the factors
driving environmental patent applications in Japan. To decompose
patent applications related to environmental technology, we use
four indicators: [1] the priority of a speci fic environmental technology
(PRIORITY), [2] the share of all patent applications focusing on environ-mental technology (ENVSHARE), [3] the ef ficiency of R&D expenditures
in patents (EFFICIENCY), and [4] the scale of R&D activities (SCALE).
We de fine the PRIORITY indicator as the number of speci fice n v i r o n –
mental technology patent applications divided by the total number of
environmental patent applications, yielding the share of patent applica-
tions for speci fic technologies. The value of the indicator increases if the
number of speci fic environmental technology patent applications in-
creases more rapidly than the total number of environmental patent ap-
plications, thus indicating that inventors have concentrated their
research resources on a speci fic environmental technology. We can
infer that inventors assign higher priority to speci fic environmental
technologies over other types of environmental technologies if
PRIORITY increases.
4
Similarly, the ENVSHARE indicator is de fined as the total number of
environmental patent applications divided by total number of patent
applications, yielding the share of total environmental patent applica-
tions. The value of this indicator increases if the number of total envi-
ronmental patent applications increases more rapidly than the
number of total patent applications, thus indicating that inventors
have concentrated their research resources on environmental technolo-
gies. Inventors assign higher priority to inventing environmental tech-
nologies over other types if ENVSHARE increases.
EFFICIENCY indicates the ef ficiency of patent generation based on
R&D expenditures. During the R&D process, expenditures are consid-
ered the input and the number of patents is treated as the output.
Thus, the number of patents produced by R&D expenditures re flects
the ef ficiency of the expenditures. This ef ficiency clearly differs by tech-
nological classi fication (e.g., technologies that require expensive mate-
rials for experiments yield few patents for a given amount of R&D
spending). This study tries to capture the ef ficiency of R&D expenditures
and the composition of patent technologies by EFFICIENCY.
Finally, the SCALE indicator is de fined by R&D expenditures and thus
represents the scale of R&D activities. Generally, the scale of R&D activ-ities is strongly related to the number and budget of each research pro-
ject. Thus, total R&D expenditures re flect the level of active R&D efforts.
Additionally, the R&D activities of companies strongly depend on the
corporate financial situation because patent applications are associated
with their R&D investment. For example, the number of patent applica-
tions declined after the global financial crisis triggered by the collapse of
Lehman Brothers ( Organisation for Economic Co-operation and1Some companies employ corporate strategies focusing on their secret core technology
and know-how. Fujii et al. (2015) focus on the relation between corporate productivity
changes and corporate R&D strategies using productivity analysis and questionnaire data.
They concluded that companies employing con fidential information systems tend to in-
crease their productivity. Meanwhile, con fidential corporate information about R&D strat-
egy is dif ficult to collect for manufacturing firms. Thus, analysis of the con fidential
information strategy effect is not a main focus of this study.
2According to TOYOTA Motor Corporation (2015) ,“Toyota Motor Corporation ( “TMC ”)
announced today the pricing of a public offering in Japan of 47,100,000 new shares of First
Series Model AA Class Shares of TMC (the “Model AA Class Shares ”), at the offer price of
10,598 yen per share. TMC expects to receive net proceeds of approximately 475.0 billionyen from the offering. TMC intends to use the proceeds for research and development ofnext-generation innovation, including the development of fuel battery vehicles, researchon infrastructure and development of computerized and sophisticated intelligence mobil-ity technology. ”
3The WIPO environmental patent classi fication is popular in patent data analysis. For
example, Albino et al. (2014) applied it to identify patents related to low-carbon energy
technology. They use 131,661 patent items for low-carbon energy technology from 1971
to 2010 in the U.S. Durán-Romero and Urraca-Ruiz (2015) examined 50,087 patent data
items related to climate change mitigation from 1978 to 2010 in Europe using the WIPOpatent classi fication.4The number of patent applications for speci fic environmental technologies increases
due to disruptive innovation. In this case, the PRIORITY indicator is also affected by the ef-fect of disruptive innovation. Therefore, a dramatic change in the PRIORITY indicator im-
plies either a corporate priority change due to disruptive innovation or a policy and
market change (e.g. new subsidy system, new business market). Thus, we believe our re-search framework can identify the possibility of disruptive innovation by consideringchanges in the PRIORITY indicator.295 H. Fujii, S. Managi / Technological Forecasting & Social Change 112 (2016) 293 –302

Development, OECD, 2009 ). Thus, companies facing financial dif ficulties
reduced their R&D activities to reduce their bankruptcy risk. This de-
crease in R&D activities followed a decline in R&D expenditures. There-
fore, the scale of R&D activity is an important factor explaining why the
number of environmental patent applications changes. The value of
SCALE can increase if R&D expenditures increase. The number of patent
applications for a speci fic environmental technology increases with the
scaling up of overall R&D activities re flected in an increase in SCALE.
Here, we introduce the decomposition approach using waste man-
agement patents as an example. The number of waste management
patent applications (WASTE) is decomposed using total environmental
patent applications (ENVIRONMENT), total patent applications
(TOTAL), and R&D expenditures (R&D) in Eq. (1).
WASTE ¼WASTE
ENVIRONMENTAL/C2ENVIRONMENTAL
TOTAL/C2TOTAL
R&D/C2R&D
¼PRIORITY /C2ENVSHARE /C2EFFICIENCY /C2SCALEð1Ț
We consider the change in waste management patent applications
from year t−1 (WASTEt-1)t oy e a r t(WASTEt). Using Eq. (1),t h e
growth ratio of waste management patent applications can be repre-
sented as follows:
WASTEt
WASTEt−1¼PRIORITYt
PRIORITYt−1/C2ENVSHAREt
ENVSHAREt−1/C2EFFICIENCYt
EFFICIENCYt−1
/C2SCALEt
SCALEt−1: ð2Ț
A natural logarithmic transformation of Eq. (2)yields Eq. (3).5
lnWASTEt−lnWASTEt−1¼lnPRIORITYt
PRIORITYt−1 !
țlnENVSHAREt
ENVSHAREt−1 !
țlnEFFICIENCYt
EFFICIENCYt−1 !
țlnSCALEt
SCALEt−1 !
ð3ȚMultiplying both sides of Eq. (3)byΩit=(WASTEt-WASTEt-1)/
(lnWASTEt-l nW A S T Et-1) yields Eq. (4).6
WASTEt−WASTEt−1¼⊿WASTEt;t−1
¼ωt
ilnPRIORITYt
PRIORITYt−1 !
țωt
ilnENVSHAREt
ENVSHAREt−1 !
țωt
ilnEFFICIENCYt
EFFICIENCYt−1 !
țωt
ilnSCALEt
SCALEt−1 !
ð4Ț
Therefore, changes in the number of patent applications in waste
management ( ⊿WASTE) are decomposed into changes in PRIORITY
(first term), ENVSHARE (second term), EFFICIENCY (third term), and
SCALE (fourth term). The term Ωitoperates as an additive weight for the
estimated number of patent applications for waste management technol-
ogies. This decomposition technique, the logarithmic mean Divisia index
(LMDI) approach, was developed by Ang et al. (1998) .Ang (2004) noted
that LMDI is the preferred method for decomposition analysis because of
its theoretical foundation, adaptab ility, ease of use and result interpreta-
tion, and lack of the residual terms generated by Laspeyres-type
methods. The LMDI approach is widely applied in environmental science
to address issues such as climate change ( de Freitas and Kaneko, 2011 )
and toxic chemical management ( Fujii and Managi, 2013 ).
The novel contribution of this research is to clarify R&D strategies
using LMDI analysis and patent application data. Many previous studies
have focused only on the number of patent applications, which is affect-
ed by the priority, ef ficiency, and scale of R&D activities. Thus, this study
clarifies the contributions of each determinant to changes in environ-
mental patent applications using decomposition analysis. Fujii (2016)
first applied a decomposition framework to patent data using two fac-
tors: priority and scale. In this study, we extend this approach to clarify
the determinants of patent applications using four factors: the priority
of a speci fic environmental technology, the importance of environmen-
tal technology, the ef ficiency of patents based on R&D expenditures, and
the scale of R&D activities.7
5Zero values in the dataset cause problems in the decomposition because of the prop-
erties of logarithmic functions. To solve this problem, the LMDI literature suggests replac-ing the zero values with a small positive number ( Ang and Liu, 2007 ).6Ωit= 0 if WASTEt=W A S T Et−1(Fujii et al., 2013 ).
7A limitation of our research framework is the dif ficulty of completely understanding
the effects of policies and international events on patent invention activity. In order tocompletely understand the causal relationship between priority changes and policy, wewould need to interview the corporate R&D managers of many companies. This approach
requires substantial time and effort. In the meantime, our research framework has the ad-
vantage of being cost-effective, which means that corporate priority changes can be clar-ified by using a decomposition framework and published patent application data, which
are freely available (e.g. IIP patent database).Table 2Breakdown of patent application and R&D expenditures of private company and public sector.Source: Survey of Research and Development (http://www.stat.go.jp/english/data/kagaku/index.htm).
YearPatent application R&D expenditure Breakdown of R&D expenditure
Company Public sector Company Public sectorPrivate company Public sector
Basic research Applied research Development Basic research Applied research Development
2001 99% 1% 76% 24% 6% 20% 74% 42% 33% 25%
2002 99% 1% 75% 25% 6% 20% 75% 43% 33% 24%2003 98% 2% 76% 24% 6% 19% 75% 43% 34% 23%2004 98% 2% 76% 24% 6% 19% 75% 41% 35% 25%2005 98% 2% 77% 23% 6% 20% 74% 41% 34% 25%2006 98% 2% 78% 22% 7% 19% 75% 40% 35% 26%2007 98% 2% 79% 21% 6% 20% 74% 41% 35% 25%2008 97% 3% 78% 22% 6% 20% 74% 40% 35% 25%
2009 98% 2% 75% 25% 7% 21% 73% 41% 35% 24%
2010 97% 3% 76% 24% 7% 19% 74% 40% 35% 25%
Note: To compare the patent application between company and public sector, the share of patent application is estimated by using number of applicatio n of company and public sector as
denominator. In this sense, we did not count the individual patent application for denominator.The Institute of Intellectual Property patent database.296 H. Fujii, S. Managi / Technological Forecasting & Social Change 112 (2016) 293 –302

3. Data
We used patent application data from the database published by the
Institute of Intellectual Property (IIP) ( Goto and Motohashi, 2007 ). The
IIP Patent database covers 12,706,640 patent applications filed from
1964 to 2012.8We speci fied the environmental technology patents fol-
lowing the green inventory patent classi fication published by the WIPO.
As explained above, this study focuses on seven environmental technol-
ogies: (1) alternative energy production, (2) transportation, (3) energy
conservation, (4) waste management, (5) agriculture/forestry, (6) ad-
ministrative, regulatory or design aspects, and (7) nuclear power gener-
ation. Table 3 provides detailed patent classi fication of green inventory
technologies.
Additionally, we use R&D expenditure data from 2001 to 2010 for
both the private and public sectors. The R&D expenditure dataset is ob-
tained from the Survey of Research and Development published by the
Statistics Bureau of Japan.9In this survey, R&D expenditure data are
available for private companies, non-pro fit research institutes, and uni-
versities. R&D expenditures include research salaries, material costs for
experiments, expenses for physical fixed assets, and lease fees. We com-
bine the non-pro fit research institute and university data into public
sector data. R&D expenditure data are de flated to 2010 prices using an
R&D de flator that is available from the Indicators of Science and Tech-
nology published by Japan's Ministry of Education, Culture, Sports, Sci-
ence and Technology. To maintain consistent data coverage for patent
applications and R&D expenditures, we construct the patent application
dataset for the private and public sectors following the grouping meth-
odology applied in the Survey of Research and Development.
According to Fujii (2016) , there are advantages and disadvantages to
using patent data to analyze R&D strategies.10The disadvantage is the
difficulty of evaluating the quality of patent applications. Generally, pat-
ent applications that are granted have higher private values than those
that are withdrawn or refused. However, patent application data do not
control for the quality of technologies. The patent-granted count meth-
od is mainly employed to examine the diffusion of technologies
(e.g., Popp (2006) ). However, patent-granted data do not include infor-
mation about unsuccessful patent applications even though these re-
flect the R&D strategy.
Meanwhile, an advantage of patent application data is that they
cover all R&D activities re flected in inventors' strategies ( Fujii, 2016 ).
In addition, an application fee is required to file the patent application.
Thus, inventors are likely to be con fident that their invention will pass
the examination process if they submit a patent application. Therefore,
we believe that patent application data are more accurate measures of
inventors' R&D activities and strategies than are data on granted pat-
ents. Focusing on these points, Cavalheiro et al. (2014) and Fujii
(2016) used patent application data to analyze R&D strategies. This
study therefore uses patent application data to represent inventors'
R&D strategies with respect to environmental technologies.
This study applied the International Patent Classi fication (IPC) green
inventory classi fication introduced by the WIPO to divide each patentapplication into seven technology groups. Some patent applications
are registered to multiple applicants. To avoid double counting patent
application data, we use only the primary applicant's information
to construct the dataset. Following this procedure, we construct an en-
vironmental patent application dataset by type of technology and
inventor.
Table 4 shows the share of environmental patent applications by in-
dustry and sector. We apply the industrial classi fication approach intro-
duced by the Japan Exchange Group.11We select four industries, which
represents a large share of environmental patent applications. The pub-
lic sector is divided into three organizations whose shares equal 100%. In
Table 4 , government includes local governments and Japanese minis-
tries. The trends in patent applications are re flected in the industry
and public sectors' characteristics presented in Table 4 .
The four industries in Table 4 represent 57.5% of total environmental
patent applications in Japan from 2001 to 2010. The electric products in-
dustry share is approximately 30%. The electric products industry also
represents a large share of alternative energy production, energy con-
servation, and nuclear power generation, which are directly related to
energy. Additionally, the electric products industry represents 34.6% of
administrative, regulatory or design aspects of technology due to the in-
crease in patent applications related to emission trading technology
after 2005.
Next, the transportation equipment industry represents more than
60% of environmental patent applications in the transportation technol-
ogyfield. Patent applications related to hybrid vehicles, fuel cell vehi-
cles, and fuel economy performance improvements represent a large
share of this group. The general machinery industry represents a rela-
tively high share of the waste management technology group because
a law concerning special measures against dioxins entered into force
in Japan in 2000, spurring companies in the general machinery industry8The patent database was constructed using standardized data from the Japan Patent
Office through its 25th release.
9The sample coverage of the Survey of Research and Development is introduced by the
Statistics Bureau of Japan as follows. “The survey covered approx. 13,400 business enter-
prises, approx. 1,100 non-pro fit institutions and public organizations, and approx. 3,700
universities and colleges, for a total of approx. 18,200. The response ratio was approx.87%. (approx. 83% of the business enterprises, approx. 99% of the non-pro fit institutions
and public organizations, and approx. 100% of the universities and colleges) ”.h t t p : / /
www.stat.go.jp/english/data/kagaku/1530.htm.
10Another approach is using scienti fic publications to analyze R&D strategies and activ-
ities. However, it is dif ficult to use scienti fic publication data for this study because there is
no detailed classi fication of environmental technologies in scienti fic publications. There-
fore, a keyword search method must be employed to obtain the data. A keyword search
method may include scienti fic journals that are not directly related to environmental tech-
nologies. Thus, obtaining scienti fic publication count data that correctly re flects inventors'
R&D strategies is dif ficult.Table 3
Description of environmental patent group.
Patent group Patent subgroup
Alternative energy
production(1) Bio-fuels, (2) integrated gasi fication combined
cycle fuel cells, (3) pyrolysis or gasi fication of
biomass, (4) harnessing energy from manmadewaste, (5) hydro energy, (6) ocean thermal energy
conversion, (7) wind energy, (8) solar energy,
(9) geothermal energy, (10) other production or useof heat, not derived from combustion, e.g., naturalheat, (11) using waste heat, (12) devices for produc-ing mechanical power from muscle energy
Transportation(1) Vehicles in general, (2) vehicles other than rail
vehicles, (3) rail vehicles, (4) marine vessel propulsion,(5) cosmonautic vehicles using solar energy
Energy conservation(1) Storage of electrical energy, (2) power supply
circuitry, (3) measurement of electricityconsumption, (4) storage of thermal energy, (5) lowenergy lighting, (6) thermal building insulation, ingeneral, (7) recovering mechanical energy
Waste management(1) Waste disposal, (2) treatment of waste,
(3) consuming waste by combustion, (4) reuse ofwaste materials, (5) pollution control
Agriculture/forestry(1) Forestry techniques, (2) alternative irrigation
techniques, (3) pesticide alternatives, (4) soilimprovement
Administrative, regulatory
or design aspects(1) Commuting, e.g., HOV, teleworking,
(2) carbon/emission trading, e.g., pollution credits,(3) static structure design
Nuclear power generation(1) Nuclear engineering, (2) gas turbine power
plants using heat source of nuclear origin
Source: IPC green inventory launched by WIPO (http://www.wipo.int/classi fications/ipc/
en/est).
11The industrial classi fication for listed companies produced by the Japan Exchange
Group is a popular approach. Most exchanges in Japan, including the Tokyo stock ex-change and Osaka stock exchange, accept this approach.297 H. Fujii, S. Managi / Technological Forecasting & Social Change 112 (2016) 293 –302

to invent new technologies for high temperature incineration plants. Fi-
nally, the chemical products industry's share of 23.9% of environmental
patent applications is related to agriculture and forestry. Most applica-
tions are related to chemical fertilizers and antiseptic dusting powders,
which increase agricultural productivity due to effective harvesting and
forest resource management.
Next, we consider the public sector data. Table 4 indicates that public
research institutes represented more than half of the environmental
patent applications filed by public sector actors. Speci fically, public re-
search institutes filed 87.3% of patent applications related to nuclear
power generation, which requires specialized knowledge and safety
equipment for experiments. Most nuclear power generation technolo-
gies are invented by the Japan Atomic Energy Agency, which has suf fi-
cient expert researchers and specialized equipment.
Surprisingly, the government filed 11.8% and 17.9% of patent appli-
cations related to waste management and agriculture/forest technolo-
gies, respectively. The majority of patent applications are for pollution
control in waste management and for organic fertilizers derived from
waste products and pesticide alternatives in agriculture and forest tech-
nology. The government invented these environmental technologies
because in Japan, the management of garbage dumps, forest resources,
and agricultural businesses is strongly dependent on local governments.
Therefore, governmental staff have incentives to invent technologies to
improve their management performance.
4. Results
Figs. 2 and 3 andTables 5 and 6 show the accumulated changes in
environmental patent applications calculated using the LMDI model.
Figs. 2 and 3 show the results of the decomposition analysis for total en-
vironmental patent applications by private companies and public sectoractors from 2001 to 2010. The scores in Figs. 2 and 3 represent the accu-
mulated changes in total environmental patent application ratios based
on a 2001 baseline. To conduct the decomposition analysis for the
change in total environmental patent applications, the priority term is
dropped from the LMDI model.12The plotted line shows the estimated
accumulated patent application change ratios, and the bar chart shows
the accumulated effects of each determinant on patent applications.
The sum of the bars is equivalent to the line.
Tables 5 and 6 show the results of the decomposition analysis for
changes in environmental patent applications by type of technology
from 2001 to 2010. In Tables 5 and 6 , positive scores indicate increases
in patent applications, while negative scores imply patent application
decreases from 2001 to 2010. Using the results presented in Tables 5
and 6 , a comparative analysis by type of technology is possible.
Here, we discuss the results for private companies, focusing on Fig. 2
andTable 5 .Fig. 2 represents the increase in total environmental patent
applications from 2002 to 2006 due to the increasing share of environ-
mental patents relative to total patent applications and the scaling up of
R&D activities. Meanwhile, the ef ficiency of patent applications contrib-
uted negatively to environmental patent application growth. We can
provide two reasons for this negative effect of ef ficiency among private
companies. First, technological inventions usually become more dif ficult
over time. This is because patent applications require novel products.
Therefore, the number of available technology items for which patent
applications can be filed decreases each year.
Fig. 2. Results of decomposition analysis for private companies' environmental patent applications.
12To conduct a decomposition analysis for changes in total environmental technology
patent applications, the left-hand side of Eq. ( 4)i sE N V S H A R Et−ENVSHAREt−1.I nt h i s
case, we do not use speci fic patent applications, such as waste management technologies.
Therefore, the priority term cannot be calculated in a decomposition analysis for total en-vironmental technology patent applications.Table 4Share of environmental patent applications by industry and public sector from 2001 to 2010.
Private company Public sector
Electric
productsTransportation
equipmentGeneral
machineryChemical
productsUniversitiesResearch
institutesGovernment
All environmental patents 29.8% 17.5% 6.7% 3.5% 39.0% 54.9% 6.1%
Alternative energy production 25.9% 22.2% 7.2% 4.5% 42.9% 51.7% 5.4%
Transportation 6.6% 60.5% 5.3% 0.1% 39.8% 59.0% 1.2%
Energy conservation 47.0% 8.5% 1.5% 3.3% 43.0% 54.2% 2.8%Waste management 10.1% 19.3% 16.7% 4.7% 32.1% 56.1% 11.8%Agriculture/forestry 1.7% 1.2% 6.5% 23.9% 37.3% 44.8% 17.9%Administrative, regulatory or design aspects 34.6% 14.9% 2.0% 0.4% 39.2% 57.4% 3.4%Nuclear power generation 49.5% 1.3% 15.6% 0.1% 12.7% 87.3% 0.0%298 H. Fujii, S. Managi / Technological Forecasting & Social Change 112 (2016) 293 –302

Second, the costs of and investments in R&D activities increase when
newer technologies are used in further investment. Additionally, the in-
vention of technologies using existing or old technologies becomes
more dif ficult each year. However, the latest technologies and rare ma-
terials, such as electron microscopes and rare metals, allow us to find
new structures and relationships at extremely minute scales and ex-
plore new chemical reactions. In such cases, the equipment and mate-
rials needed for experiments require large R&D expenditures. The
efficiency factor is de fined by the number of total patent applications
for a given amount of R&D spending; thus, this expensive R&D produces
an indicator that worsens each year, as shown in Fig. 2 .
According to the 2015 Japan Patent Of fice Annual Report, total pat-
ent applications in Japan decreased by 10.8% (42,232 items) from
2008 to 2009. This sharp decline is mainly observed in six technology
fields: engineering elements or units, measuring and testing instru-
ments, computing and calculating, general vehicles, electronic commu-
nication techniques, and basic electrical elements. The patent
applications filed in these six technology fields declined by 16,479
items, which represents a decrease in total patent applications of ap-
proximately 40% from 2008 to 2009. The Japan Patent Of fice Annual Re-
port indicates that the main reason for this dramatic decline is the
economic recession triggered by the collapse of Lehman Brothers in
2008. Additionally, R&D expenditures by private companies declined
during this period because corporate financial performance, which is
strongly related to bankruptcy risk, deteriorated. Therefore, we suggest
that the smaller effect of the scale factor after 2008 was due to the finan-
cial crisis.
The decrease in the ef ficiency factor since 2008 is also due to other
crisis-related issues: R&D expenditures, including researcher salaries,
are dif ficult to cut immediately during a financial crisis. Researchers in
private companies are usually employed in permanent positions,which are strongly protected by the Labor Standard Act in Japan. Mean-
while, the procurement of expensive materials and equipment for R&D
activities becomes dif ficult in an uncertain economic situation or during
periods of unstable corporate financial performance. Therefore, private
companies cut material and equipment costs but maintained researcher
salaries. The gap between researchers and the resources they need to in-
novate is another reason for the declining ef ficiency factor after 2008.
Next, we discuss the decomposition analysis results for private com-
panies by type of technology. Table 5 represents the change in cumula-
tive patent applications from 2001 to 2010 by type of environmental
technology. From Table 5 , all technology types exhibit similar trends:
a negative effect of the ef ficiency factor and a relatively small effect of
the scale factor. Meanwhile, the effects of the priority and environmen-
tal patent share factors differ by type of technology.
We identify three trends in the decomposition analysis by focusing
on the priority and environmental patent share factors. The first trend
observed is in the administrative, regulatory or design aspects of tech-
nology field. In this area, patent applications increased mainly due to
the stronger effect of the priority factor. One interpretation of this result
is that new business opportunities in emission trading systems werecreated by the Kyoto Protocol in 2005. The creation of a new and
large-scale market encouraged private companies to seize these oppor-
tunities and to improve their market competitiveness using their R&D
resources. Thus, we can explain the increasing number of patent appli-
cations for administrative, regulatory or design aspects through the pri-
ority factor, which was in turn strongly affected by the Kyoto Protocol.
The second trend is observed in the transportation and energy con-
servation technology fields. In these two technology areas, patent appli-
cations increased mainly due to the growth of the environmental patent
share. The priority factor also contributes to the increase in patent
applications, albeit weakly. Therefore, the contributions of these two
Fig. 3. Results of decomposition analysis for public sector actors' environmental patent applications.
Table 5
Change in private company patent applications by technology type from 2001 to 2010.
Patent application change Decomposed factor contribution
(Item) (%) PRIORITY ENVSHARE EFFICIENCY SCALE
All environmental patents 1930 12% N.A. 65% −58% 5%
Alternative energy production 88 2% −13% 65% −57% 7%
Transportation 191 47% 35% 80% −69% 2%
Energy conservation 1338 31% 18% 67% −59% 4%
Waste management −2222 −49% −59% 39% −37% 8%
Agriculture/forestry −85 −59% −61% 27% −28% 6%
Administrative, regulatory or design aspects 2676 326% 313% 198% −176% −9%
Nuclear power generation −56 −14% −23% 40% −37% 5%299 H. Fujii, S. Managi / Technological Forecasting & Social Change 112 (2016) 293 –302

technologies are different from the first trend introduced above. The
positive trend in patent applications filed for these two fields is similar
to that of environmental technology out of total patent applications.
One interpretation of this trend is that transportation and energy
conservation technologies are strongly related to the market competi-
tiveness of products because the operating costs of transportation
equipment and electrical products depend on these technologies.
Therefore, the incentives for transportation and energy conservation
technology innovation become stronger with growing environmental
technology demand. Additionally, the demand for environmental tech-
nology is widely affected by the cost of energy resources. Because inter-
national oil prices increased dramatically from 2001 to 2010, the
demand for environmental technology increased. Thus, the increased
number of patent applications in these two market competitiveness-re-
lated areas was mainly due to the increase in the environmental tech-
nology share observed from 2001 to 2010.
The third trend is observed in alternative energy production, waste
management, agriculture/forestry, and nuclear power generation tech-
nologies. In these four technology areas, priority negatively affected
the change in patent applications from 2001 to 2010, especially for
waste management and agriculture/forest technologies. The priorityfactor decreased because the market demand for these technologies de-
creased in 2010 relative to 2001. During the 1990s and early 2000s,
waste management became a key issue for balancing economic devel-
opment and environmental protection on Japan's small land area. Addi-
tionally, many environmental policies, such as the basic environmental
law of 1993, the home appliance recycling law of 1998, and the law
concerning special measures against dioxins of 2000 have entered into
force. Therefore, market demand for waste management was high dur-
ing this period.
Meanwhile, climate change and Japanese energy security issues be-
came more serious in the late 2000s. These serious environmental and
resource problems provide strong incentives to invent energy conserva-
tion and climate change mitigation technologies. Thus, one reason for
decreasing patent applications in waste management technology from
2001 to 2010 is that the priority technology in corporate R&D strategies
shifted from waste management to energy conservation and climate
change mitigation due to changes in market demand. This priority
change is observed in the Japanese science and technology basic plan.
In the 2nd (2001 to 2005) and 3rd (2006 to 2010) basic plans, the cre-
ation of a resource-circulating society was a high-priority research
field. However, the 4th basic plan (2011 to 2015) set technology related
to creating a low-carbon society and ef ficient energy use through smart
infrastructure as high priorities. This policy change by the Japanese gov-
ernment helps explain the priority change in corporate R&D strategy
observed from 2001 to 2010.
Next, we discuss the results for the public sector. Fig. 3 represents
the results of a decomposition analysis for the period from 2001 to
2010. From Fig. 3 ,t h ee f ficiency factor strongly contributed to the
increase in environmental patent applications until 2007, which is a dif-
ferent trend from that observed for private companies. The contribution
of efficiency could have been caused by strengthening business-academic collaborations, which were promoted by the Japanese
Bayh –Dole Act in 1999 and the National University Corporation Act
in 2003.
Rules governing patent ownership for innovations developed using
government research funding were changed by the Japanese Bayh –
Dole Act in 1999. After this law entered into force, inventors who used
these research funds for R&D activities could obtain ownership of the
resulting patents. Therefore, researchers using government funds had
strong incentives to invent new technologies and obtain patent owner-
ship ( Kato and Odagiri, 2012 ). Thus, this change in patent ownership
could have increased the ef ficiency factor.
A second reason for the increase in public sector ef ficiency is a
change in the R&D strategies of national universities due to the National
University Corporation Act passed in 2003. This Act reclassi fied national
universities as independent administrative entities, which use an inde-
pendent accounting system. Before this Act, universities focused mainly
on student education and academic publication, especially of basic re-
search. However, after the Act entered into force, universities were re-
quired to increase their market competitiveness in both education and
technology development to secure funding. This R&D strategy change
among universities encouraged them to proactively form research col-laborations with private companies to publish academic papers and
fi
le patents ( Motohashi and Muramatsu, 2012 ). However, the ef ficiency
factor decreased after 2007. We suggest that the main reason is the
global economic recession, as business-academic collaboration is dif fi-
cult during periods of high financial risk for private companies.
Next, we discuss the change in the environmental technology share
for the public sector. During the early 2000s, the share of environmental
patent applications decreased. The application shares of several tech-
nology fields, such as chemicals, medicine, semiconductors, image com-
munication, and electric digital data processing technologies, increased.
However, patent applications related to the emission trading system in-
creased rapidly from 41 items in 2005 to 95 items in 2006, which in-
creased the contribution of the share factor.
Fig. 3 indicates that the main change in environmental patent in-
creases in the public sector is due to the ef ficiency factor rather than
the share factor. This finding is detectable only by conducting separate
decomposition models for the private and public sectors. The small con-
tribution of the share factor and the high contribution of the ef ficiency
factor are similar in the results according to type of technology de-
scribed in Table 6 .
As seen in Table 6 , the priority factor strongly contributes to the in-
crease in patent applications for alternative energy production filed by
the public sector. Alternative energy technologies include many basic
research fields, such as solar panel materials and biofuel enzymes. As
we noted, public sector actors have a comparative advantage in basic re-
search fields. Therefore, public sector actors apply for patents to gain
market competitiveness in this field. The government has increased re-
search budgets in the renewable energy field, and R&D strategies are
strongly dependent on the research budget allocations of the govern-
ment. From 2001 to 2010, the Japanese government increased research
budgets for renewable energy to achieve a low-carbon society. Thus, weTable 6
Change in public sector patent applications by technology type from 2001 to 2010.
Patent application change Decomposed factor contribution
(Item) (%) PRIORITY ENVSHARE EFFICIENCY SCALE
All environmental patents 96 23% N.A. −2% 20% 5%
Alternative energy production 121 92% 65% 1% 21% 6%Transportation −5 −45% −47% −7% 4% 4%
Energy conservation 32 41% 22% 2% 13% 4%
Waste management −37 −33% −66% −4% 33% 4%
Agriculture/forestry 7 140% 89% 1% 47% 2%Administrative, regulatory or design aspects −8 −12% −13% −6% 2% 4%
Nuclear power generation −14 −58% −79% −10% 27% 3%300 H. Fujii, S. Managi / Technological Forecasting & Social Change 112 (2016) 293 –302

consider these two reasons for the increased role of the priority factor in
alternative energy production in the public sector.
Another finding in Table 6 is the contribution of the priority factor
for administrative, regulatory or design aspects to the dramatic increase
in private sector patent applications in this area. Meanwhile, the com-
position of patent applications changed from 2001 to 2010. Before
2005, the main technology area was commuting technologies, whilepatent applications for the emission trading system increased from
2005 to 2006. After 2006, public sector patent applications related to ad-
ministrative, regulatory or design aspects gradually decreased each
year. Therefore, enforcement of the Kyoto Protocol affected the R&D
strategy of the public sector, increasing patent applications related to
emission trading over the short term. However, this effect did not per-
sist over the long term, as seen in the private sector results.
5. Conclusions
This study analyzes the determinants of environmental patent appli-
cations in the private and public sectors in Japan. We focus on the differ-
ent characteristics of each environmental technology type. We apply
the LMDI approach to patent application changes from 2001 to 2010
for seven technology groups. Our main findings are summarized as
follows.
First, the determinants of environmental patent applications differ
between the private and public sectors. Private companies increased en-
vironmental patent applications mainly due to their share of environ-
mental patent applications. Meanwhile, the public sector increased
applications by improving the number of patent applications filed for
given R&D expenditures, that is, it improved the ef ficiency of R&D activ-
ities. This ef ficiency improvement in the public sector may have been
caused by the Japanese Bayh –Dole Act and the strengthening of
business-academic collaborations in Japan in the 2000s.
Second, the priority of environmental patents differs in each tech-
nology group. Patent applications in the administrative, regulatory or
design area increased more than threefold from 2001 to 2010 amongprivate companies. This dramatic increase could have been related to
new business opportunities created by the introduction of an emission
trading system by the Kyoto Protocol in 2005. Because private compa-
nies fund research through their corporate pro fits, new business oppor-
tunities provide strong incentives to invent new technologies.
Therefore, a market designed to create environmental business is im-
portant to enhancing private creation of environmental technologies.
Third, the public sector did not react strongly to new business oppor-
tunities produced by the Kyoto Protocol. These findings indicate that
private companies may be more sensitive than the public sector to mar-
ket demands for technologies, as public sector R&D strategy depends on
the research budget allocated by the government, which is not intended
to maximize pro fits. In other words, R&D activities in the public sector
focus on the technology areas in which private companies struggle
due to high corporate risk and low pro fitability. Therefore, the govern-
ment needs to consider the characteristics of environmental technology
during the research budget planning process, especially how these tech-
nologies are related to demand in the market over the short term or to
demand in society over the long term.
Finally, we consider why the environmental patent application share
increased during the lost decades in Japan. The above findings of the de-
composition analysis of environmental patent applications suggest
three main reasons. The first reason is changes made to rules governing
patents. The main changes were produced by the Japanese Bayh –Dole
Act in 1999 and the National University Corporation Act in 2003.
These two acts provide strong incentives for research institutes, which
accept research funding from the government, to file patent applica-
tions in order to survive in the market during economic depressions.
A second reason is the growth of market demand for environmental
technologies, especially for climate change mitigation, due to interna-
tional environmental policies. These policies create new businessmarkets in which private companies gain competitiveness if they devel-
op higher quality patented products. Thus, we argue that the growth of
the share of environmental patents from 2005 to 2006 was caused by
private companies' quick response to the enforcement of the Kyoto Pro-
tocol, which made the emission trading system a high priority.
A third reason is that the relative priority of environmental technol-
ogy increased due to the economic recession triggered by the collapse ofLehman Brothers in 2008. During financial crises, companies reduce
their R&D activities. Meanwhile, the Japanese government started the
eco-point system and granted subsidies for eco-friendly cars as part of
its emergency economic measures. These policies provided strong in-
centives for private companies to invent technologies for energy conser-
vation and transportation. Therefore, the share of environmental
patents increased after 2008 due to the re-prioritization of certain tech-
nologies, especially in the energy conservation and transportation tech-
nology fields.
These findings explain why environmental patent applications in-
creased during the economic recession in Japan, which is useful for es-
tablishing effective environmental policies and allocating government
research budgets to achieve sustainable development. Additionally,
the research framework and application of a decomposition model for
patent invention could be helpful for estimating the determinants of
R&D activities and for conducting policy evaluations. We suggest that
policy makers and decision makers within companies' R&D divisions
apply this R&D decomposition research framework to evaluate the ef-
fects of subsidies and policies on environmental technology innovation.
The results imply that policies and subsidies affect the invention of en-
vironmental technologies through the four determinant factors. This in-
formation is helpful in forecasting how policy may affect future
environmental technology inventions.
Further research is needed on the following two points. First, factor
decomposition analysis that considers the time-lag between investment
and patent invention is needed. The time-lag span between environ-
mental inventions and patents is diverse and depends on the character-
istics of patent technology. Therefore, it was dif ficult for our research
framework to clearly consider the time-lag effect, which is an important
factor in understanding patent invention activities.
Second, future research should include a comparative analysis of de-
veloped and developing countries. Such an analysis could clarify the pri-
ority gap in the R&D strategies of each environmental technology type.Based on the determinants of each environmental technology area, we
can suggest effective domestic and international policies to enhance
the development of future environmental technologies for a sustainable
society.
Acknowledgments
This research was funded by the grant-in-aid for Specially Promoted
Research [26000001B]; the Ministry of Education, Culture, Sports,
Science, and Technology (MEXT), Japan; and grant-in-aid for research
activity start-up [26881006B], MEXT, Japan; and research program “Pol-
icy studies of environmental economics ”[J155102J09] by the Japanese
Ministry of the Environment. The results and conclusions of this article
do not necessarily represent the views of the funding agencies.
References
Albino, V., Ardito, L., Dangelico, R.M., Petruzzelli, A.M., 2014. Understanding the develop-
ment trends of low-carbon energy technologies: a patent analysis. Appl. Energy 135,
836–854.
Ang, B.W., Zhang, F.Q., Choi, K.H., 1998. Factorizing changes in energy and environmental
indicators through decomposition. Energy 23 (6), 489 –495.
Ang, B.W., 2004. Decomposition analysis for policymaking in energy: which is the pre-
ferred method? Energ Policy 32 (9), 1131 –1139.
Ang, B.W., Liu, N., 2007. Handling zero values in the logarithmic mean Divisia index de-
composition approach. Energ Policy 35 (1), 238 –246.
Cavalheiro, M.C., Joia, L.A., Gonçalves, A.C., 2014. Strategic patenting in the upstream oil
and gas industry: assessing the impact of the pre-salt discovery on patent applica-
tions in Brazil. World Patent Inf. 39, 58 –68.301 H. Fujii, S. Managi / Technological Forecasting & Social Change 112 (2016) 293 –302

de Freitas, L.C., Kaneko, S., 2011. Decomposition of CO 2emissions change from energy
consumption in Brazil: challenges and policy implications. Energy Policy 39 (3),
1495 –1504.
Durán-Romero, G., Urraca-Ruiz, A., 2015. Climate change and eco-innovation: a patent
data assessment of environmentally sound technologies. Innov. ManagPolicy Practice
17 (1), 115 –138.
Ermoliev, Y., Ermolieva, T., Jonas, M., Obersteiner, M., Wagner, F., Winiwarter, W., 2015.
Integrated model for robust emission trading under uncertainties: cost-effectiveness and environmental safety. Technol. Forecast. Soc. Chang. 98, 234 –244.
Fujii, H., 2016. Decomposition analysis of green chemical technology inventions from
1971 to 2010 in Japan. J. Clean. Prod. 112 (5), 4835 –4843.
Fujii, H., Managi, S., 2013. Decomposition of toxic chemical substance management in
three U.S. manufacturing sectors from 1991 to 2008. J. Ind. Ecol. 17 (3), 461 –471.
Fujii, H., Managi, S., Kaneko, S., 2013. Decomposition analysis of air pollution abatement in
China: empirical study for ten industrial sectors from 1998 to 2009. J. Clean. Prod. 59
(15), 22 –31.
Fujii, H., Edamura, K., Sumikura, K., Furusawa, Y., Fukuzawa, N., Managi, S., 2015. How en-
terprise strategies are related to innovation and productivity change: an empiricalstudy of Japanese manufacturing firms. Econ. Innov. New Technol. 24 (3), 248 –262.
Goto, A., Motohashi, K., 2007. Construction of a Japanese Patent Database and a first look
at Japanese patenting activities. Res. Policy 36 (9), 1431 –1442.
Hamada, K., Okada, Y., 2009. Monetary and international factors behind Japan's lost de-
cade. J. Jpn. Int. Econ. 23 (2), 200 –219.
Jin, W., 2016. International technology diffusion, multilateral R&D coordination, and glob-
al climate mitigation. Technol. Forecast. Soc. Chang. 102, 357 –372.
Kato, M., Odagiri, H., 2012. Development of university life-science programs and
university –industry joint research in Japan. Res. Policy 41 (5), 939 –952.
Lindman, A., Söderholm, P., 2015. Wind energy and green economy in Europe: measuring
policy-induced innovation using patent data. Appl. Energy. http://dx.doi.org/10.1016/
j.apenergy.2015.10.128 .Lise, J., Sudo, N., Suzuki, M., Yamada, K., Yamada, T., 2014. Wage, income and consumption
inequality in Japan, 1981 –2008: from boom to lost decades. Rev. Econ. Dyn. 17 (4),
582–612.
Motohashi, K., Muramatsu, S., 2012. Examining the university industry collaboration pol-
icy in Japan: patent analysis. Technol. Soc. 34 (2), 149 –162.
Organisation for Economic Co-operation and Development (OECD), 2009C. Patents in
Environment-related Technologies. in: OECD Science, Technology and Industry
Scoreboard 2009. OECD Publishing, Paris. http://dx.doi.org/10.1787/sti_scoreboard-
2009-18-en .
Organisation for Economic Co-operation and Development (OECD), 2014C. Measuring
Environmental Innovation Using Patent Data: Policy Relevance. OECD Publishing,
Paris.
Popp, D., 2006. International innovation and diffusion of air pollution control technolo-
gies: the effects of NOX and SO2 regulation in the US, Japan, and Germany.
J. Environ. Econ. Manag. 51 (1), 46 –71.
Sun, Y., Lu, Y., Wang, T., Ma, H., He, G., 2008. Pattern of patent-based environmental tech-
nology innovation in China. Technol. Forecast. Soc. Chang. 75 (7), 1032 –1042.
Motor Corporation, T.O.Y.O.T.A., 2015. TMC to Commence Steps to Issue Model AA Class
Shares for Medium to Long Term Investors. http://newsroom.toyota.co.jp/en/detail/
7780146 .
Hidemichi Fujii is an Associate Professor at the Graduate School of Fisheries and Environ-
mental Sciences, Nagasaki University, Japan.
Shunsuke Managi is a Distinguished Professor of Technology and Policy & Director of Ur-
ban Institute at the Kyushu University, Japan.302 H. Fujii, S. Managi / Technological Forecasting & Social Change 112 (2016) 293 –302