OECD Studies On Environmental Innovation Environmental Policy Technological Innovation and Patents OECD Studies On Environmental Innovation

-:HSTCQE=UY[]V]:
ISBN 978-92-64-04681-8
97 2008 06 1 P
OECD Studies on Environmental
Innovation
Environmental Policy,
Technological Innovation
and Patents
OECD Studies on Environmental Innovation
Environmental Policy, Technological
Innovation and Patents
Technological innovation can help realise environmental objectives in a less costly
manner than would otherwise be the case. Thus, understanding the role that
technological innovation can play in achieving environmental objectives is important for
policy debates.
However, the relationship between environmental policy and technological innovation
remains an area in which empirical evidence is scant. In an attempt to bridge this gap,
the OECD has examined the relevant issues, using patent activity as a measure of
technological innovation.
Three case studies have been undertaken: abatement technologies for wastewater
efuent from pulp production, abatement of motor vehicle emissions, and development
of renewable energy technologies. On the basis of patent data, the nature, extent, and
causes of innovation in each of these areas have been explored. While a particular focus
has been placed on the role of environmental policy in bringing about the innovation
documented, it is recognised that other factors play a key role in inducing innovation
that has positive environmental implications.
The full text of this book is available on line via these links:
www.sourceoecd.org/environment/9789264046818
www.sourceoecd.org/scienceIT/9789264046818
Those with access to all OECD books on line should use this link:
www.sourceoecd.org/9789264046818
SourceOECD is the OECD online library of books, periodicals and statistical databases.
For more information about this award-winning service and free trials, ask your librarian, or write to
us at SourceOECD@oecd.org.
www.oecd.org/publishing
O
E
C
D

S
t
u
d
i
e
s

o
n

E
n
v
i
r
o
n
m
e
n
t
a
l

I
n
n
o
v
a
t
i
o
n

E
n
v
i
r
o
n
m
e
n
t
a
l

P
o
l
i
c
y
,

T
e
c
h
n
o
l
o
g
i
c
a
l

I
n
n
o
v
a
t
i
o
n

a
n
d

P
a
t
e
n
t
s
and Patents
ORGANISATION FOR ECONOMIC CO-OPERATION
AND DEVELOPMENT
The OECD is a unique forum where the governments of 30 democracies work
together to address the economic, social and environmental challenges of globalisation.
The OECD is also at the forefront of efforts to understand and to help governments
respond to new developments and concerns, such as corporate governance, the
information economy and the challenges of an ageing population. The Organisation
provides a setting where governments can compare policy experiences, seek answers to
common problems, identify good practice and work to co-ordinate domestic and
international policies.
The OECD member countries are: Australia, Austria, Belgium, Canada, the
Czech Republic, Denmark, Finland, France, Germany, Greece, Hungary, Iceland,
Ireland, Italy, Japan, Korea, Luxembourg, Mexico, the Netherlands, New Zealand,
Norway, Poland, Portugal, the Slovak Republic, Spain, Sweden, Switzerland, Turkey,
the United Kingdom and the United States. The Commission of the European
Communities takes part in the work of the OECD.
OECD Publishing disseminates widely the results of the Organisations statistics
gathering and research on economic, social and environmental issues, as well as the
conventions, guidelines and standards agreed by its members.
Also available in French under the title:
tudes de lOCDE pour linnovation environnementale
Politique environnementale, innovation technologique et dpts de brevets
Corrigenda to OECD publications may be found on line at: www.oecd.org/publishing/corrigenda.
OECD 2008
Photo credit: David Wasserman/Brand X/Corbis
You can copy, download or print OECD content for your own use, and you can include excerpts from OECD publications,
databases and multimedia products in your own documents, presentations, blogs, websites and teaching materials,
provided that suitable acknowledgment of OECD as source and copyright owner is given. All requests for public or
commercial use and translation rights should be submitted to rights@oecd.org. Requests for permission to photocopy
portions of this material for public or commercial use shall be addressed directly to the Copyright Clearance Center
(CCC) at info@copyright.com or the Centre franais d'exploitation du droit de copie (CFC) contact@cfcopies.com.
This work is published on the responsibility of the Secretary-General of
the OECD. The opinions expressed and arguments employed herein do not
necessarily reflect the official views of the Organisation or of the governments
of its member countries.
Frontmatter Page 2 Thursday, October 23, 2008 10:31 AM
FOREWORD
ENVIRONMENTAL POLICY, TECHNOLOGICAL INNOVATION AND PATENTS ISBN 978-92-64-04681-8 OECD 2008
3
Foreword
Understanding the role that technological change can play in achieving environmental
objectives is important since innovations can allow for improved environmental quality at
lower cost. However, the relationship between environmental policy and technological
innovation remains an area in which empirical evidence is scant. In an attempt to
bridge this gap, the OECD has examined these issues.
Three different case studies have been undertaken: abatement technologies for
wastewater effluent from pulp production; abatement of motor vehicle emissions; and
development of renewable energy technologies. While particular focus has been placed
on the role of environmental policy in bringing about the innovation documented, it is
recognised that other factors play a key role in inducing innovation which has positive
environmental implications.
The work was overseen by the delegates to the Working Party on National
Environmental Policies of the OECD, who provided valuable comments and inputs at
all stages of the project. In addition, the work has been presented at a number of
international conferences, and comments received have served to improve the study
significantly.
The authors would also like to thank Dominique Guellec and Hlne Dernis of the
OECD Directorate for Science, Technology and Industry for their foresight and hard
work in developing the OECDs patent database upon which much of this study is
based. And finally, Claire-Line Martin provided excellent support in the preparation of
the manuscript, which was very much appreciated.
TABLE OF CONTENTS
5
Table of Contents
List of Acronyms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
Executive Summary. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
Chapter 1. Environmental Policy, Technological InnovatIon
and Patent Activity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
2. The economics of innovation and eco-innovation . . . . . . . . . . . . . . . . 20
3. Measures of innovation and eco-innovation . . . . . . . . . . . . . . . . . . . . . 24
4. The policy determinants of eco-innovation. . . . . . . . . . . . . . . . . . . . . . 38
5. The case studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44
Notes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
Annex 1.A1. Sources of Patent Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52
Annex 1.A2. Patent Classification Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . 57
Annex 1.A3. Number of EPO Applications in Different
Environmental Areas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59
Chapter 2. Environmental Regulation and International Innovation
in Automotive Emissions Control Technologies. . . . . . . . . . . . . . 63
1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
2. Environmental regulation in the automobile sector. . . . . . . . . . . . . . . 65
3. Innovation in automotive emissions control technologies . . . . . . . . . 75
4. Empirical analysis and results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90
5. Conclusions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
Notes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95
Annex 2.A1. Overview of Technologies and Corresponding Patent Classes . . 97
TABLE OF CONTENTS
6
Chapter 3. Policy Versus Consumer Pressure: Innovation and Diffusion
of Alternative Bleaching Technologies in the Pulp Industry. . 107
1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108
2. The pulp and paper industry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 109
3. Pollution and the pulp and paper industry . . . . . . . . . . . . . . . . . . . . . . 114
4. Regulatory responses. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118
5. Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121
6. Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 122
7. Conclusions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 131
Notes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 132
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 134
Annex 3.A1. Relevant Patent Classes for Pulp Bleaching Technologies. . . . 138
Chapter 4. Renewable Energy Policies and Technological Innovation:
Energy Source and Instrument Choice . . . . . . . . . . . . . . . . . . . . 139
1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 140
2. The renewable energy sector: trends, technologies and policies . . . . 140
3. Patent applications for renewable energy . . . . . . . . . . . . . . . . . . . . . . . 144
4. Empirical analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 153
5. Conclusions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 161
Notes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 162
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 162
Annex 4.A1. First Page of Sample Patent Application . . . . . . . . . . . . . . . . . . 164
Chapter 5. Policy Conclusions and Further Work . . . . . . . . . . . . . . . . . . . . . 167
1. Policy conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 168
2. Further work . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 170
Notes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 172
Annex 5.A1. Glossary of Relevant Patent and Related Terms . . . . . . . . . . . . 173
List of tables
1.1. Comparison of the patent systems of the United States, Japan
and Europe (circa 2000) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
1.2. IPC patent classification system for solar concentrating devices
used for the generation of mechanical power . . . . . . . . . . . . . . . . . . 36
1.3. Characteristics of the case studies. . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
2.1. Technologies covered. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77
2.2. Fixed-effects model estimates for engine re-design technologies . . . 92
2.3. Fixed-effects model estimates for post-combustion devices. . . . . . 92
TABLE OF CONTENTS
7
3.1. Pulp producers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110
3.2. Percentage of exports to each country: paper and paperboard. . . . 111
3.3. Summary of Ecolabel programs related to pulp and paper
manufacturing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117
3.4. Summary of key regulations. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121
3.5. Number of domestic chlorine and non-chlorine patents,
selected years . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 124
3.6. Top domestic patent assignees . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 126
4.1. Share of electricity production from renewable sources
(excluding hydro) (%) by country . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 143
4.2. Examples of policies aimed at supporting renewable energy . . . . . 145
4.3. IPC classifications for renewable energy. . . . . . . . . . . . . . . . . . . . . . . 147
4.4. Number of EPO patent filings in renewable energy technologies
(Annual average 1978-2003, by inventor country) . . . . . . . . . . . . . . . 149
4.5. Number of EPO patent filings in renewable energy technologies
(Annual average 1978-2003, per unit of GDP, by inventor country) . . . 150
4.6. Number of EPO patent applications in renewable energy
technologies, normalised by overall patenting activity
(1978-2003) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 151
4.7. Descriptive statistics of explanatory variables (1978-2003) . . . . . . . 155
4.8. Estimated coefficients of the negative binomial fixed effects
models with individual policy variables . . . . . . . . . . . . . . . . . . . . . . . 156
4.9. Correlation coefficients between policy variables . . . . . . . . . . . . . . . 157
models with a composite policy variable . . . . . . . . . . . . . . . . . . . . . . 158
models with clusters of policy variables . . . . . . . . . . . . . . . . . . . . . . . 160
List of figures
1.1. Share of environmental R&D in total government R&D, 1981-2005 . . 26
1.2. Proportion of facilities by country with budgets
for environment-related R&D. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
1.3. Proportion of facilities by employee size class with budgets
for environment-related R&D. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
1.4. Share of new products in turnover . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
1.5. Number of TPF patent applications by inventor country . . . . . . . . . 34
1.6. Number of EPO Environmental patent applications
and total EPO patent applications . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
1.7. Number of EPO Environmental patent applications . . . . . . . . . . . 37
1.A3.1a. Number of EPO Applications in Different Environmental Areas. . 60
TABLE OF CONTENTS
8
1.A3.1b. Number of EPO Applications in Different Environmental Areas (suite) 61
2.1. Evolution of US HC and NO
X
standards for passenger cars
(gasoline) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66
2.2. Evolution of US CO Standards for passenger cars (gasoline) . . . . . . 67
2.3. Evolution of Japanese CO, HC and NO
X
(gasoline) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68
2.4. Evolution of Japanese CO, HC, NO
X
and PM standards
for passenger cars (diesel) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68
2.5. Evolution of European CO standards for passenger cars
(gasoline and diesel) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69
2.6. Evolution of European HC and HC + NO
X
standards
for passenger cars (gasoline) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
2.7. Evolution of European HC + NO
X
(gasoline and diesel) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
2.8. Evolution of European HC and NO
X
(gasoline) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71
2.9. Evolution of European PM standards for passenger cars (diesel) . . 71
2.10. Evolution of patent applications at the USPTO, 1975-2001. . . . . . . . 78
2.11. Evolution of patent applications at the GPTO, 1975-2001. . . . . . . . . 79
2.12. Evolution of patent applications at the JPO, 1975-2001) . . . . . . . . . . 80
2.13. Evolution of patent applications at the EPO, 1975-2001 . . . . . . . . . . 82
2.14. Technology shares within engine re-design group at different
patent offices (1975-2001 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82
2.15. Source countries for patents (1975-2001) . . . . . . . . . . . . . . . . . . . . . . 83
2.16. Average patent family size by country and year . . . . . . . . . . . . . . . . 84
2.17. Source countries of USPTO engine re-design and post-combustion
patents. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86
2.18. Source countries of German engine re-design and post-combustion
patents. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87
2.19. Source countries of JPO engine re-design and post-combustion
patents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88
2.20. JPO engine re-design and post-combustion patents by domestic
inventors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89
3.1. Domestic ECF and TCF patents by country. . . . . . . . . . . . . . . . . . . . . 123
3.2. Average patent family size by country and year . . . . . . . . . . . . . . . . 127
3.3. ECF and TCF patent trends, selected countries . . . . . . . . . . . . . . . . . 128
3.4. Diffusion of ECF bleaching technologies . . . . . . . . . . . . . . . . . . . . . . . 129
3.5. Diffusion of ECF and TCF bleaching technologies . . . . . . . . . . . . . . . 130
4.1. Annual growth rates for renewable energy in the world
and the OECD (1990-2004) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 141
4.2. Renewable energy sources in the OECD in 2004 . . . . . . . . . . . . . . . . 142
TABLE OF CONTENTS
9
4.3. Percentage of energy to be provided by renewable energy . . . . . . . 144
4.4. Introduction of policies by type for renewable energy
in OECD countries (1973-2003) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 146
4.5. Number of EPO patent applications by type of renewable . . . . . . . . 148
4.6. Number of EPO patent technologies for renewables by country. . . 148
4.7. EPO patent filings in renewable energy technologies
(annual mean, per unit of GDP) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 151
4.8. Relationship between point of introduction of policies
and patent counts. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 152
4.9. Dendogram of policy variable clustering . . . . . . . . . . . . . . . . . . . . . . 159
LIST OF ACRONYMS
11
List of Acronyms
ADt Air-dried Tonne
AOX Adsorbable Organic Halide
BAT Best Available Technology
BOD Biological Oxygen Demand
CAAA Clean Air Act Amendments
CAF Corporate Average Fuel Economy
CARB California Air Resources Board
CAT Catalytic Converters and Catalytic Regeneration Technology
CIS Community Innovation Survey
ClO2 Chlorine Dioxide
CO Carbon Monoxide
COD Chemical Oxygen Demand
CVCC Compound Vortex Controlled Combustion
DSTI Directorate for Science, Technology and Industry (OECD)
ECF Elemental Chlorine-Free
ECLA European Patent Classification System
EGR Exhaust Gas Recirculation
EPO European Patent Office
GBAORD Government Budget Appropriations and Outlays for R&D
GERD Gross Domestic Expenditures on R&D
GPTO German Patent and Trademark Office
H2O2 Hydrogen Peroxide
HC Hydrocarbons
IPR Intellectual Property Rights
JPO Japanese Patent Office
NORSCAN US, Canada, Sweden, Finland and Norway
NOX Nitrogen Oxides
NSF National Science Foundation
O3 Ozone
OBD On-Board Diagnostics
PAH Polycyclic Aromatic Hydrocarbons
Pb Lead
PCT Patent Cooperation Treaty
PM Particulate Matter
LIST OF ACRONYMS
12
R&D Research and Development
SO2 Sulphur Dioxide
TCF Totally Chlorine-Free
TPES Total Primary Electricity Supply
TPF Triadic Patent Family
TRIPS Trade-Related Aspects of Intellectual Property Rights
TSS Total Suspended Solids
USPTO United States Patent and Trademark Office
WIPO World Intellectual Property Organization
ISBN 978-92-64-04681-8
Environmental Policy, Technological Innovation and Patents
OECD 2008
13
Executive Summary
Technological innovation can help realise environmental objectives in a
less costly manner than would otherwise be the case. Thus, understanding
the role that technological innovation can play in achieving environmental
objectives is important for policy debates. However, the relationship between
environmental policy and technological innovation is an area in which empirical
evidence remains limited. In an attempt to bridge this gap, the OECD has
examined these issues, using patent activity as a proxy for technological
innovation. This work has assessed the role that environmental policies have
played in inducing eco-innovation.
Three different case studies have been undertaken: abatement technologies
for wastewater effluent from pulp production; abatement of motor vehicle
emissions; and development of renewable energy technologies. These cases
were selected in order to ensure variation in the issues addressed (product vs.
process innovation; integrated abatement technologies vs. end-of-pipe
technologies; degree of tradability of the technologies). While there is no ideal
measure of innovation, patent data have been widely used to assess the
effects of policy and other factors on technological innovation in general. As
such, in this report patent data is used to assess the nature, extent, and causes
of innovation specifically in the environmental context.
While particular focus has been placed on the role of environmental
policy in bringing about the innovation documented, it is recognised that other
factors play a key role in inducing innovation which has positive environmental
implications. The factors which drive innovation in general are also likely to
encourage eco-innovation specifically. For instance, factors such as macro-
economic stability, functioning of capital markets, the degree of economic
openness, and the quality of education systems affect innovation rates in
general, but also the specific case of eco-innovation by extension.
There are, however, some distinct concerns which arise with respect to
eco-innovation. Most importantly, there are two externalities involved in the
eco-innovation case: the positive externalities associated with information
spillovers resulting from the innovation process; and, the negative externality
associated with the environmental impacts. In the case of knowledge
spillovers those who are responsible for innovation bear the full costs, without
receiving all the benefits. In the absence of policy interventions, the rate of
EXECUTIVE SUMMARY
14
innovation will therefore be sub-optimal, and thus an economy will be less
competitive and productive over time.
In the case of environmental externalities, those who are responsible for
emitting pollution receive the full benefits from doing so, but without paying
the full costs. In effect the price of polluting is too low. And as with the
general theory of induced innovation, this will provide incentives for innovation
which uses the under-priced factor input intensively. Therefore, in the absence
of policy interventions to internalise the externality, innovation will bend in a
direction which is relatively more pollution-intensive than would otherwise
be the case.
Environmental and innovation externalities are usually the responsibilities
of different Government Ministries. Policy coordination is, therefore, key.
However, innovation and environmental policy have different objectives. While
the former is largely concerned with internalising knowledge spillovers (and
thus increasing competitiveness and productivity), the latter is concerned
with addressing negative environmental externalities. While no single
instrument is likely to be able to address both market failures, co-ordination
between the two is vital, if both the rate and the direction of innovation are to be
optimal.
As such, this volume focuses on the role of environmental policy in
bending innovation in a more environmentally-friendly direction. Several
interesting conclusions have emerged from the three case studies that were
undertaken:
Environmental policy does have an effect on technological innovation. For
instance, in the study on renewable energy, the implementation of different
policy measures had a measurable impact on innovation, with tax measures
and quota obligations being statistically significant determinants of patent
activity. However, the effect of the different policies varied by the type of
renewable energy involved.
General scientific capacity matters. Again in the case study on renewable
energy innovation, the variable reflecting expenditures on targeted R&D
was statistically significant in every model estimated.
Relative prices induce particular kinds of innovation. In the case of motor
vehicle emissions abatement, fuel prices encouraged investment in
integrated innovation (in which fuel efficiency gains also arose), but not in
post-combustion technologies. In the case of renewable energy, the role of
electricity prices was rarely significant, except for solar energy. However, as
fossil fuel prices rise (and renewables become more competitive), the price
substitution effect is likely to become more important.
Other market factors can also be important spurs to innovation. In the case
of bleaching technologies in the pulping process, public concerns about the
EXECUTIVE SUMMARY
15
environment appeared to spur the development of ECF and TCF technologies,
pre-dating the introduction of regulatory standards. Interestingly, eco-labelling
did not appear to have an influence on innovation in this case.
The type of innovation changes through time. In the renewables sector,
different energy sources have reached maturity at different points, and
there have been different generations of innovation within particular
renewable energy sources. In the case of motor vehicle emissions abatement,
there has been a shift from post-combustion technologies to integrated
technologies.
International diffusion of environmental innovation is common. In the case
of both bleaching technologies and motor vehicle emissions, abatement
patent families (for some countries) are large, reflecting significant technology
transfer. In the case of motor vehicle emissions abatement, the transfer of
Japanese technologies to the US is striking.
In other areas, there is some evidence of a first-mover advantage. In the pulp
and paper sector, the early policy interventions introduced by Finland and
Sweden resulted in a strong comparative advantage in TCF technologies.
While not addressed directly in this project, two more general conclusions
emerge from the literature. First, investing in R&D is risky. As such it is important
that the environmental policy framework not add to this risk, but rather provide
investors with a stable horizon in which to undertake research investments. If
markets have difficulty efficiently dealing with commercial risk associated
with innovative activity, they will be even less likely to deal efficiently with the
uncertainties associated with unstable policy conditions. Second, the policy
framework should allow for a variety of technological options to be adopted.
Governments have limited resources at their disposal, as well as limited
information about optimal technological trajectories. Moreover, with the
potential for lock in, it is important to develop policies which minimise the
downside risks of picking losers. In general, this means targeting
environmental policies as close as possible to the environmental objective itself,
not some proxy.
The volume concludes that patent data is a helpful means by which to
examine eco-innovation and suggestions for future policy research are proposed:
a) development of robust indicators of eco-innovation across a wider spectrum
of environmental areas (e.g.green chemistry, abatement of air pollution,
carbon capture and mitigation, building energy efficiency, waste prevention,
etc.); b) assessment of the environmental and economic returns on patented
eco-innovation inventions in selected areas; and c) examination of the links
between environmental policy, economic globalisation and eco-innovation,
with a focus on international technology transfer.
ISBN 978-92-64-04681-8
OECD 2008
17
Chapter 1
Technological InnovatIon and Patent Activity
by
Nick Johnstone and Ivan Hascic (OECD Environment Directorate)
and Katrin Ostertag (Fraunhofer Institute)*
Technological innovation can allow for the realisation of
environmental objectives in a manner which is less costly than would
otherwise be the case. As such, understanding the role that
technological innovation can play in achieving environmental
objectives is important for policy debates. This chapter provides a
review of the theory and evidence with respect to the role that
environmental policies can play in inducing innovation, and
provides an introduction to case studies undertaken. On the basis
of patent data, the nature, extent, and causes of innovation with
respect to renewable energy, wastewater treatment and motor vehicle
emissions control were explored. While particular focus has been
placed on the role of public policy in bringing about the innovation
documented, it is recognised that other factors play a key role in
inducing innovation which has positive environmental implications.
* Comments from Davis Popp and Frans de Vries gratefully acknowledged.
1. ENVIRONMENTAL POLICY, TECHNOLOGICAL INNOVATION AND PATENT ACTIVITY
18
1. Introduction
Technological innovation can help realise environmental objectives in a
less costly manner than would otherwise be the case. For example, some
innovations may drive down the cost of pollution abatement. This results in
welfare gains. On the one hand, the same environmental quality may be achieved
with fewer factor inputs devoted to abatement. On the other hand, for the same
amount of factor inputs, environmental quality may be improved.
Thus, an improved understanding the role that technological innovation
can play in achieving environmental objectives is important for policy makers.
However, the relationship between environmental policy and technological
innovation remains an area in which empirical evidence is scant.
1
Moreover,
there are almost no empirical studies with international coverage. Part of the
reason for this is the absence of appropriate comparative data to reflect the
innovation process. In this report patent activity is used as a proxy for
Three different case studies have been undertaken: abatement technologies
for wastewater effluent from pulp production; abatement of motor vehicle
emissions; and development of renewable energy technologies. These cases were
selected in order to ensure variation in the issues addressed and the
characteristics of the innovation. Specifically:
autos are a traded commodity, and as such international influences matter.
In addition, autos are a consumer good and thus the objectives of regulations
are to reduce end-use pollution, not pollution from production;
in the case of pulp and paper, unlike autos, we are focusing on pollution at
the source of production. However, the focus is on a process technology,
rather than end-of-pipe solutions. The analysis of innovation with respect
to a process technology represents a new contribution to the literature; and,
renewable energy sources are both a process (generating electricity) and a
product (e.g. the turbines themselves). Electricity is not generally a traded
commodity, which contrasts with autos. However, the renewable equipment
may be traded.
On the basis of patent data, the nature, extent, and causes of innovation
in each of these areas were explored. Particular focus has been placed on the
role of public policy in bringing about the innovation documented. This
includes both policy stringency and the nature of the environmental policy
19
instruments applied. However, it is recognised that other factors play a key
role in inducing innovation which has positive environmental implications.
Factors such as fuel and energy prices, sectoral growth rates, and general
scientific capacity have been controlled for in the more formal analyses
included in the case studies.
Several interesting conclusions have emerged from the three case studies
that were undertaken:
policy measures had a measurable impact on innovation, with tax
measures and quota obligations being statistically significant determinants
of patent activity. However, the effect of the different policies varied by the
type of renewable energy involved.
environment appeared to spur the development of ECF and TCF
technologies, pre-dating the introduction of regulatory standards.
Interestingly, eco-labelling did not appear to have an influence on
innovation in this case.
renewable energy sources. In the case of motor vehicle emissions
abatement, there has been a shift from post-combustion technologies to
integrated technologies.
patent families (for some countries) are large, reflecting significant
technology transfer. In the case of motor vehicle emissions abatement, the
transfer of Japanese technologies to the US is striking.
2
20
In other areas, there is some evidence of a first-mover advantage. In the
pulp and paper sector, the early policy interventions introduced by Finland and
This introductory Chapter begins with a review of the literature related to
the key determinants of innovation, with specific focus on environmental
innovation. It also provides a discussion of alternative measures of innovation
and eco-innovation, before proceeding to a discussion of the impact of
environmental policy on environmental innovation. The Chapter concludes with
a summary of the characteristics of the case studies that were undertaken. The
following Chapters then present the case studies, followed by a concluding
Chapter which summarises the main policy conclusions and proposes areas
for further research.
2. The economics of innovation and eco-innovation
When assessing the determinants of innovation which results in reduced
environmental impacts (hereafter eco-innovation), it is important to first
understand the determinants of innovation in general. The role that general
market and policy factors play in encouraging innovative activities is therefore
crucial to an understanding of the factors that facilitate innovation which is
more specifically environmental in nature. Since the primary interest of this
report is the analysis of the influence of environmental policy instruments on eco-
innovation, the following review of the more general determinants of overall
innovation will be very brief.
2.1. The determinants of innovation in general
There is considerable variation in innovative activity both across OECD
countries and within OECD countries between sectors of the economy (see OECD,
2007a). Researchers have sought to identify the reasons for this variation, and a
number of factors appear to be significant. These can be distinguished as
i) market and firm-level factors; and ii) policy factors.
2.1.1. Market and firm-level factors
Over six decades ago, Schumpeter (1942) argued that there was a positive
correlation between market concentration and innovation. In theory, this is
because a monopoly is in a better position to prevent imitation, and because it
will have more resources with which to finance R&D activities. A monopoly
may therefore be better able to bear the risk (and reap the benefits) associated
with R&D investments. However, in a seminal paper, Arrow (1962) argued that
the efficiency incentives associated with perfect competition are actually
more conducive to innovative activities.
21
Given these conflicting theoretical views concerning the potential role of
market structure on incentives to invest in R&D, it is not surprising to find that
the empirical evidence is also mixed. While Gerosky (1990) found support for
the positive effects of competition, Kraft (1987) and Acs and Audretsch (1987)
found support for the Schumpeterian hypothesis.
3
Therefore, theoretical and
empirical support has been found for both a negative and a positive relationship
between the level of innovation and the degree of market competition.
The spatial scope of the market is also thought to be an important driver
of innovation. In effect, the more global the market in which a firm operates,
the more likely the firm will be to innovate. Criscuolo et al. (2005) found that
markets serve as a conduit of information; as such, the pool of information
available to global firms is greater than that which is available to national and
local firms. For similar reasons, it is sometimes argued that foreign direct
investment can be an important conduit, expanding the potential knowledge
pool upon which the firm can draw. Some recent empirical evidence (Jaumotte
and Pain 2005a) indicates that openness is among the most important
determinants of investment in R&D.
Another important feature of industrial R&D arises out of the frequent
need for a firm to self-finance its R&D investments. The inherent riskiness of
these investments, the skewed nature of potential returns, and the potential
for asymmetric information between prospective borrowers and lenders make
external sources of financing more difficult to secure (Jaumotte and Pain,
2005a). Under such conditions, capital markets may have difficulty assessing
the optimality of potential investments in R&D (Scherer and Harhoff 2000).
Consequently, it is argued that firms with greater internal financial resources
are more likely to invest in R&D. Two explanations have been advanced for this
(see for example, Kamien and Schwartz, 1978):
Outside lending may be difficult to obtain, because a failed R&D project
leaves few valuable tangible assets. Given the risk associated with R&D
projects, external lenders may be reluctant to finance such projects without
substantial collateral.
Firms may be unwilling to reveal private information that could make the
project interesting to outside lenders, because they fear that this information
will then become available to rivals.
Such financing problems are not thought to be as important for firms
quoted on the stock market because these firms have easier access to capital
than other firms (Syrneonidis, 1996). Moreover, the recent deepening of venture
capital markets in some countries/regions (North America, Netherlands, and UK),
has tended to reduce some of the need to rely upon internal finance (OECD,
2006a).
22
Small firms may face particular difficulties in financing their R&D projects.
As noted by Jaffe et al. (2003), these firms have less internally generated cash and/
or less access to financial markets. Firm size may also be important if there are
significant economies of scale associated with R&D. The evidence in this area
is mixed as well (Syrneonidis, 1996). However, it would appear that the effect
of firm size is non-linear, becoming less important once a minimum threshold
is reached. Empirical evidence indicates that this threshold may be as low as
approximately 100 employees (Syrneonidis, 1996).
2.1.2. Public policy factors
According to work by the OECD a number of general policy framework
conditions are also likely to be important drivers of innovation. First, stable
macroeconomic conditions are positively related to innovation. Low and
stable interest rates are particularly important for risky investments, such as
those associated with R&D (OECD, 2006a). Second, firms which are exposed to
international market competition are more likely to be innovative. As such,
more open international trade and foreign investment policies will tend to
result in more innovation. Third, and in a similar vein, when a firms products
operate in markets that are less heavily regulated, innovation tends to be stronger
(Jaumotte and Pain, 2005c).
The effects of more targeted innovation and science policies have also
been assessed. One intrinsic characteristic of the knowledge produced through
R&D is that it is very difficult to exclude others from the benefits that arise from
that R&D. This makes knowledge a public good as such, it will be underprovided
by the market, because the private returns to R&D investments are much lower
than the social returns [for an empirical analysis, see for example Mansfield
et al. (1997)]. In addition, and as noted earlier, even in the absence of the public
good elements of R&D, there can be particular difficulties in financing R&D at
optimal levels. There is, therefore, a need for policy intervention, whether to
increase the returns on innovation or to reduce the cost of its generation. [See
OECD (2004) for a review of current practice in OECD countries.]
OECD governments have provided financial support for R&D for many
years. In recent years, the form in which this support is provided has changed,
with a transition being made from grants to various forms of tax credits
(OECD, 2007b). However, the evidence on the returns to public sector support
for R&D is mixed. While Hall and van Reenen (2000) find positive evidence that
tax incentives encourage investment in R&D, a meta-analysis undertaken by
Garcia-Quevado (2004) did not find systematic evidence in support of the more
general contention that subsidies encourage greater private sector R&D.
The effectiveness of subsidies depends upon the degree of additionality
associated with the subsidies being provided i.e. the extent to which they
23
lead to a higher overall level of R&D expenditure. On the one hand, crowding
out can occur if the funds provided displace private sector expenditures that
would otherwise have been made. Falk (2004) found evidence that government
subsidies have a significant positive effect on business R&D, but only when
previous R&D is controlled for suggesting that the beneficiaries of subsidies
may be mainly firms that have previously undertaken R&D. On the other hand,
crowding in may also occur: if there are capital constraints and economies of
scale in innovation, provision of public support may lead to greater privately-
financed R&D. Interestingly, Lach (2002) found that subsidies to Israeli SMEs
had a significant (and positive) impact on company-financed R&D, but this
was not the case for larger firms.
In order to address the public good nature of the knowledge produced by
R&D, legal protection can be provided to help innovators capture potential
rents. Thus, the strength of a countrys intellectual property rights (IPR) regime
is often thought to be a key driver of R&D. However, recent cross-sectional
empirical work [e.g. Cohen et al. (2000)] generally found that the IPR regime is
a weak predictor of R&D investment in OECD countries. This may be due to the
fact that there is little difference in the stringency of IPR regimes across
countries, so isolating their effect is complicated. Nonetheless, in work
undertaken at the OECD, it was found that a 1% increase in the standard
deviation of the Park index of IPRs resulted in an 8% increase in the number
of patents and 1-1.5% in R&D expenditures (Jaumotte and Pain, 2005a).
In addition, many governments have put in place programmes which
facilitate cooperation between public research institutions and industry. [See
Jaumotte and Pain (2005b) for a discussion.] While many of these programmes
are motivated by a wish to realise greater economic benefits from publiclyfunded
R&D, they may also serve as a spur to private sector R&D. If well-designed, these
programmes can encourage the internalisation of knowledge externalities
between public and private bodies and (most importantly) between different
private bodies.
2.2. The specific case of eco-innovation
The lessons learned concerning innovation in general, and the role that
public policy can play in inducing this innovation are, of course, directly
relevant to a more specific discussion of eco-innovation. For instance, market
structure may be important because some of the sectors where environmental
impacts are potentially most important (e.g. chemicals, pulp and paper,
energy) also have concentrated markets.
There are, however, some distinct concerns which arise with respect to
eco-innovation. Most importantly, there are two externalities involved in the
eco-innovation case: the positive externalities associated with information
24
spillovers resulting from the innovation process and the negative externality
associated with the environmental impacts (see Jaffe et al., 2005). In the case
of knowledge spillovers those who are responsible for innovation bear the full
costs, without receiving all the benefits. In the absence of policy interventions,
the rate of innovation will therefore be sub-optimal, and thus an economy will
be less competitive and productive over time.
In the case of environmental externalities, those who are responsible for
emitting pollution receive the full benefits from doing so, but without paying
the full costs. In effect the price of polluting is too low. And as with the
general theory of induced innovation,
4
this will provide incentives for innovation
which uses the under-priced factor input intensively. Therefore, in the absence
of policy interventions to internalise the externality, innovation will bend in a
direction which is relative more environment-intensive than would otherwise
be the case.
In general, distinct policies need to be implemented to resolve both of
these problems. Such policies have generally been the responsibilities of
different Ministries. This is appropriate, because of the differing underlying
policy objectives. However, in some cases, it can lead to policy incoherence.
For instance, the objectives of environmental policy measures with respect to
eco-innovation may be undermined by innovation policy measures which
support more polluting technologies (see Goel and Hsieh, 2006).
Many OECD governments have made significant efforts to co-ordinate
these two sets of policies (e.g. the European Unions Environmental Technology
Action Plan). On the one hand, the innovation effects of environmental
policies have become an increasingly important criterion in policy assessment
for Ministries of the Environment. On the other hand, the environmental
effects of innovation policies have become an important criterion for policy
assessment in Ministries of Science, Technology and Industry (see Kivimaa
and Mickwitz, 2006 and Kivimaa and Mickwitz, 2004).
Indeed, in the OECDs Science, Technology and Industry Outlook (2004), a
majority of countries cite environment-related concerns in their science and
technology priorities, including: Australia (environmentally sustainable
Australia); Austria (environment, energy and sustainability); France (development
of renewable energy); Germany (clean processes and production technologies);
Hungary (environmental protection); Norway (energy and environment); UK
(sustainable energy) and the US (climate, water and hydrogen).
3. Measures of innovation and eco-innovation
Given the importance of technological innovation in modern economies,
identifying reliable measures of technological innovation has long preoccupied
economists (and continues to do so). The task is further complicated when
25
measures of specific types of innovation are sought such as eco-innovation
even if a reliable measure of innovation can be identified, it may be difficult to
determine whether it is specifically environmental in nature.
3.1. Input and output measures of innovation
Due to the complexity of the innovation process, all potential measures
of innovation (e.g. budgets for research and development, number of scientific
personnel) are at best imperfect indicators of the innovativeness of an
economy. [For annual data on research and development and other relevant
indicators, see OECD Research and Development Statistics (2006c), and Main
Science and Technology Indicators (2006b).]
5
Following the different phases of the R&D process [i.e. basic and applied
research and (experimental) development] through to the different stages of
innovation (from invention to diffusion), a range of different indicators for
measuring technological change can be found in the literature. Three basic
groups can be distinguished: resource (or input) indicators; indicators of R&D
results (i.e. output indicators of the R&D process); and progress indicators
(i.e. output indicators focussing on the economic impacts of innovations)
(Grupp, 1998).
Input indicators follow the rationale that, for technological change to take
place, the necessary resources have to be invested in knowledge acquisition.
Among the most commonly applied input indicators are: i) R&D expenditures
and ii) R&D personnel. Generally, data on R&D personnel is not as widely available
as data on R&D expenditure. However, input indicators may also refer to other
ways of knowledge acquisition, such as investment in R&D intensive goods, or
expenditure for licenses.
The OECD has also sought to disaggregate public R&D data by socio-
economic objective (OECD 2002). (No effort has been made to collect private
sector R&D data disaggregated by socio-economic objective.) In principle, the
allocation of expenditures to specific objectives is determined on the basis of
managerial intentions at the time of commitment of the funds. However,
given the uncertainty associated with general R&D, further disaggregating
public R&D data by socio-economic objective may be even more difficult to
establish with confidence, particularly for more basic forms of research.
Moreover, even if this could be done, drawing boundaries between the different
objectives
6
is by no means straightforward. Bearing these caveats in mind, data
on public sector R&D expenditures for control and care of the environment is
presented in Figure 1.1. Based on this figure, many OECD countries have
clearly been increasing their investment in environmental research and
development (R&D), in an effort to boost technological developments that
improve environmental quality.
7
26
At the micro level, a survey of 4 000 manufacturing facilities in seven
countries undertaken by the OECD Environment Directorate in 2003 provided
data on those facilities which reported having undertaken environment-
related R&D.
8
In the sample, 58.7% of facilities reported that they had
incurred R&D expenditures, and 9.3% of facilities reported having invested in
environment-related R&D. Figure 1.2 shows the proportion of facilities
reporting that they had environment-related R&D expenditures (by country).
Figure 1.1. Share of environmental R&D in total government R&D, 1981-2005
Source: OECD (2005).
Figure 1.2. Proportion of facilities by country with budgets
for environment-related R&D
Source: Johnstone (2007).
%
6
5
4
3
2
1
0
1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2001 2003 2005
Korea Germany France United Kingdom Japan United States
0.18
0.14
0.10
0.06
0.00
0.16
0.12
0.08
0.04
0.02
Canada
Mean +/ 1 std. error
France Germany Hungary Japan Norway United States Total
27
Norway had the highest percentage, with just under 15% reporting having do
ne so, while for Germany, it is only 3.6%. For four of the seven countries
(France, Canada, Japan and the US), the proportion was approximately 10%.
Figure 1.3 shows the proportion of facilities with environment-related
R&D budgets (according to facility size). As can be seen clearly, larger facilities
(more than 500 employees) are more likely to have undertaken such investments
(with almost 20% reporting having done so), while this figure is less than 10% for
facilities with less than 250 employees. In the sample as a whole, the average
size of facilities responding affirmatively had over 720 employees; for those who
responded negatively, the relevant figure was less than 300.
The data appears to be consistent with results obtained from other
surveys. For instance, according to statistics on research and development
from the US National Science Foundation,
9
55.3% of US companies with more
than 5 employees in the manufacturing sector reported R&D expenditures
for 2001. In the OECD sample, 51% of US manufacturing facilities (with more
than 50 employees) are engaged in R&D. Among those, 16% reported having a
R&D budget specifically for environmental matters.
In the OECD survey, the sectors with the highest percentage of facilities
reporting having undertaken investments in environment-related R&D were:
coke, petroleum and refining (13.6%); chemicals and chemical products
(12.8%); paper and paper products (12.5%); and motor vehicles (12.4%).
Figure 1.3. Proportion of facilities by employee size class
with budgets for environment-related R&D
Source: Johnstone (2007).
0.20
0.15
0.10
0.05
0.00
< 100 100-249 250-499 > 500
0.25
Mean +/ 1 std. error
Number of employees
28
Although input indicators (and R&D expenditures in particular) reflect an
important element of the innovation system, there are a number of
disadvantages associated with their use as an indicator of innovation. For
example, and as noted earlier with respect to private R&D expenditures, the data
are incomplete. Further, the data are only available at an aggregate level and
cannot be broken down by technology group. In addition, efforts to unbundle
specific environmental R&D from other types of R&D can be difficult,
particularly as efforts to improve environmental performance become more
integrated with general business strategies (see Johnstone and Labonne, 2006).
However, the greatest shortcoming associated with the use of R&D data is that
it measures inputs to the innovation process, rather than successful outputs.
Given the uncertainty associated with the innovation process, the economic
value of equivalent R&D expenditures can also vary widely.
As such, an output measure of innovation is broadly preferable. There
are two principal candidates for indicators of the output of R&D: bibliometric
data (scientific publications) and technometric data (patent publications). The
use of bibliometric data as a measure of innovation has been given renewed
impetus with the growth of the internet, combined with increasingly efficient
search engines. Using keywords and indexing codes, searches of relevant
databases (e.g. the Science Citation Expanded Index) are typically undertaken.
Data on author, affiliation, date of publication, etc. can be extracted, and
counts developed to assess relative innovative capacity (see Meyer, 2002).
This kind of indicator is particularly useful as for analysing the diffusion
of knowledge among inventors (and between countries), based on
co-publications and citations. However, there are some shortcomings associated
with the use of bibliometric data. In particular, while such data is indeed an
output indicator of innovation, it is only an indicator of an intermediate output.
Publication in a peer-reviewed journal reflects a scientific advance, but not
necessarily one which has commercial applications. It is difficult to use
citations as an index of quality, let alone economic importance.
For these reasons, patent data are more commonly used as output
indicators of innovation. Compared to bibliometric data, patents are more
closely linked to applied research and experimental development. They are
therefore closer to the commercial stage of innovation. In addition, for reasons
explained below, patents lend themselves to cross-country comparisons of
innovation performance in specific technology fields.
Downstream output-based measures of innovation necessitate the
identification of specific innovations which have been developed and
commercialised. Data of this kind is usually collected through surveys of firms;
and self-reporting of the introduction of novel goods/technologies or successful
inventions. However, since the answers provided in these surveys are highly
29
subjective, they are often difficult to interpret. Most commonly, a companys
turnover with new products is used.
For instance, the European Community Innovation Survey (CIS) requests
that respondents indicate the value share of novel products in total turnover.
10
While it is difficult to assess the reliability of responses to such questions, they do
give an indication of self-perceived innovation. Unfortunately, the CIS did not
previously ask respondents to provide information of specific relevance to
environmental concerns. Figure 1.4 provides the aggregate data from the third
CIS, which reported on responses from the 1998-2000 survey.
3.2. Patents as a measure of innovation
For this report, patents have been used as a key measure of innovation.
11
With the exception of the European Patent Office, patents are granted by
national patent offices (usually specialised agencies) in individual countries.
They give the holder the right to exclude others from the production of a
specific good (or from the use of a specific process) for a defined number of
years. This period may vary, depending on the nature of the innovation. In
order to be eligible for a patent, the innovation must be novel, involve a non-
obvious inventive step, and be useful (see Dernis and Guellec, 2001). Definitions of
novelty differed in the past among countries. Today generally, the concept of
universal novelty is implemented in national patent legislation. It requires that
no publication of any sort () may occur anywhere in the world prior to the filing
date of the invention (Adams, 2006).
Figure 1.4. Share of new products in turnover
Source: Eurostat, Third Community Innovation Survey.
25
20
15
10
5
0
Share in total turnover (%) 1998-2000
I
c
e
l
a
n
d
N
o
r
w
a
y
L
u
x
e
m
b
o
u
r
g
G
r
e
e
c
e
F
r
a
n
c
e
A
u
s
t
r
i
a
D
e
n
m
a
r
k
B
e
l
g
i
u
m
P
o
r
t
u
g
a
l
I
t
a
l
y
S
p
a
i
n
F
i
n
l
a
n
d
G
e
r
m
a
n
y
30
Patent protection is only valid in the country that grants the patent.
Inventors who desire patent protection in other nations must file separate
applications in those nations, either directly or indirectly [e.g. through a
regional patent office such as the European Patent Office (EPO) or the World
Intellectual Property Organisation (WIPO)]. Although the broad principles are
the same, there are differences in administrative procedures and nature of
protection granted between patent offices. Some of the main characteristics
are presented in Table 1.1.
12
Table 1.1. Comparison of the patent systems of the United States, Japan
and Europe (circa 2000)
Prosecution Features United States Japan Europe
Basis of deciding patent
priority
First to invent First to file First to file
Native language filing
permitted?
Any language (English
translation for record)
Japanese or English Any EPO member language,
with English, French or German
translation; translations required
for each country designated
Patent term 20 years from filing 20 years from filing 20 years from filing
Publication of patent Yes, 18 months from filing Yes, 18 months from filing Yes, 18 months from filing
Examination deferral No Yes, 7 years from filing Yes, 6 months from publication
Third-party contestation
of patent
Post-grant opposition Post-grant oppositions Post-grant oppositions
Patentability standards Least unfavourable
to applicants
Strict standards-claims need
working examples
Moderate standards; some
variation by nationality, limited
grace period
Breadth of claim awarded Very Broad Narrow Broad
Overall comparable costs
(1993)
$13.000 $30.000 $120.000
Strength of enforcement Very strong Weak Medium (varies) enforcement
by national patent offices
Interpretation of claims Literal interpretation,
pro patentee court systems,
equivalents applied
Narrow interpretation linking
back to specifications,
equivalents not applied
Interpreted less literally;
equivalents applied, but under
different principles
Judicial style Adversarial court trials,
jury trials often seen
Brief intermittent hearing
with judge, written testimony
is widely used
Largely by written testimony,
some oral hearings, court
experts used
Uncertainty in outcomes High, due to lay judges
(in District Courts) and juries.
No in UK
High, especially due to delays,
and pressure to settle out of
court
Relatively low in all jurisdictions
Time taken (1993) 18-24 months 36 months + opposition
(3-5 years)
30 months + opposition
(expeditious)
Source: Adapted from Somaya (2000) and Gallini (2002).
31
While differences remain, there has been considerable convergence in
the characteristics of intellectual property rights regimes over time. Indeed,
applying the Park Index
13
of the strength of IPRs from1980 to 2000, Jaumotte
and Pain (2005c) found that there is relatively little variation across OECD
countries. Recent data suggest that the coefficient of variation in the index for
OECD countries fell from 0.32 in 1990, to 0.07 in 2005 (Park and Lippoldt 2007).
EPO patent applicants may designate as many of the 32 EPO member-states
for protection as desired. The application is first examined and if successful
granted by the EPO. The patent is then transferred to the individual national
patent offices designated for protection. Because EPO applications are more
expensive, European inventors typically first file a patent application in their
home country, and then apply to the EPO if they desire protection in other
European countries. However, by filing with EPO, the total costs will be much less
than if individual applications are made to each country (Popp, 2005).
The procedure through the WIPO is based on the Patent Co-operation
Treaty (PCT). In this case, any of the more than 140 signatory nations can be
designated for patent protection. The WIPO issues a preliminary examination
report, but does not actually grant the patent. This is the responsibility of the
national patent office, to which the application is subsequently transferred.
Applications abroad must be filed within one year of the priority date,
which is the year in which the initial application was filed. If the inventor does
file abroad within one year, the inventor will have priority over any similar
patent applications received in those countries since the priority date. This
legal concept of priority was first introduced in the Paris Convention in 1883.
Additional fees apply for each application. Because of these features of patent
law, only the more valuable inventions are filed in several countries. Moreover,
filing a patent application in a given country is a signal that the inventor
expects the invention to be profitable in that country.
Patents are typically sorted by the priority year. If a patent is granted,
protection begins from the priority date. It corresponds quite closely to the
date when the inventive activity took place, as patent applications are usually
filed early in the innovation process. Initially, patents were only published
when granted. However, due to long delays of such publications and the
resulting disadvantages (e.g. duplication of R&D), starting in the 1960s, most
patent offices adopted the deferred examination process. This includes the
requirement that an application be published while it is still pending
typically around 18 months after the earliest filing date (Adams, 2006). In
empirical research, it is therefore important to distinguish between data on
patent applications and data on patents granted.
32
Patent data provide a rich data source for empirical studies. In particular,
they have the following strengths:
There are very few examples of major inventions which have not been
patented (Dernis and Guellec, 2001).
Compared to R&D expenditure, patents are an output indicator (i.e. they
measure productive innovation activity).
Further, the data classified into detailed technologies. A distinction can,
therefore, be made between, for example, air pollution control devices
designed to reduce NO
X
emissions and devices designed to control SO
2
emissions (see e.g. Popp, 2005).
Patent data is generally readily available (although not always in a
convenient format). They are based on objective standards that change
slowly on the basis of invention novelty and utility (e.g. Griliches, 1990).
Patent data are also discrete, and thus easily subject to statistical analysis.
Moreover, given the objective standard of assessing patents, these statistics
provide some uniformity in comparing innovation across countries.
And finally, patent data have the potential to be linked to sectoral data
(see e.g. EPO and OECD work on concordance between ISIC and IPC codes)
and to firm level data.
14
It is important to recognise that patents cannot be used to develop a
measure of all innovations. First, they are designed only to protect technological
inventions. Other IPR regimes exist to protect innovations in other fields for
example, copyrights for literature, trademarks for words or graphic devices
which distinguish a product and the registration of designs, and where
protection is focussed on the appearance of a product (Adams, 2006). Less
formal ways (than intellectual property rights) to protect technological
inventions also exist notably industrial secrecy, or purposefully complex
technical specifications (Frietsch and Schmoch, 2006).
15
In return for receiving
the monopoly rights inferred by a patent, the inventor is required to publicly
disclose the invention. Rather than make this disclosure, inventors may prefer
to keep their invention secret. Surveys of inventors indicate that the rate at
which new innovations are patented varies across industries (Cohen et al.,
2000 and Blind et al., 2006). For meaningful empirical analyses, it is therefore
important to control statistically for these differences in the propensity to
patent.
A further critique of patent data relates to the fact that not everything
that is patented is eventually commercialised and adopted. There are, however,
significant fees attached to the examination of a patent application (and to
renewal fees, once the patent has been granted). So it is safe to assume that, at
least in the expectations of the applicant or patent holder, the prospects for
33
commercialisation and adoption are good. Nonetheless, the economic value of
patents varies (Popp, 2005). The OECDs Triadic Patent Family only includes
patents which have been deposited at all of the European Patent Office (EPO),
the Japanese Patent Office (JPO), and the US Patent and Trademark Office
(USPTO). This ensures that the patents included in the TPF database are of
wide commercial value.
In addition, different patent offices apply different rules to define the
scope of a patent. In particular, the JPO often requires multiple patents for a
single patent at the USPTO or the EPO. When comparing patent counts across
countries, this can result in a significant bias. Most patent databases apply a
consolidation filter, in which all patents which are derived from the same
original invention are included in a single patent family. The latter is thus
defined as the set of patent applications in multiple countries pertaining to
the same invention (and thus, sharing the same priority details
16
). The
inventor country of a patent family is the country of residence of the inventor,
as stated in the priority filing. Drawing on the feature that additional fees
apply for each patent filing abroad, Lanjouw and Schankerman (2004) have
used data on patent families as proxies for the quality of individual patents.
However, when comparing patent families across different countries of origin,
the general patterns of patent filings abroad have to be kept in mind. Since
these are motivated by commercial strategies, they depend on regional
specificities, size of the domestic market, and the export orientation of an
economy (Grupp 1998).
As a result, the propensity for patent applicants to file internationally
varies significantly. These factors explain why, for European countries, the
percentage of patent families that are first filed in one European country, and
then followed by subsequent patent applications for the same invention
abroad, is relatively high (e.g. 56% for Germany in the years 2000-2005). In
contrast, the same indicator for the US (with its very large domestic market)
and for Japan (as a more closed economy) is 42% and 19%, respectively
(WIPO 2007).
17
In terms of host countries, the patent offices in the US, China
and the EPO receive the most non-resident patent filings (WIPO, 2007).
On the basis of the TPF database, the OECD has developed a wide variety
of indicators of the innovation capacity of different OECD economies. This
includes data on the total number of patents filed by country at both the
European Patent Office and the USPTO, as well as in these two offices and
the Japanese Patent Office. Figure 1.5 provides data on the number of TPF
applications by inventor country for selected OECD countries. Further data is also
being made available (e.g. citation data, inventor data) with on-going
development of the PATSTAT database; it is anticipated that this additional work
will also be exploitable for future empirical work related to eco-innovation.
34
While the TPF database eliminates certain biases in patent data, the
geographic and consolidation filters applied in its development may not be
sufficient for examining very specific areas of technological innovation. Since a
very large number of applications and grants do not pass through the relevant
filters, the resulting patent counts may be very small. The more specific the area
under analysis, the more important this problem will be. Perhaps more
Figure 1.5. Number of TPF patent applications by inventor country
1
9
8
3
1
9
8
3
20 000
18 000
16 000
14 000
12 000
10 000
8 000
6 000
4 000
2 000
0
1
9
8
4
1
9
8
5
1
9
8
6
1
9
8
7
1
9
8
8
1
9
8
9
1
9
9
0
1
9
9
1
1
9
9
2
1
9
9
3
1
9
9
4
1
9
9
5
1
9
9
6
1
9
9
7
1
9
9
8
1
9
9
9
2
0
0
0
2
0
0
1
2
0
0
2
1 200
1 000
800
600
400
200
0
1
9
8
4
1
9
8
5
1
9
8
6
1
9
8
7
1
9
8
8
1
9
8
9
1
9
9
0
1
9
9
1
1
9
9
2
1
9
9
3
1
9
9
4
1
9
9
5
1
9
9
6
1
9
9
7
1
9
9
8
1
9
9
9
2
0
0
0
2
0
0
1
2
0
0
2
A. Germany, Japan and United States
Germany Japan United States
Australia Austria Belgium
B. Other OECD Countries
Canada Denmark Finland
France Hungary Ireland Italy Korea Netherlands
Norway Spain Sweden Switzerland United Kingdom
35
significantly, these filters may not be appropriate for assessing the effects of
domestic environmental policies on innovation (see Popp, 2005). As such, the TPF
database may not always be the most appropriate source of data for very specific
research questions. In the work undertaken, patent data was therefore also
obtained from a commercial database (Delphion) for two of the three case studies.
(See Annex 1.A1 for a discussion of different sources of patent data). Delphion
applies a consolidation filter similar to that used in the TPF and as such,
differences in the breadth of patents between different countries have been
corrected for.
3.3. Patents as a measure of eco-innovation
Arguably, the identification of those innovations which are specifically
environmental in nature is facilitated through the use of patent data due to
the nature of the classification systems used for IPRs. Specifically, relevant
patents can be identified using the International Patent Classification (IPC),
developed at the World Intellectual Property Organisation (www.wipo.org).
While there are other classification systems, most researchers use the IPC
classification system when searching for patents. This classification system
involves a hierarchy of codes, structured into different levels. When patents
are granted, they are given technology classifications and sub-classifications
by various patent offices. These classifications can then be used to identify all
relevant patents for a given type of technology.
18
The IPC uses a technology-oriented approach. (Annex 1.A2 provides an
overview of the IPC system.) Since very detailed engineering-based descriptions
are given for each class, this allows for the identification of technologies which
have important environmental implications. Table 1.2 gives an idea of the
hierarchical structure, taking the example of solar concentrating devices used
for the generation of mechanical power.
This example, taken from the case study on renewable energy, is relatively
straightforward. All patent applications within class F03G 6/08 are clearly related
to renewable energy, and there are likely to be very few patents involving solar
energy concentrating technologies which do not list this class, particularly
since each application can list multiple classes.
A broader effort to identify environmental patents was also undertaken by
the OECD, building upon a search algorithm developed at the Directorate for
Science, Technology and Industry. Using the DSTI definition of environmental
patents, counts of patents deposited by different countries in a number of
different environmental areas have been derived. For most policy areas, Germany,
the US, and Japan dominate (see Figure 1.6). There has apparently been
continuous growth over recent years (particularly in air and water pollution
innovations), except for solid waste and recycling, which peaked in the
36
early 1990s. However, the rate of growth is generally lower than for overall TPF
patent activity.
Interestingly, there also appears to be some degree of specialisation that is
occurring with the US being particularly innovative with respect to water
pollution; and Japan and Germany with respect to air pollution. In recent years,
there have been relatively few patents in the area of waste disposal (Figure 1.7).
Table 1.2. IPC patent classification system for solar concentrating devices
used for the generation of mechanical power
Subdivision
Number
of Subdivisions
Example of an IPC Code
Symbol Title
Section 8 F Mechanical engineering; lighting; heating; weapons;
blasting
Subsection 21 F01-F04 Engines or pumps
Class 120 F03 Machines or engines for liquids; wind, spring, or
weight motors; producing mechanical power or
a reactive propulsive thrust, not otherwise provided
for
Subclass 628 F03G Spring, weight, inertia, or like motors; mechanical-
power-producing devices or mechanisms,
not otherwise provided for or using energy sources
not otherwise provided for
Main group Ca. 6.900 F03G 6 Devices for producing mechanical power from solar
energy.
Subgroup Ca. 62.100 F03G 6/08 With solar energy concentrating means
Figure 1.6. Number of EPO Environmental patent applications
and total EPO patent applications
120 000 600
1978
500
400
300
200
100
0
100 000
80 000
60 000
40 000
20 000
0
1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002
Waste disposal Recycling Air pollution abatement
Water pollution abatement Noise control Monitoring Total (right axis)
37
Annex 1.A3 provides data for three further sub-areas of environmental
technology and several additional countries.
While such figures are interesting at a very aggregate level, they should
be interpreted with some care. It may be relatively straightforward to identify
relevant classes in areas in which innovations take the form of end-of-pipe
pollution abatement such as electrostatic precipitators or catalytic converters to
remove air pollutants or membrane technology to remove water pollution.
However, this is more difficult when the innovation takes the form of changes in
production processes (clean production) or product characteristics (clean
products).
19
In this latter case, only a subset of applications within each class
will be relevant, and it will be necessary to use keywords to identify both the
relevant classes and the relevant applications within each class. Thus, a four-
stage procedure was adopted in this report for the two cases in which
process technologies are likely to be most important (pulp and paper bleaching
technologies and motor vehicle emissions abatement):
compile a list of relevant keywords, based on a thorough literature review;
search patent database by keywords, in order to identify relevant patent
classes;
test the keyword yield both individually and in combinations with the
patent classes identified;
20
Figure 1.7. Number of EPO Environmental patent applications
Source: Data drawn from the OECD Project on Environmental Policy and Technological Innovation
(www.oecd.org/env/cpe/firms/innov).
160
1978
140
120
100
80
60
40
20
0
1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002
DE-Waste JP-Waste US-Waste DE-Air JP-Air
US-Air DE-Water JP-Water US-Water
38
develop a final search strategy, on the basis of iterative assessment of the
testing exercise above.
Since changes in production processes and product characteristics are
increasingly prevalent as a means of addressing environmental concerns, the
problem of defining an appropriate patent search strategy is likely to become
more acute over time. Another problem related to this trend is that, generally,
product innovations are patented more often than process innovations [for
the latter, more informal ways of protecting knowledge often exist (Popp,
2005)]. With respect to environmental patents, this creates a potential bias
towards inclusion of end-of-pipe technologies in patent counts, so efforts
were made to correct for this in the search strategy adopted.
4. The policy determinants of eco-innovation
To some extent, the policy factors which drive eco-innovation will be the
same as those which drive the rate of innovation in general i.e. stable
macroeconomic conditions, economic openness, protection of intellectual
property, etc. However, steering innovation in a direction which is less
environmentally-damaging can also be done by using policy measures which
target environmental externalities directly. Whether explicitly or implicitly,
environmental policies result in either a change in the cost of different factor
inputs or a change in the relative prices of goods and services that are eventually
produced. As such, there are likely to be increased returns to the identification of
production processes and product characteristics which are environment-
saving. In this section, the implications of different measures are discussed,
along with relevant empirical evidence.
4.1. The theory of environmental policy choice and innovation
Different policy instruments will affect the incentives for firms and
households to develop and adopt environmentally-beneficial technologies in
different ways. An ideal policy provides incentives to search for both the
optimal rate of innovation, and the optimal direction of that innovation. Any
policy measure which affects relative prices or constrains production choices
will induce innovation. It is therefore not sufficient to assess the effects of a given
policy on the rate of innovation, however those effects may be measured.
Whether this innovation is welfare-improving will also depend on the direction of
innovation induced (Johnstone, 2005).
Efficiency with respect to the direction of innovation has two aspects.
First, innovation in one area can crowd out innovation in other areas. If
policy support for environmental R&D reduces R&D in other areas (i.e. health
technologies), this needs to be taken into account in policy assessment. This
issue has not been generally addressed in previous empirical work. Second,
39
within the environmental sphere, the direction of innovation should be cost-
minimising with respect to the ultimate environmental objective. In this
regard, the point of incidence of the policy is key, and it is widely argued that
a policy which is technology-neutral and which targets the environmental
objective directly, is more likely to result in a cost-minimising technological
trajectory.
A strong case has been made for the use of market-based instruments
(e.g. taxes, tradable permits), rather than direct regulation (e.g. technology-
based controls, performance standards) in order to induce innovation (see Jaffe et
al. 2002 for a review). In particular, it is argued that the rate of innovation under
market-based instruments is more likely to be optimal since a greater proportion
of benefits of technological innovation and adoption will be realised by the firm
itself than is the case for many direct forms of regulation. Moreover, since
market-based instruments are not prescriptive, they are more likely than
many types of direct regulation to ensure that the direction of technological
change is cost-minimising with respect to the avoidance of damages
(see Downing and White, 1986; Milliman and Prince, 1989; Nentjes and Wiersma,
1987; and, Jung et al., 1996). The incitation given to innovation in the US with
respect to SO
2
abatement which followed the replacement of technology-
based standards for a tradable permit regime for power plants is an oft-cited
example.
However, different forms of market-based instrument may have different
dynamic effects on the technological trajectory of the economy. For equivalent
environmental targets, auctioned tradable permits and taxes are likely to have
comparable effects. However, if policy targets are not adjusted in light of increased
information on abatement costs or environmental impacts, the effects of the two
measures may differ markedly because one is a price-based measure and the
other is a quantity-based measure. Thus, Jung et al. (1996) found that when
governments pre-commit to a given tax rate (or to a given number of permits), the
effect under the two regimes will differ since in the case of taxes the price of
emissions remains constant, even as innovation reduces abatement costs.
Moreover, with a grandfathered permit system, the innovator will face
adverse financial effects from reduced permit prices. If the innovating firm is
a seller of permits, it will have less incentive to allow for the diffusion to other
firms (unless the innovation is patented, in which case, the incentives will
depend on relative rates of return for permit and technology sales (see Milliman
and Prince, 1989). Under these (rather restricted) conditions, grandfathered
permits may perform worse than direct controls (e.g. mandated emission
reductions) in terms of incentives to induce technological diffusion. With
direct controls, the only costs will be those associated with abatement; under
grandfathering, permit sellers will also lose from increased diffusion
21
(Albrecht,
2001).
40
However, the stark juxtaposition of the innovation effects of market-
based instruments and more direct forms of regulation is a caricature. It is
important to compare measures not just in terms of the form they take (tax,
standard, permit, etc.), but also their point of incidence in terms of proximity
to the environmental externality. An environmental tax which only targets
the environmental objective in a proximate sense is actually no more likely to
lead to an optimal direction of innovation with respect to this externality than
a technology-based standard is. In this case, innovation will be misdirected
toward those technologies which save on the tax, but do not reduce
environmental damages per se. Since there can be high administrative costs to
target environmental externalities directly with environmental taxes and other
market-based instruments, this is a significant concern.
Analogously, a performance-based standard which targets the
environmental objective directly will provide similar incentives for the direction of
innovation to those of a tax which has the same point of incidence. For example,
whether a tax is imposed on the lead content of fuels, or a performance standard
is imposed in terms of lead content, innovators will seek to identify lead-
minimising technologies. However, for a given environmental objective,
incentives with respect to the rate of innovation will still be less under a
regulatory approach than under most market-based instruments. This is
because, under performance standards, savings will only develop up to the
point at which the regulatory standard is met.
In general, technology-based standards suffer from the same shortcomings
as performance-based standards with respect to the rate of innovation.
Innovators (and adopters) do not have any incentives to go beyond the standard,
unless they can induce a further regulatory response. However, by definition,
technology-based standards are unable to target the externality directly. In
order to induce an optimal direction of innovation, the regulator must also be
able to identify the cost-minimising technological trajectory, and to adjust
standards in line with this trajectory. This demands a great deal of the
regulator. Indeed, one of the most important criteria by which to assess the
innovation effects of different policy measures is the information requirements
they demand of regulators.
While the theoretical literature focuses mainly on the distinction between
market-based instruments and direct regulations, there are, of course, a variety of
other policy measures which are frequently applied. These each have different
impacts on innovation. In particular, and as noted earlier, there are a number
of policies which are specifically targeted at the innovation process. For
instance, the use of subsidies in support of environmental R&D is common,
whether in the form of grants or tax credits. There is some positive evidence
of their impacts in the area of energy efficiency (Jaffe et al., 2002 and Jaffe et al.,
2005). However, there are three main problems associated with these subsidies.
41
First, in the presence of asymmetric information between the regulator and
potentially-innovating firms, it may be difficult for regulators to identify either
recipients or promising technologies in an efficient manner i.e. there may
be problems of selection. Second, depending on the design of the subsidy
scheme, there may be significant administrative costs for both the regulator
and potentially innovating firms. And third, even if they provide incentives for
the substitution of dirty technologies for clean technologies, they have
positive scale impacts, increasing the output of the activities supported.
On the demand side, some governments have introduced various types of
technology prizes to encourage environmentally beneficial technological
change. In some cases, a market is guaranteed for the successful invention,
while in other cases, a financial prize is awarded directly. Perhaps the most
famous example is the US Super-Efficient Refrigerator Program (SERP),
which encouraged the development of a more energy-efficient refrigerator
which was CFC-free (Newell and Wilson, 2005). The winning firm (Whirlpool)
received some financial support, but the refrigerator did not actually perform
very well in the market.
22
Also on the demand side, public procurement can be used to induce
environmental innovations. Starting in 1988, the Swedish Government
established a major energy conservation programme which included about
30 technology procurement initiatives.
23
Key elements in the programme were
the bundling of demand between different consumers, the harmonisation of
specifications for the product to be procured, a competitive tendering procedure
based on these specifications, and a range of measures securing a certain initial
market (Westling, 1996; Energimyndigheten, 1998). It thus combines demand-
pull and technology-push forces for innovation (see Ostertag and Dreher,
2002).
24
An evaluation of the market transformation effects of five procurement
projects showed that, on average, the procured technologies were 30% more
efficient than prevailing products, their sales volumes were increasing, and (in
some cases) unit-costs of technologies decreased (Lund, 1997).
The use of information-based measures can also serve as a spur to
innovation, by addressing information problems either at the invention
stage, or with respect to diffusion. For instance, demonstration projects are
very common and have been used to spur investment in wind turbines in
Germany (Walz, 2007). Targeting consumers directly through eco-labels can
also have upstream impacts on invention. In a study of the determinants of
the energy efficiency of air conditioners offered for sale, Newell et al. (1998)
found that the energy labelling of appliances had a significant (and positive)
effect on the efficiency of models offered for sale, thereby complementing
energy price incentives.
42
In Section II the benefits which might arise through government initiatives
to support innovation partnerships or clusters were discussed. By bringing
together innovating firms, university laboratories, and downstream users, some
of the positive externalities which might otherwise be lost may be internalised.
The government can serve an important matchmaking role in this context,
by ensuring that positive spillovers are realised, without firms jeopardising
competitive advantages. In the environmental sphere, two examples are the
Danish Partnerships in Innovation (Danish EPA, 2007) and the Finnish
Environmental Cluster Research Programme (Honkasalo and Alasaarela,
2003).
4.2. Empirical evidence with respect to patents
There is increasing empirical evidence to support the contention that
environmental policies do lead to technological innovation. Jaffe, Newell and
Stavins (2002) and Vollebergh (2007) both provide recent reviews of the empirical
literature on this theme. The following paragraphs draw on their results.
An early paper was that of Lanjouw and Mody (1996), who examined the
relationship between the number of patents granted and environmental
policy stringency, measured in terms of pollution abatement expenditures
at the macroeconomic level, for Japan, the US, and Germany. For the
period 1971-1988, they found that pollution abatement cost affects the
number of patents successfully granted, but with a 1-2 year lag. However, their
study was not entirely satisfactory, because other factors that are likely to
affect technical innovation were not controlled for in the analysis.
Using US industry-level data, Jaffe and Palmer (1997) extended Lanjouw
and Modys study, by incorporating various factors that potentially affects
environmental innovation. They examined the relationship between stringency
and innovation more broadly (not only environmental patents) for a set of US
manufacturing industries in the period 1977-1989, where innovation was
captured in terms of both R&D expenditures and patents. They found that
increased environmental stringency (higher level of PACE) does increase R&D
expenditures. But the study did not support the hypothesis that the number of
patents increased in response to environmental regulation. They also stressed
the necessity to assess the relative strength of the effects of flexible (versus
prescriptive) environmental policy regulation regarding environmental
innovation.
Brunnermeier and Cohen (2003) built on Jaffe and Palmers work, by
narrowing innovation down to purely environmental patents. They used US
manufacturing industry data and empirically analyzed factors that determined
environmental technological innovation paying close attention to the fact that
emission reduction pressures come not only from domestic regulatory
43
authorities, but also evolve from international competition. For indicators of
emission reduction pressures, they used pollution abatement costs and the
number of inspections undertaken by the direct regulatory institutions.
Contrary to Jaffe and Palmer, they found that the PACE variable has a statistically
significant (and positive) effect on environmental innovation, whereas
subsequent monitoring does not. Moreover, they found that international
competition stimulates environmental innovation; although once again, the
effect of inspections was not confirmed.
Taylor et al. (2003) studied the time path of innovation in sulphur dioxide
(SO
2
) control, especially activities related to flue gas desulfurization. Analyzing a
100 year time span (1887-1995), they found that consistently more patent
applications were placed after SO
2
regulation was introduced in the 1970s. In
addition to SO
2
regulation, Popp (2006) also examined NO
X
regulation in the
US, as well as the German and Japanese electricity sectors to explore whether
these regulations affected (inter)national innovation and diffusion. One of Popps
main findings was that it is mainly domestic regulation that fosters innovative
activities in the home country. But he also found an important role being
played by foreign innovation in the development of these patents.
Very few studies have examined the role of policy instrument choice.
Using patent data, Popp (2003) examined the effects of the introduction of the
tradable permit system for SO
2
emissions as part of US Clean Air Act
Amendments on the technological efficiency of flue-gas desulphurisation.
Comparing patent applications after the introduction of the tradable permit
scheme with those submitted under the previous technology-based regulatory
system, he found evidence of the improved removal efficiency of scrubbers.
In one of the few available cross-country studies, de Vries and Withagen
(2005) investigated the relationship between environmental policy regarding
SO
2
and patent applications in relevant patent classifications. Applying three
different models which vary according to the manner in which policy
stringency was modelled, they found some evidence that strict environmental
policies lead to more innovation. However, they recognised that the modelling
of environmental policy in their analysis required further refinement. Moreover,
they expressed concerns about their ability to identify all relevant patent classes.
Crabb and Johnson (2007) assessed the effects of fuel prices (and thus
taxes) on innovation in automotive energy-efficient technologies in the US in
the period 1980-1999. Using USPTO patent classes, they found that applications
for patents on relevant automotive products and processes were induced by
increases in domestic wellhead extraction costs (but not by increases in the
import price of oil or the price of gasoline). This is consistent with the induced
innovation hypothesis, if it is assumed that domestic sources can substitute
for imported oil, at least temporarily.
44
4.3. Summary
It has generally been argued that environmental policy stringency should
induce innovation in environment-related technologies. However, the nature
of the policy instrument that is used seems to matter. Generally, market-based
instruments are thought to provide stronger incentives for innovation than
direct forms of regulation. Since there is an opportunity cost associated with
any level of emissions, market instruments provide innovators with the
potential to earn much greater returns. Moreover, incentives based on market
approaches tend to be continuous, and not dependent upon continually
ratcheting up the regulatory stringency.
It is also frequently argued that market-based approaches are less
prescriptive, resulting in a greater space across which innovations can be
applied. This leads, in principle, to a cost-minimising technological trajectory.
However, whether market-based approaches are, in fact, less prescriptive than
regulatory ones depends in part on the policy design. The point of incidence of
a policy is key, and there is no one-to-one relationship between instrument
choice and point of incidence. This aspect is often neglected in the literature
dealing with policy choice.
However, the empirical evidence remains limited, both with respect to
the overall effects of environmental policy on technological innovation, as
well as the more specific question of the extent to which this is reflected in
patent activity. The empirical evidence with respect to the use of other policy
measures (subsidies for environmental R&D, technology prizes, communication
tools, support for networks) is even more limited.
5. The case studies
As noted earlier, case studies were undertaken for this report in three
areas: i) renewable energy; ii) motor vehicle emissions abatement; and
iii) bleaching technologies in pulp and paper. In all cases, an effort was made
to assess the role of public policy (and other factors) on innovation. However,
due to differences in the scope of the studies, different approaches were
actually applied in each case (Table 1.3).
Issues that have been addressed in the case studies include:
the effect of environmental policy stringency in inducing technological
innovations which reduce environmental impacts;
the relative importance of different types of environmental policy
instruments in encouraging these innovations;
the potential international market opportunities for new technologies
which arise for those countries which introduce environmental policies;
45
the roles of foreign policy and market factors on the generation of domestic
environmental innovations;
the extent to which environmental innovations take the form of integrated
product or process changes (rather than end-of-pipe solutions); and
the innovation impacts associated with regulating multiple pollutants from
a single source in a coherent and integrated manner (rather than one-by-
one).
The main results from the work undertaken in these three areas are
reviewed in the Chapters which follow.
Notes
1. See Vollebergh (2007) and Jaffe et al. (2002) for recent reviews.
2. Danish exports of wind power technologies to the US are also important, but
patent family data was not collected in this case.
3. For a more detailed review of the empirical work in this area, see Syrneonidis
(1996).
4. See Binswanger (1974), Binswanger and Ruttan (1978) and Hicks (1932).
5. www.oecd.org/department/0,2688,en_2649_34451_1_1_1_1_1,00.html.
Table 1.3. Characteristics of the case studies
Renewable Energy
Generation
Wastewater from
Pulp and Paper
Motor Vehicle Emissions
Abatement
Technologies Generation of renewable
energy from wind, solar,
geothermal, wave-tide,
biomass, and waste-to-energy
sources
Elementally-chlorine free
and totally-chlorine free
bleaching technologies
Post-combustion (e.g. catalytic
converters) and integrated
(e.g. on-board diagnostics)
technologies
Major issues Policy instrument choice;
Type of innovation
Government policy
vs. consumer pressure
as market determinants;
First-mover advantage
Type of innovation;
International technology
transfer
Countries Austria, Australia, Belgium,
Brazil, Canada, Switzerland,
Germany, Denmark, Spain,
Finland, France, United
Kingdom, Greece, Hungary,
Ireland, Italy, Japan, Korea,
Netherlands, Norway, New
Zealand, Poland, Portugal,
Sweden, Taiwan, United States
Canada, United States,
Germany, Finland, Sweden,
Japan
Germany, Japan, United States
Time period 1978-2003 1975-2003 1978-2001
Data source OECD patent database Delphion Delphion
46
6. Exploration and exploitation of the earth, infrastructure and planning of land use,
control and care of the environment, protection and improvement of human
health, production, distribution and rational use of energy, agricultural production
and technology, industrial production and technology, social structures and
relationships, exploration and exploitation of space, defence.
7. Government budget appropriations or outlays for R&D provided by the
Government Budget Outlays or Appropriations of R&D (GBOARD) database were
used (OECD, 2005).
8. For further information on the survey, see Johnstone (2007) and www.oecd.org/env/
cpe/firms.
9. www.nsf.gov/statistics/nsf05305/htmstart.html.
10. Differences in sampling across countries mean that the data is not strictly
comparable across countries. For the most recent version of the CIS, see Eurostat
(2004).
11. For overviews of the use of patents as measures of innovation, see Adams (2006),
Schmoch and Hinze (2004), Grupp (1998), and Grilliches (1990).
12. The table is based on Table 9.1 and 9.3 in Somaya (2000) and Gallini (2002).
13. See Park and Wagh (2002) and Park (2002). The index includes variables which
reflect patent coverage, duration, and enforcement, as well as membership in
international treaties, and protection from restrictions on patent rights.
14. Work is underway on the harmonisation of patent applicants names, so that
patents by the same applicant have a single identifier.
15. For instance, the German Machinery and Industrial Equipment Manufacturers
Association (VDMA) has noted that patents are an ineffective means of capturing
rents when enforcement of IPRs is imperfect. See www.vdma.org/original.
16. Priority details include the Priority Application Number(s), the Priority Application
Date and the Country in which priority application was filed (see Adams 2006, p. 21).
17. Despite this low figure in 2005, Japan together with the US was still the largest
filer of patent applications in other countries, accounting for 23% of non-resident
patent filings worldwide (WIPO, 2007). Underlying these figures is Japans generally
very high number of patent applications.
18. However, care must be taken when choosing patent classifications, to ensure that
they are valid for the entire time-frame being searched. Fortunately, IPC manuals
include a record of when each classification was added. (The IPC is currently on its
8th edition.) Revisions are typically made every five years. Definitions of
classifications that have been added since the 1st edition include numbers next to
their definition, stating when the class was added.
19. See Popp (2005).
20. special algorithms can be used to calculate weights, based on the frequency of
selected keywords in titles and abstracts.
21. It must be emphasised that this distinction is not important if innovations are
generated by specialist firms which are external to the sector, and thus not
themselves involved in the permit market.
22. A variant on technology prizes arises when the government agrees to acquire
the right to a patentable invention and then to compensate the winner directly. In
theory, this should spur both invention and diffusion, since the firm will not
47
capture monopoly rents once the invention has been commercialised. While this has
been used in other areas (i.e. medicines), it has not been applied in the environmental
sphere (see Mandel, 2005).
23. For an overview and evaluation of results, see Neij (2001), Suvilehto and
fverholm (1998) and Lund (1997).
24. See BERR/DIUS/DEFRA (2007) for a discussion of the merits of such an approach in
the UK context.
References
Adams, S.R. (2006), Information Sources in Patents, Muenchen: K. G. Saur Verlag.
Albrecht, J. (2001), Tradable CO2 Permits for Cars and Trucks in Journal of Cleaner
Production, Vol. 9, pp. 179-189.
Arrow, K.J. (1962), Economic Welfare and the Allocation of Resources to Invention, in
R. Nelson (ed.) The Rate and Direction of Inventive Activity, Princeton University Press.
Binswanger, Hans (1974), A Microeconomic Approach to Induced Innovation in The
Economic Journal, Vol. 84, pp. 940-958.
Binswanger, H.P. and V.W. Ruttan (1978), Induced Innovation: Technology, Institutions,
and Development, John Hopkins University Press, Baltimore and London.
Blind, Knut et al. (2006), Motives to Patent: Empirical Evidence from Germany in
Research Policy, Vol. 35, pp. 655-672.
Brunnermeier, S. B. and M.A. Cohen (2003), Determinants of environmental innovation in
US manufacturing industries in Journal of Environmental Economics and Management,
Vol. 45, pp. 278-293.
Cohen, W.M. et al. (2000), Protecting Their Intellectual Assets: Appropriability Conditions
and Why US Manufacturing Firms Patent (or Not), NBER Working Paper No. 7552.
Crabb, Joseph M. and D.K.N. Johnson (2007), Fueling the Innovation Process: Oil Prices
and Induced Innovation in Automotive Energy-Efficient Technology, Working
Paper, Colorado Department of Economics and Business, May 2007.
Criscuolo, Chiara, Jonathan E. HASKEL and Matthew J. Slaughter (2005), Global
Engagement and the Innovation Activities of Firms, NBER Working Paper No. 11479.
De Vries, Frans P. and Cees Withagen (2005), Innovation and Environmental Stringency:
The Case of Sulfur Dioxide Abatement, Center Discussion Paper #2005-18, Tilburg
University.
Denmark Environmental Protection Agency (2007), Danish Solutions to Global
Environmental Challenges (DEPA).
Dernis, Hlne and Dominique Guellec (2001), Using Patent Counts for Cross-Country
Comparisons of Technology Output in STI Review 27 (www.oecd.org/LongAbstract/
0,3425,en_2649_33703_21682516_1_1_1_1,00.html).
Dernis, Hlhne and Mosahid Kahn (2004), Triadic Patent Families Methodology,STI
Working Paper 2004/2, Organisation for Economic Co-operation and Development,
Paris.
Downing, Paul B. and Lawrence J. White (1986), Innovation in Pollution Control in
Journal of Environmental Economics and Management, Vol. 13, pp. 18-29.
48
Energimyndigheten (ed.) (1998), Procurement for Market Transformation for Energy-efficient
Products (Eskilstuna: Energimyndigheten).
Eurostat (2004), Innovation in Europe: Results for the EU, Iceland and Norway,
Luxembourg, Eurostat.
Falk, M. (2004), What Drives Business R&D Intensity Across OECD Countries?, WIFO
Working Paper No. 236/2004.
Frietsch, R. and U. Schmoch (2006), Technological structures and performance as
reflected by patent indicators, in U. Schmoch, C. Rammer, H. Legler (eds.) National
Systems of Innovation in Comparison. Structure and Performance Indicators for Knowledge
Societies (Dordrecht: Springer).
Gallini, Nancy T. (2002), The Economics Of Patents: Lessons From Recent US Patent
Reform, in Journal of Economic Perspectives, Vol. 16, No. 2, pp. 131-154.
Quevado, G.-Garcia (2004), Do Public Subsidies Complement Business R&D? A Meta-
Analysis of the Econometric Evidence, Kyklos, Vol. 57, pp. 87-102.
Goel, Rajeev K. and Edward W.T. HSIEH (2006), On Coordination Environmental and
Technology Policy, in Journal of Policy Modelling, Vol. 28, pp. 897-908.
Griliches, Zvi (1990), Patent Statistics as Economic Indicators: A Survey, in Journal of
Economic Literature, Vol. 28, No. 4, pp. 1661-1707.
Grupp, H. (1998), Foundations of the Economics of Innovation Theory, Measurement and
Practice, Cheltenham, Edward Elgar.
Hall, B.H. and J. Van Reenan (2000), How Effective are Fiscal Incentives for R&D? A
Review of the Evidence, Research Policy, Vol. 29, pp. 449-469.
Hassan, E. (2003), Mapping the Knowledge Base for Fuel Cells: A Bibliometric Approach,
IEA Conference on Innovation in Energy Technologies, Washington, Sept. 2003.
Hemmelskamp J (1999), The Influence of Environmental Policy on Innovative Behaviour:
An Econometric Study, IPTS Working Paper, Seville.
Hicks, J.R. (1932), The Theory of Wages (Reprinted in 1948 by Macmillan, London).
Honkasalo, Antero and Erkki Alasaarela (2003), On the Cluster Approach to
Environmental Research and Development, Finnish Ministry of the Environment,
Working Paper 653.
Jaffe A., R. Newell and R.N. Stavins (2002), Technological Change and the
EnvironmentEnvironmental and Resources Economics, Vol. 22, pp. 41-69.
Jaffe A.B. and R.N. Stavins. (1995), Dynamic Incentives of Environmental Regulations:
The Effects of Alternative Policy Instruments on Technology Diffusion, in Journal
of Environmental Economics and Management, Vol. 29, pp. 43-63.
Jaffe, Adam B. and Karen Palmer (1997), Environmental Regulation and Innovation: A
Panel Data Study, in The Review of Economics and Statistics, Vol. 79, No. 4, pp. 610-619.
Jaffe, A.B. et al., (2005), A Tale of Two Market Failures, in Ecological Economics, Vol. 54,
pp. 164-174.
Jaumotte, F. and N. Pain. (2005a), From Ideas to Development: The Determinants of
R&D and Patenting, OECD Economics Department Working Paper No. 457 [ECO/
WKP(2005)44].
Jaumotte, F. and N. Pain, (2005b), An Overview of Public Policies to Support Innovation,
OECD Economics Department Working Paper No. 456 [ECO/WKP(2005)43].
49
Jaumotte, F. and N. Pain, (2005c), Innovation in the Business Sector, OECD Economics
Department Working Paper No. 459 [ECO/WKP(2005)46].
Johnstone, N. (2005), The Innovation Effects of Environmental Policy Instruments, in
Jens Horbach (ed.) Indicator Systems for Sustainable Innovation, Berlin, Springer-
Verlag, pp. 21-42.
Johnstone, N. (2007), Environmental Policy and Corporate Behaviour, Cheltenham, Edward
Elgar.
Johnstone, N. and J. Labonne (2006), Environmental Policy, Management and R&D, in
OECD Economic Studies, No. 43, 2006/2.
Jung, C. et al. (1996), Incentives for Advanced Pollution Abatement Technology at the
Industry Level, in Journal of Environmental Economics and Management, Vol. 30,
pp. 95-111.
Kivimaa, Paula and Per MICKWITZAKIVIMA (2006), The Challenge of Greening
Technologies Environmental Policy Integration in Finnish Technology Policies,
in Research Policy, Vol. 35, pp. 729-744.
Kivimaa, Paula and Per Mickwitzamakivi (2004), Driving forces for environmentally
sounder innovations, in K. Jacob et al. (eds), Governance for Industrial Transformation,
Berlin, Environmental Policy Research Centre.
Kneese A. and C. Schultze (1975), Pollution, Prices and Public Policy, Washington, John
Wiley.
Lach, S. (2002), Do R&D Subsidies Stimulate or Displace Private R&D? Evidence from
Israel, Journal of Industrial Economics, Vol. L, pp. 369-390.
Lanjouw, J.O. and A. MODY (1996), Innovation and the International Diffusion of
Environmentally Responsive Technology, in Research Policy, Vol. 25, pp. 549-571.
Lanjouw, J.O. and M. Schankerman (2004), Patent Quality and Research Productivity:
Measuring Innovation with Multiple Indicators, in Economic Journal, Vol. 114,
No. 495, pp. 441-465.
Lund, P.D. (1997), Evaluation of the Swedish Programme for Energy Efficiency: Successful
Examples of Market Transformation Through Technology Procurement, Proceedings of
ECEEE 97 Summer Study, Czech Republic, June 1997, 6 pages.
MandeL, G.N. (2005), Promoting Environmental Innovation with Intellectual Property
Innovation: A New Basis for Patent Rewards, Available at SSRN: http://ssrn.com/
abstract=756844.
Meyer, Martin (2002), Tracing Knowledge Flows in Innovation Systems, in Scientometrics,
Vol. 54, No. 2, pp. 193-212.
Milliman S.R. and R. Prince (1989), Firm Incentives to Promote Technological Change
in Pollution Control, in Journal of Environmental Economics and Management, Vol. 17,
pp. 247-265.
Neij, Lena (2001), Methods to Evaluate Market Transformation Programs Experience
of the Swedish Market Transformation Program, in Energy Policy, Vol. 29, pp. 67-79.
Nentjes A, and D. Wiersma (1987), Innovation and Pollution Control, in International
Journal of Social Economics, Vol. 15, pp. 51-71.
Newell R., G.A. Jaffe and R. Stavins (1998), The Induced Innovation Hypothesis and
Energy-Saving Technological Change, in Quarterly Journal of Economics, Vol. 114,
No. 3, pp. 941-975.
50
Newell R.G. and N.E. Wilson (2005), Technology Prizes for Climate Change Mitigation,
RFF Discussion Paper, June 2005.
OECD (2002), Frascati: Manual Proposed Standard Practice for Surveys on Experimental
Research and Development, OECD, Paris.
OECD (2004), Science, Technology and Industry Outlook, OECD, Paris.
OECD (2005), Research and Development Statistics, OECD, Paris.
OECD (2006a), Economic Policy Reforms: Going for Growth, OECD, Paris.
OECD (2006b), Main Science and Technology Indicators, OECD, Paris.
OECD (2006c), Research and Development Statistics, OECD, Paris.
OECD (2007a), Compendium of Patent Statistics (www.oecd.org/dataoecd/5/19/37569377.pdf)
OECD (2007b), Science, Technology and Industry: Scoreboard 2007, OECD, Paris.
Ostertag, K. and C. Dreher (2002), Cooperative procurement: Market transformation
for energy-efficient products in P. Clinch et al. (eds.), Greening the Budget. Budgetary
Policies for Environmental Improvement, Cheltenham, Northampton: Edward Elgar,
pp. 314-332.
Park, W. and D. Lippoldt (2007), Technology Transfer and the Economic Implications
of the Strengthening of Intellectual Property Rights in Developing Countries,
Trade Policy Working Paper, OECD, Paris, underlying data.
Park, G. Walter (2002), Patent Rights and Economic Freedom: Friend or Foe?, in
Journal of Private Enterprise, Vol. 18, No. 1, pp. 84-121.
Park, G. Walter and Smita Wagh (2002), Index of Patent Rights, in James Gwartrey et
al. (eds) Economic Freedom of the World: 2002 Annual Report (Cato Institute).
Popp, David (2003), Pollution Control Innovations and the Clean Air Act of 1990, in
Journal of Policy Analysis and Management, Vol. 22, No. 4, pp. 641-660.
Popp, David (2005), Using the Triadic Patent Family Database to Study Environmental
Innovation, OECD ENV/EPOC/WPNEP/RD(2005)2).
Popp, David (2006), International Innovation and Diffusion of Air Pollution Control
Technologies: The Effects of NO
X
and SO
2
Regulation in the US, Japan, and Germany,
in Journal of Environmental Economics and Management, Vol. 51, Issue 1, pp. 46-71.
Rollinson, J. and D. Lingua (2007), The Development of PATSTAT, Presentation at the
EPO and OECD Conference Patent Statistics for Policy Decision Making, Venice,
Italy 2-3 October 2007.
Rollinson, J. and R. Heijna (2006), EPO Worldwide Patent Statistical Database (PATSTAT),
Presentation at the EPO and OECD Conference on Patent Statistics for Policy Decision
Making, Vienna, Austria, 23-24 October 2006.
Schmooch, U. and S. Hinze (2004), Opening the Black Box, in H.F. Moed, W. Glnzel,
U. Schmoch (eds.) Handbook of Quantitative Science and Technology Research. The Use of
Publication and Patent Statistics in Studies of S&T Systems (Dordrecht: Kluwer
Academic Publishers), pp. 215-235.
Somaya, Deepak (2000), Obtaining and Protecting Patents in the United States, Europe
and Japan, in Kagan, Robert A. and Lee Axelrad (eds) Regulatory Encounters (Berkeley:
University of California Press).
51
M.-Suvilehto, H. and E. fverholm (1998), Swedish Procurement and Market Activities
Different Design Solutions on Different Markets, Proceedings of the 1998 ACEEE Summer
Study on Energy Efficiency in Buildings, Berkeley, California and Washington D.C.
UK BERR/DIUS/DEFRA (2007), Commission on Environmental Markets and Economic
Performance: Report (November 2007), www.defra.gov.uk/environment/business/
commission/index.htm
Vollebergh, Herman (2007), Impacts of Environmental Policy Instruments on
Technological Change, OECD COM/ENV/EPOC/CTPA/CFA(2006)36/FINAL.
Walz, Rainer (2007), The Role of Regulation for Sustainable Infrastructure Innovations:
The Case of Wind Energy, in International Journal of Public Policy, Vol. 2, Nos. 1/2,
pp. 57-88.
Westling, Hans (1996), Cooperative Procurement. Market Acceptance for Innovative
Energy-Efficient Technologies (Eskilstuna: Energimyndigheten).
52
ANNEX 1.A1
Sources of Patent Data
1
The TPF Database
In the Triadic Patent Family (TPF) Database, patents are included if they
have been deposited in the EPO, the Japanese Patent Office (JPO), and the US
Patent and Trademark Office (USPTO). Because of the additional costs of filing
abroad, along with the one-year waiting period that gives inventors additional
time to gauge the value of their invention, only the most valuable inventions
are filed in several countries. This ensures that the patents included in the TPF
database are of wide commercial value. Moreover, filing a patent application
in a given country is a signal that the inventor expects the invention to be
profitable in that country.
In addition to the geographic filter, a consolidation filter is applied
(Dernis and Khan, 2004). This is necessary because different patent offices
apply different rules for the scope of a patent. It is difficult to define the
boundaries of an invention in a precise manner and different offices tend to
apply different criteria. In particular the JPO often requires multiple patents
for a single patent at the USPTO or the EPO. When comparing patent counts
across countries this difference can result in a significant bias. For example,
three separate priorities, representing three separate Japanese patents, might
be shared by just one or two USPTO and EPO patents.
In the TPF database, this problem is addressed by including in a single
family all patents which share at least one common priority right. Thus, all
later patents which are derived from this original invention are included in
a single patent family. Any patents that are directly or indirectly linked to
other patents are counted as a single patent family. For example, if patent A
shares priorities 1 and 2, and patent B shares priorities 2 and 3, priorities 1
and 2 are directly linked, as are priorities 2 and 3. As they both share a direct
link with priority 2, priorities 1 and 3 are indirectly linked. All patents sharing
one of these three priorities are assembled in a single patent family in the
OECD TPF database (Popp, 2005).
53
As a result of the application of these two filters, counts of environmental
innovations using the OECD patent families will result in smaller counts than
found when data from single patent offices is used. On the one hand, all
patents which do not find wide commercial markets internationally will not
be included. On the other, patent applications with a common priority will
be bundled into a single family. As Popp (2005) makes clear, this has some
benefits; however, it also has costs.
Perhaps more significantly, the geographic and consolidation filters
applied in the development of the TPF database may not be appropriate for the
assessment of the effects of domestic environmental policies on innovation.
For instance, there may be a greater tendency for innovators from laggard
countries to deposit patents in all three offices, since their potential market
will be greater than for innovators from leader countries. Popp (2005) found
that the TPF data is downwardly biased with respect to German NO
x
and SO
2
abatement innovation, because German environmental policy predated that
of other countries, and innovators had little incentive to deposit patents in
countries with lax environmental policies.
Esp@cenet
An alternative data set is espacenet. This is a searchable on-line database
provided by the European Patent Office (EPO) that can be accessed at http://
ep.espacenet.com/. While the website includes access to four patent databases
(EP, WIPO, Japan, and worldwide), it is the worldwide database that is of most
use.
2
The worldwide database currently includes information on patents from
78 patent offices.
3
The range of data availability varies by country, so
researchers should check the availability before beginning research to
determine the time frame suitable for analysis. The most relevant data
limitations are that only recent data is available for some developing
countries, and European Patent Classifications (ECLA) are not readily available
for many non-EPO countries (with the US being an exception). This
classification scheme is based on the often-used International Patent
Classification system (IPC), but includes classifications at a greater level of
detail. More importantly, as new classifications are added, the EPO updates
the ECLA of older patents in its database.
Researchers can use the esp@cenet search pages to generate a list of
patents meeting a specific criterion (e.g. all patents from a specific country and
classification in a given year). These searches can be done using the advanced
search page. To specify a patent country, simply enter the two letter country
abbreviation in the publication number field. This will yield all patents
published in a specific country. Unfortunately, it is not possible to limit
searches to all patents with inventors from a specific country. The closest
54
approximation is to use the priority number. Since most applicants first file in
their home country, entering the same country in both the publication
number and priority number will approximate the number of domestic
patents. However, this will not yield a perfect count, as some patents have
multiple priority numbers. For example, a search of US publications published
in 2000 in ECLA class B01D 53/56 yields 35 patents. Limiting this to US priority
numbers results in 14 patents. These 14, however, include one patent from
Japanese inventors with an earlier Japanese priority number, as well as a
patent from Canadian inventors with a US priority number.
One other limitation to espacenet searches arises if the technology being
searched includes several possible ECLA codes. Only four search terms can be
combined at a given time. If more than 4 classifications are being searched,
multiple searches will need to be run. However, these searches may yield
duplicate results. For example, searches for ECLA B01D 53/34 and for
ECLA B01D 53/56 may each yield a patent that includes both ECLA. Thus,
simply adding the search results will lead to double counting. This can be
avoided by saving the individual patent numbers generated from the search,
and then manually removing the duplicates. Unfortunately, one drawback of
the espacenet site is that data on individual patents cannot be downloaded.
Thus, this would be done by first manually cutting and pasting the search
results into a separate programme.
Delphion
An alternative source that does allow users to download descriptive
data is the Delphion on-line database.
4
This database, available only by
subscription, allows the user to download multiple patent records in machine-
readable formats. Descriptive data available include the priority, application
and issue dates for each patent, the home country of the inventor, and data on
patent families. Unfortunately, updated ECLA classifications are not available
from Delphion. Popp (2006) matched this data to lists of patents from relevant
ECLA, by first getting lists of relevant patent numbers from espcenet. He then
downloaded patents from Delphion that fell into related IPC classifications,
and merged these data with the lists of patent numbers to restrict the analysis
to those falling into the relevant ECLA classes. Various counts can be obtained
from the raw data using standard data management software.
Finally, like the espacenet data, the Delphion database does not include
detailed descriptive data for Japanese patents. Some additional data on these
patents can be obtained from the Japanese Patent Office (JPO) website.
5
The
JPO does not use the ECLA. However, it uses its own system: the F-term, which,
like the ECLA, provides greater detail than the IPC. F-terms consist of a 5-digit
theme code, a 2-digit viewpoint symbol, and a 2-digit number. For example,
55
consider technologies classified as 4D002 AA12. The 5-digit theme
code 4D002 pertains to processing of waste gases. The two letter viewpoint
symbol AA pertains to compounds to be processed. Finally, the 2-digit
number specifies that NO
X
is the component processed. By searching by year,
researchers can get lists of patents for a given application year. However, the
priority date of these patents is not available, nor is it possible to separately
identify foreign and domestic Japanese patents using this database.
The EPO/OECD World Patent Statistical Database (PATSTAT)
Over the last several years, the OECDs Directorate for Science,
Technology and Industry, jointly with other members of the OECD Patent
Statistics Taskforce,
6
have developed a patent database that is suitable for
statistical analysis the OECD Patent Statistics Database.
7
Recently, further work has been undertaken by the Taskforce members
towards developing a worldwide patent database the EPO/OECD World
Patent Statistical Database (PATSTAT). The European Patent Office (EPO) took
over responsibility for development and management of the database, which
is drawn directly from the EPOs master database (Rollinson and Lingua,
2007). The PATSTAT database has been developed specifically for use by
(inter)governmental organisations and academic institutions, and is
optimised for use in the statistical analysis of patent data (Rollinson and
Heijnar, 2006).
The PATSTAT database has a worldwide coverage (over 80 patent offices)
over a time period stretching back to 1880 for some countries, and contains
over 70 million patent documents. It is updated on a regular basis biannually.
Patent documents are categorised using the international patent classification
(IPC), European classification (ECLA),
8
and national classification systems. In
addition to the basic bibliometric and legal data, the database also includes
patent descriptions (abstracts) and harmonized citation data.
Notes
1. This Annex is taken from Popp (2005) (www.oecd.org/dataoecd/35/33/38283097.pdf).
2. Note that the worldwide database also includes data on patents from the EPO,
WIPO, and Japan. One difference is that the worldwide database does not include
translations of patent abstracts from Japan, whereas the Japanese database does.
3. An updated list of available data can be found at http://ep.espacenet.com/ep/en/
helpV3/detailedcoverage.html.
4. Available at www.delphion.com.
5. This database can be found at www.ipdl.jpo.go.jp/homepg_e.ipdl.
56
6. Other Taskforce members include the European Patent Office (EPO), Japan Patent
Office (JPO), United States Patent and Trademark Office (USPTO), World Intellectual
Property Organisation (WIPO), National Science Foundation (NSF), Eurostat, and DG
Research of the European Commission.
7. The TPF Database is one of the significant outputs of this work.
8. The ECLA classification system is an extension of the IPC and contains
132 200 subdivisions (i.e. about 62 000 more than the IPC).
57
ANNEX 1.A2
Patent Classification Systems*
The TPF database contains three potential classifications for each patent.
First, the database includes US Patent Classifications (USPC) placed on US
patents by the USPTO. Second, it contains International Patent Classifications
(IPC) placed on the patent by both the EPO and the USPTO. Note that the IPC
classifications assigned to patents in the same family may differ by patent
office. The EPO uses the IPC as its main classification system. However, the
USPTO relies on the USPC system. It assigns IPC classifications to patents
based on a concordance between the USPC and IPC. However, these two
systems differ in fundamental ways. The USPC uses a function-oriented
classification system, whereas the IPC uses an application-oriented
approach. As such, most researchers use the IPC classification system when
searching for patents.
In addition, since the mid-1990s, the EPO has made use of the European
Classification System (ECLA). This scheme is based on the IPC, but includes
classifications at a greater level of detail. This additional detail allows for
separate identification of technologies, based on the pollutants they control.
For example, IPC classification B01D 53/86 includes catalytic processes for
pollution control. ECLA class B01D 53/86F2 specifies catalytic processes for
reduction of NO
X
, and B01D 53/86B4 specifies catalytic processes for reduction
of SO
2
. These classes can also help rule out irrelevant technologies. Using the
example of catalytic converters from above, ECLA class B01D 53/94 and
F01N 3/08B2 relate to catalytic converters for autos, and were thus not
included in the Popp (2005) study. Moreover, as new classifications are added,
the EPO updates the ECLA of older patents in its database. As EPO patents in
the late 1990s included these extra letters at the end of the IPC record, the TPF
database appears to use the ECLA for these more recent patents. However, it
* This Annex is taken from Popp (2005) (www.oecd.org/dataoecd/35/33/38283097.pdf).
58
does not appear to include ECLA assignments given to older patents after their
initial publication.
Because the TPF database includes original patent classifications, care
must be taken when choosing patent classifications to ensure that they are
valid for the entire timeframe being searched. Fortunately, IPC manuals
include a record of when each classification was added. The IPC is currently on
its eighth edition. Revisions are typically made every five years. Definitions of
classifications that have been added since the first edition include numbers
next to their definition stating when the class was added.
59
ANNEX 1.A3
Number of EPO Applications
in Different Environmental Areas
60
Figure 1.A3.1a.
100
120
70
1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000
60
50
40
30
20
10
0
2002
1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002
1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002
80
60
40
20
0
100
80
60
40
20
0
AT CH DE FR GB
IT JP NL SE US
Patent count
Patent count
Patent count
B. Recycling
C. Air pollution abatement
A. Waste disposal
61
Figure 1.A3.1b. (suite)
40
60
140
1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000
120
100
80
60
40
20
0
2002
1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002
1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002
30
20
10
0
50
40
30
20
10
0
AT CH DE FR GB
IT JP NL SE US
Patent count
Patent count
Patent count
E. Noise protection
F. Environmental monitoring
D. Water pollution abatement
ISBN 978-92-64-04681-8
OECD 2008
63
Chapter 2
Environmental Regulation
and International Innovation in
Automotive Emissions Control Technologies
by
Frans de Vries (Department of Economics, Stirling University)
and Neelakshi Medhi (Maxwell School, Syracuse University)*
This chapter focuses on innovative activities undertaken by the
automobile industry in the US, Europe (with an emphasis on
Germany), and Japan, focusing on the extent to which differential
environmental policy measures have affected technological
innovation in automotive emissions control technologies, within
and across these regions. The chapter distinguishes between: a)
different types of policy measures; and b) patents which relate to
post-combustion technologies (such as catalytic converters) and
more fundamental changes (such as engine re-design). Significant
evidence of technology transfer is found. In addition, the increasing
role of integrated abatement strategies is documented, with evidence
that this is affected by both fuel prices and environmental policy
design.
* Assistance comments and suggestions from David Popp Nick Johnstone and Ivan
Hascic are gratefully acknowledged.
2. ENVIRONMENTAL REGULATION AND INTERNATIONAL INNOVATION...
64
1. Introduction
As noted in the Introduction, regulatory policies which provide strong
incentives to undertake innovative activities are crucial for encouraging cost-
effective technological development in pollution control (e.g. Jaffe et al., 2003;
Vollebergh, 2007 and references cited therein). This Chapter focuses on
innovative activities undertaken by the automobile industry in the US, Europe
(with an emphasis on Germany), and Japan, focusing on the extent to which
differential environmental policy measures have affected technological
innovation in automotive emissions control technologies, within and across
these regions.
Automotive emissions are among the most important sources of local air
pollution. The major pollutants emitted by motor vehicles include carbon
monoxide (CO) and nitrogen oxides (NO
X
). In OECD countries, car emissions
account for 55% of CO and 36% of the ozone-causing NO
X
emissions.
1
Other
automobile pollutants include hydrocarbons (HC) and particulate matter (PM),
contributing 21% and 12% respectively to air emissions in OECD countries
(OECD, 2007).
2
Given the relatively large contribution of automotive emissions
to overall air pollution, reducing the amount of emissions generated by motor
vehicles can contribute significantly to improving local air quality.
The objective of this Chapter is to examine the links between environmental
regulation and technological innovation in automotive emissions control. The
following questions are addressed.
How did environmental regulation in the field of automotive emissions
control develop over time (1970-2000) in North America, Europe and the
Asia Pacific region?
What effect did environmental regulations in these regions have on the
incentives to innovate in automotive emissions abatement technologies?
Is there significant evidence of international technology transfer with
respect to different types of pollution control technologies?
Are there important differences in the factors which encouraged the
development of integrated pollution control technologies and end-of-pipe
solutions?
Patent applications and grants are used as a proxy for innovative activity.
The use of patent statistics as an indicative measure of innovation is particularly
interesting here, because these statistics are readily available, they are directly
65
tied to inventiveness (as a measure of R&D output rather than input), and they
are based on an objective standard that changes slowly on the basis on novelty
and utility (e.g., Grilliches, 1990). The availability of patent statistics allows the
analyst to trace the patent classes that specifically contain technologies that
are tailored towards controlling vehicle emissions, hence providing detail and
coverage on the relevant technologies under investigation. Moreover, given
the objective standard of assessing patents, these statistics provide some
uniformity in comparing innovation across countries.
By addressing innovation in emissions control technologies, this report
complements previous reports (e.g. Taylor et al., 2003; Jaffe and Palmer, 1997;
Brunnermeier and Cohen, 2003; Lanjouw and Mody, 1996) reviewed in the
Introduction. In this case, however, the technologies relate to a final commodity,
i.e. cars.
3
As a consequence, car manufacturers in every country need to be
concerned with regulations in both home and overseas markets.
The structure of the Chapter is as follows. Section II gives an overview of
the relevant regulations of the automobile sector in the US, Europe and Japan.
In Section III, the different types of automotive emissions control technologies
are discussed (both qualitatively and quantitatively), in terms of patent counts.
Section IV presents preliminary results of formal estimations. Section V
concludes and discusses policy implications.
2. Environmental regulation in the automobile sector
There has been significant evolution in the key regulatory measures
affecting the automobile sector in the US, Japan and the EU in recent decades.
Alongside the summary provided here, Annex 2.A1 includes all the relevant
legislation and corresponding pollutants that are currently regulated. For a
more extensive discussion of regulation and other strategies to cut back
emissions from motor vehicles, see OECD (2004).
2.1. Emission standards
2.1.1. US emission standards
In principle, compliance with performance standards, such as emission
standards for different types of pollutants, is technology-neutral in the
sense that car manufacturers can use the technology of their choice to reduce
emissions. While car manufacturers based their compliance technology
strategies generally on the use of catalytic converters in the early 1970s, by the
early 1980s, some car manufacturers were applying three-way catalysts in a
closed-loop emissions control system, using sophisticated electronic devices
for controlling engine functions. Others relied solely on the use of three-way
catalyst without these electronic devices (e.g. Bresnahan and Yao, 1985).
66
The US standards have generally had a technology-forcing character,
with the introduction of performance standards that cannot be met with
existing technology and as such have not been demonstrated in practice (e.g.
Gerard and Lave, 2005). Technology-following standards are less strict, and
can be met with existing technology. It has been argued that this is the type of
standards that was mainly adopted in Europe (Faiz et al., 1996).
Figure 2.1 depicts how the US level of emission standards for HC and NO
X
evolved in the period 1970-2004 for gasoline-driven passenger cars. Figure 2.2
does the same for CO.
4
Relative to the initial levels applied in the early 1970s,
standards are now extremely stringent. Compared to the level of 1970, the 2004
HC standard was reduced by almost 97%, whereas the 2004 standard for NO
X
has
been reduced by about 94%, relative to the level in 1973. The current (stringent)
levels for both HC and NO
X
emissions represent the Tier II standards, which
were phased in from2004, as a follow up of the Tier I standards, which had
come into effect in 1994. The Tier II for light-duty vehicles (cars and light
trucks) was fully phased in by 2007; heavy light-duty trucks will be fully
phased in by 2009. Regarding CO, the 2004 standard is 1.7 g/mile, implying
a 95% reduction, compared to the 1970 level. Since 1994, standards with
respect to PM have also been implemented, which are specified at a level of
0.08 g/mile.
5
Figure 2.1. Evolution of US HC and NO
X
(gasoline)
1970 1974 1978 1982 1986 1990 1994 1998 2002
4.5
4.0
3.5
3.0
2.5
2.0
1.5
1.0
0.5
0
Gram/mille
HC NO
x
Year
67
2.1.2. Japanese emission standards
In 1968, Japan introduced the Air Pollution Control Law. The subsequent
standards, set in 1972, required a 90% reduction of CO and HC emissions
by 1975, and a 90% reduction of NO
X
emissions by 1976 (Zhu et al., 2006).
Additional Japanese automobile regulations are embedded in the 1992 Motor
Vehicle NO
X
Law, which specified performance standards for NO
X
emissions
from in-use vehicles. In 1996, in joint cooperation with the Japan Automobile
Manufacturers Association (JAMA) and the Petroleum Association of Japan
(PAJ), the Government established the Japan Clean Air Program (JCAP).
Figures 2.3 and 2.4 how the change in Japanese emission standards for
the relevant pollutants for gasoline and diesel engines, respectively. After the
mid-1970s, the standards for gasoline driven passenger cars remained stable
until 2000. The biggest increase in stringency was then implemented for CO
emissions.
6
The standards for HC and NO
X
followed the same pattern and
have converged since 1978. The CO standard for diesel engines also coincides
with the standard for gasoline in the period 1986-1999. Compared to the
increased stringency level for gasoline driven cars in 2000 (from 2.1 g/km
to 0.67 g/km), the CO standard for diesel became more strict in 2002 (from
2.1 g/km to 0.63 g/km). The HC and NO
X
standards for diesel are generally less
strict than the corresponding standards for gasoline, except for 2005 where
the HC standard for diesel became more strict (0.024 g/km for diesel versus
0.05 g/km for gasoline). Furthermore, since 1994 PM standards have been set
for diesel, and were gradually reduced from 0.23 g/km to 0.0135 g/km in 2005; an
increase in stringency of about 94%.
Figure 2.2. Evolution of US CO standards for passenger cars (gasoline)
1970 1974 1978 1982 1986 1990 1994 1998 2002
40
35
30
25
20
15
10
5
0
Gram/mille
CO
Year
68
2.1.3. EU emission standards
The basis for EU-wide standards on automobile emissions is laid down in
Directive 70/220/EEC. This (1970) Directive specified the maximum limits for
CO and HC vehicle emissions. In subsequent decades, the stringency levels
have been increased, through a series of amending Directives. Directive
83/351/EEC imposes further stringency levels on CO emissions and introduces
Figure 2.3. Evolution of Japanese CO, HC and NO
X
standards
for passenger cars (gasoline)
Figure 2.4. Evolution of Japanese CO, HC, NO
X
and PM standards
for passenger cars (diesel)
1976 1980 1984 1988 1992 1996 2000 2004
2.5
2.0
1.5
1.0
0.5
0
Gram/km
HC CO NO
x
Year
1986 1989 1992 1995 1998 2001 2004
2.5
2.0
1.5
1.0
0.5
0
Gram/km
HC CO NO
x
PM
Year
69
limits on the combined HC + NO
X
emissions; Directive 88/76/EEC also regulates
these pollutants, as well as imposing restrictions on NO
X
emissions separately;
Directive 91/441/EEC introduces standards for PM emissions. The most important
amendments are Euro 1-Euro 5. For example, the 1992 Euro 1 performance
standard forced manufacturers to install the three-way catalytic converters in
gasoline vehicles. Since Euro 2 (1996), there have been different standards for
diesel and gasoline vehicles, and the emission limits became more stringent with
the introduction of Euro 2 to Euro 5.
Figures 2.5 to 2.9 summarise the evolution of environmental stringency
for different pollutants. CO standards for gasoline were already in place
in 1971; for diesel, they were introduced in 1984. From that time on, the CO
standards for both diesel and gasoline became more stringent, with the
standards for diesel cars relatively more stringent than gasoline after 1995
(Figure 2.5). For both gasoline and diesel cars, the environmental stringency of
CO emissions has involved a reduction of 97% in 2008, compared to their
initial levels in 1971 and 1984, respectively.
Figure 2.6 shows the historical development of EU limits for HC and
combined HC + NO
X
emissions. With respect to HC, the EU specified a limit for
gasoline cars only for 1971-1983. In this particular period, stringency increased by
almost 15% (from 2.49 g/km to 2.12 g/km). Since 1984, combined limits on HC +
NO
X
emissions came into place, since they are synergistic precursors to ozone.
Indeed, it is more efficient to regulate the sum of emissions (HC + NO
X
) than the
two pollutants separately. For gasoline-driven cars, combined HC + NO
X
Figure 2.5. Evolution of European CO standards for passenger cars
(gasoline and diesel)
1971 1975 1979 1983 1987 1991 1995 1999 2003 2007
40
35
30
25
20
15
10
5
0
Gram/km
CO petrol CO diesel
Year
70
standards were imposed until 1999; for diesel cars, they are specified until 2008.
Figure 2.7 shows the development of environmental stringency of the combined
HC+NO
X
standard for both gasoline and diesel passenger cars.
7
The standard
for HC + NO
X
emissions for gasoline cars was reduced by approximately 91%
in the period 1984-1999 (from 5.43 g/km to 0.5 g/km). For diesel, the associated
decrease in the HC + NO
X
standard in 1984-2008 was about 95% (from 5.43 g/
km to 0.25 g/km).
Figure 2.6. Evolution of European HC and HC + NO
X
standards
for passenger cars (gasoline)
Figure 2.7. Evolution of European HC + NO
X
(gasoline and diesel)
1971 1975 1979 1983 1987 1991 1995 1999
6
5
4
3
2
1
0
Gram/km
HC HC + NO
x
Year
1984 1988 1992 1996 2000 2004 2008
6
5
4
3
2
1
0
Gram/km
HC + NO
x
petrol HC + NO
x
diesel
Year
71
As illustrated in Figure 2.8, since 2000, separate standards for HC and NO
X
emissions were again implemented for gasoline-driven cars. For these types of
cars, the limits on NO
X
emissions are relatively stricter, as compared to HC
emissions in 2000-2008.
The EU has also set very ambitious targets on PM emissions for diesel
cars. The increase in environmental stringency is shown in Figure 2.9. The 1990
target of 0.27 g/km has been reduced by 98% to 0.005 g/km for 2008. As of 2008,
gasoline-driven cars will also face the same PM standard.
Figure 2.8. Evolution of European HC and NO
X
(gasoline)
Figure 2.9. Evolution of European PM standards for passenger cars (diesel)
2000 2002 2004 2006 2008
0.25
0.20
0.15
0.10
0.05
0
Gram/km
HC NO
x
Year
2000 2002 2004 2006 2008 1990 1992 1994 1996 1998
0.30
0.20
0.25
0.15
0.10
0.05
0
Gram/km
PM
Year
72
Comparing the policy history (from 1975 to today) between the three
regions with respect to emission standards, one aspect immediately stands
out. Whereas US and EU regulation show a more gradual adjustment
towards higher stringency levels, the Japanese standards have been relatively
constant for a longer period of time. However, the Japanese standards were
initially set at a relatively more stringent level.
8
For example, the Japanese CO,
HC and NO
x
standards for gasoline vehicles did not change from the mid-
1970s until 1999. Only Japanese NO
x
and PM standards for diesel vehicles
feature some gradual phase-in from the mid-1980s onward. The different
temporal patterns of regulatory tightening may have some implications for
the innovation process.
In addition, the EU experience with joint standards for different
pollutants (HC and NO
x
) may have implications for the nature of innovation.
By encouraging an integrated strategy, such standards may have provided
different incentives for innovation than under the separate standards which
preceded (and followed) this particular policy strategy.
2.2. Other relevant regulations in the automobile sector
2.2.1. On-board diagnostic systems
In addition to the use of emission standards, a growing field of technological
innovation relates to the development and implementation of On-Board
Diagnostic (OBD) systems. Two generations of OBD systems can be distinguished:
OBD-I and OBD-II. The OBD-I system was the first generation of technology that
makes use of electronic means to diagnose engine problems and to control
engine functions, such as fuel and ignition. Sensors are also used to measure
the performance of the engine, as well as the level of automotive emissions. In
addition, the sensors are helpful in providing early diagnostic assistance (B&B
Electronics, 2005). OBD-II systems are more sophisticated and ensure that
vehicles remain as clean as possible over their entire life, by monitoring
virtually every component that can affect vehicle emission performance. In case
of a problem, the system automatically adjusts performance characteristics (EPA,
2004).
In the US, the first implementation of OBD requirements occurred in
California in 1988, and was implemented by the California Air Resources Board
(CARB). This (OBD-I) system had to be installed on all 1988 and subsequent
model-year vehicles. The requirements were specifically related to the control of
fuel and ignition functions of the vehicle. Induced by this first generation of OBD
systems in California, the 1990 US Clean Air Act Amendments (CAAA) then
mandated the implementation of a more sophisticated and expanded OBD
system for the US as a whole. This advanced (OBD-II) system, developed by the
Society of Automotive Engineers (SAE), became the basis for the legal standard
73
set by the US EPA. The 1990 CAAA mandated that all light-duty vehicles and
trucks built in 1996 had to be equipped with an OBD-II system.
OBD requirements in Japan ran more or less in parallel with Japans
pursuit of a more stringent environmental policy with respect to exhaust gas
emissions, and which were introduced on 1 October 2000. Since then, OBD
systems have also been explicitly mandated. However, by 1996 Japanese car
manufacturers had already developed cars equipped with OBD-II systems
similar to those mandated in the US.
In the European Union the most stringent standards (Euro 3, Euro 4 and
Euro 5 for the years 2000, 2005 and 2008, respectively) are included in
Directive 1999/96/EC. Article 4 of this Directive states that From 1 October 2005,
new types of vehicles, and from 1 October 2006, all types of vehicles, should be
equipped with an OBD system or an On-Board Measurement (OBM) system to
monitor in-service exhaust emissions. In addition to the monitoring of in-
service exhaust emissions, monitoring of durability requirements and in-
service control, and limits for non-regulated pollutants such as Polycyclic
Aromatic Hydrocarbons (PAHs), ultra-fine particulate and formaldehyde that
may become important as a result of the widespread introduction of new
alternative fuels are also covered by this Directive. With respect to OBD
requirements, Directive 98/69/EC (Euro 3 and Euro 4) already mandates the
introduction of such systems for emission control (Euro 3).
In summary, the forerunner in the development and implementation of
both OBD legislation and OBD systems was the US, which introduced the first
generation of OBD systems (OBD-I). At the federal level, the 1990 CAAA
mandated the installation of the more advanced OBD-II system for 1996
model year vehicles. Although Japan formally mandated the implementation
of OBD-II systems as of October 2000, Japanese automobile manufacturers had
already installed this type of system by 1996. European legislation on OBD
systems, included in Directive 1999/96/EC, lags behind the OBD policy
implementation in both the US and Japan. The first European-wide installation of
OBD systems was only required by 1 October 2005.
Regulation tailored to OBD systems differs from regulation that is explicitly
di rected towards reducing exhaust emi ssi ons from vehicl es. The
environmental motivation for OBD regulation seems to have been of a more
technology-following character. For example, the principal motivation for
developing OBD was related more to (general) improvements in engine
performance and engine design, rather than on reducing pollution emissions per
se. The latter effects were incidental benefits associated with the development
of more advanced OBD systems. The policy implication is that the public
good motivation for OBD regulation came into being once these systems
revealed their potential for environmental benefits, but only after they had
74
already proven to be useful for non-environmental reasons (i.e. engine
performance improvement).
This makes the nature of OBD regulation fundamentally different from
regulation mandating the development and installation of catalytic converters. In
the former case, there is a mix of private and social benefits, while in the latter
case, the benefits are exclusively social. Whereas post-combustion devices only
generate environmental benefits at the end-of-pipe, the environmental
benefits from OBD systems could be classified as being complementary with
other financial benefits, such as increased fuel efficiency and reduced
maintenance costs. Incentives for innovation in the two cases will be very
different.
2.2.2. The lead phase-down
Alongside CO, HCs, NO
x
and PM emission standards, lead (Pb) standards
were also introduced over the course of the last couple of decades. The
catalysts used to reduce other local air pollutants are rendered inactive by
lead. This incompatibility necessitated the use of unleaded fuels in vehicles
with catalytic converters, leading to a gradual phase-down of leaded gasoline
in the US during the 1970s and 1980s (e.g. Kerr and Newell, 2003).
The phase-out of lead in Japan began during the 1970s, and by the
early 1980s, only 1-2% of gasoline contained lead (Michaelowa, 1997). The
production and use of leaded gasoline has now been fully eliminated in Japan.
In Europe, Germany was the first country to adopt standards to control
the lead content of gasoline ranging from 0.4 grams lead per liter in 1972
(benchmark of 0.6 grams per liter) to 0.15 grams of lead per liter in 1976.
In 1985, Germany also passed a law to reduce total automobile emissions, and
included the introduction of unleaded gasoline, because the largest reductions
of NO
x
and CO could be achieved by catalytic converters that were incompatible
with lead (Von Storch et al., 2002). As of 1981, the EU set a standard of 0.4 grams
lead per liter (Council Directive 78/611/EEC), which lagged almost a decade
behind the German law. From October 1989, all EU member States had to offer
unleaded gasoline, with a maximum of 0.15 grams of lead per liter. The 1998
Aarhus Treaty mandated the use of only unleaded gasoline by 2005.
2.2.3. Fuel economy
Emission standards as examined above represent an upper limit to the
amount of a certain pollutant that can be emitted per vehicle kilometer (mile)
driven. American Corporate Average Fuel Economy (CAFE) standards refer to
the amount of fuel that is required for a vehicle to travel a certain distance (for
example, in terms of litres/kilometer). Although fuel economy is obviously of
primary importance to the automobile sector, this Chapter does not try to link
75
innovation in fuel economy of cars to CAFE regulations. Further OECD work
looking at both fuel efficiency and emission of local air pollutants is currently
envisioned.
9
3. Innovation in automotive emissions control technologies
3.1. Automotive emissions control technologies
Automobile pollution control technologies comprise all technologies that
are used to reduce pollutants generated and released into the atmosphere by
automobiles. These automotive-generated emissions fall broadly into two
categories, based on the point of emission: a) tailpipe or exhaust emissions;
and b) evaporative emissions.
Tailpipe emissions are produced as a by-product of the (imperfect)
combustion of fuels to power the vehicle.
10
These comprise the pollutants that
are released from the vehicles exhaust system. Evaporative emissions are
produced as a result of the evaporation of fuel, due to heating of the vehicle or
release of vapor while refueling. The gas tank heats up due to high day-time
atmospheric temperatures and/or from engine heat (even after coming to a
halt). Gasoline in the tank evaporates when the temperature of the tank
increases, raising the pressure inside the tank, which has to be released and
vented into the atmosphere. Any vacant space in a gasoline tank is filled with
HC vapor, which is forced out of the tank and into the air during refueling.
These particles are heavier, thus remaining closer to the ground level.
As such, effective abatement of pollution from vehicles must target both
tailpipe emissions and gas tank venting. Searches conducted for these
technologies are primarily based on specific regulations that were imposed on
the automobile sector like the US-Tier standards, as well as the EUs Euro
standards.
11
Tailpipe emissions can be controlled by increasing engine efficiency,
increasing vehicle efficiency, increasing driving efficiency, or cleaning up the
emissions emitted. Of these four control methods, driving efficiency is not
improved through technological innovations, but depends on non-
technological aspects, such as driving techniques and levels of congestion.
These issues are not considered here. Aspects that increase vehicle efficiency
using light-weight materials or aerodynamic design of the external body of
the vehicle are not addressed.
Which type of pollutant (and how much) is emitted from motor vehicles
is largely determined by the type of vehicle engine installed and the type of
fuel consumed. Two of the most common types of engines are spark-ignition
engines and diesel engines. Emissions from spark-ignition engines can be
reduced by re-designing the engine, changing conditions under which
combustion takes place, and treatment of post-combustion of pollutants by
76
using catalysts before releasing into the atmosphere. Technologies that
improve fuel efficiency also reduce exhaust emissions, while some emissions
control requirements have also resulted in improved fuel quality and fuel
efficiency.
Advances in engine and vehicle technology continually reduce the amount
of pollutants generated, but this is generally considered insufficient to meet
environmental objectives. Technologies designed to react with, and thus clean
up, the remaining emissions (such as catalytic converters, catalytic regeneration
technology, and particulate filters) are therefore an important part of emissions
control efforts. While tailpipe emissions result from the combustion process,
evaporative emissions can arise even when the engine is idle. Technologies to
control evaporative emissions require additional modifications which are not
necessarily related to engine design.
3.2. Patent data: identification of classes and descriptive results
3.2.1. Identification of classes
The World Intellectual Property Organization (WIPO)
12
descriptions on
the IPC classification (8th Edition) was used to identify IPC classes that
matched the emissions control technologies described above. A total of
67 different IPC classes were identified that dealt with the purification of
gases and emissions control (see Annex 2.A1).
The 67 identified IPC classes are broadly categorised into the three major
technology groups identified above: 1) those that relate to improvements in
engine re-design, and therefore generate less emissions; 2) those that treat
pollutants produced before they are released to the atmosphere; and 3) those
that reduce evaporative emissions. Unfortunately, the latter category is
somewhat opaque, because there is no IPC sub-classification that clearly
defines improvements to nozzles and/or canisters. However, both modifications
to fuel refilling nozzles as well as charcoal canisters show up under classes
already grouped under some tailpipe emissions category. For example, charcoal
canisters are included under IPC code F01N3/28.
The three broad categories have been further subdivided into nine
technology groups, reflecting the technologies discussed in the preceding
Sections. Eight of these subgroups pertain to tailpipe emissions, and only one
to evaporative emissions. Six subgroups are classified under improved engine
re-design and two under post-combustion devices. Two-way, three-way,
and lean-NO
X
catalyst/catalytic converters are grouped under one technology
heading (catalytic converters). Similarly, electrical, electronic and plasma-
based technologies are placed under one subgroup. Table 2.1 summarises the
technologies covered in the analysis.
77
Finally, note that (except for the US) all patent data includes both granted
patents and the number of patent applications; the US patents cover only
granted patents. The period of focus is 1975-2001.
13
3.3. Descriptive results
This Section presents the main descriptive results of the patent searches.
It shows the development of the number of patent filings over time in the
different regions. Given differences in the data (i.e. only granted patents in the
US), the results should be primarily interpreted in terms of trends, rather than
absolute differences in patent counts between these regions. Details are also
presented about who files where (i.e. which country is the home-country of the
inventor the source country) and in which country the inventors file for
patent protection. The latter provides some insight into the role of regulation
in market segmentation and technology transfer.
3.3.1. Trends in patent applications
Figures 2.10-2.11 illustrate the number of patent applications that have
been filed in the US and Germany. (Recall that the US patents only include
those that have been granted, and not the total number of applications.) The
Figures also show the evolution of the two main technology groups: engine re-
design and post-combustion patents, as well as a group called integrated.
This latter group includes both engine design and post-combustion treatment
components. Therefore, they could not be assigned solely to one of the other
two groups.
Given the starting date of the examination period (1975), the first relevant
implementation of standards in the US was the tightening of the upper CO
limit in 1981. The second increase in the stringency of environmental policy
concerned the HC and NO
X
standards in 1994. There was hardly any effect in
the number of patent applications prior to the imposition of the 1981 CO
limits. Between 1981 and 1994, however, a positive trend in all three types of
patent applications can be discerned.
Table 2.1. Technologies covered
Improved engine re-design technologies Post-combustion technologies
Air-fuel ratio devices Catalytic converters and catalytic regeneration technology
(CAT)
Exhaust gas recirculation (EGR) valves Particulate filters and regeneration
Oxygen, NO
X
and temperature sensors
Electronic control systems and plasma-based
technologies
Fuel injection systems
Crankcase emissions and control
78
A more prominent effect might be the anticipated impact of the OBD
requirements. In the period prior to the implementation of OBD-I (1988),
patents on engine re-design exhibited a moderate increasing trend. However,
a significant increase in the trend rate of the number of patents related to
engine re-design is evident, starting from1992-1993. This increase makes
sense, because OBD was specifically tailored towards diagnosing engine
problems, as well as towards controlling engine functions. However, the
beginning of the increase in the trend rate falls between the point of introduction
of OBD-I and OBD-II, perhaps reflecting anticipation of the latter regulation. Even
after its introduction, engine re-design patents exhibited strong growth.
Engine re-design patents filed in Germany increased from1992 onwards,
preceded by a three-year decrease in patent activity (Figure 2.11). Inventors do
not seem to have anticipated the introduction of stricter standards for CO,
HC+NO
X
and PM emissions that were set in the EU in 1992. Rather, a higher
degree of innovative activity took place after the standards were implemented.
The increase in the number of patent applications in Germany continues after
the implementation of stricter EU stringency levels for CO, HC + NO
X
and PM
in 1996. The effects of these policy shocks on the innovative output in post-
combustion technologies is less clear. There is a moderate positive trend, with
less pronounced changes in the region of the policy shocks. This also holds for
patenting activity related to integrated technologies: a steady increase can be
seen in patent filings as of 1987. It is only after 1996 that a sharp increase in
integrated technology patents is observed.
In 1985, Germany was the first European country to pass a law to reduce
total automobile emissions, including the introduction of unleaded gasoline.
Figure 2.10. Evolution of patent applications at the USPTO, 1975-2001
1993 1997 1995 1999 2001 1983 1987 1985 1989 1975 1979 1977 1981 1991
1 800
1 600
1 400
1 200
1 000
800
600
400
200
0
Patent applications US
NO
x
HC CO
NO
x
OBD-l OBD-ll HC
NO
x
Engine re-design Post-combustion Integrated
Priority year
79
The latter was initiated because the largest reductions of NO
X
and other
pollutants could be realised with catalytic converters (catalysts that were
incompatible with the use of leaded fuel). Catalytic converters that could deal
with lead were already in use in the US, because the US automobile industry
was faced with relevant standards (Von Storch et al., 2002). From1982-1985,
there was a 236% increase in post-combustion patent filings. It is likely that
this effect came about from inventors anticipating this new law, particularly
since Germany had already drafted the law in 1984. Patents for engine re-
design technologies also increased in the years prior to 1985, but not as much
as catalytic converters: 22% in 1983-1985.
Prior to the introduction of a stricter policy for NO
X
in Japan in 1978, there
was actually a decrease in post-combustion patents. Japanese car manufacturers
successfully developed the first three-way catalyst in 1977, perhaps partly in
anticipation of US regulations. The first US-invented three-way catalyst was
not commercialised until 1981 (Gerard and Lave, 2005). Following the US in
their pursuit of a more stringent environmental policy, in 1978 and 1979, Japan
implemented the most stringent limit on NO
X
emissions world-wide for
gasoline and diesel engines. Domestic regulation also directly triggered the
innovative anticipation of Japanese car producers that resulted in the
development of the three-way catalytic converter (Zhu et al., 2006).
In addition to the NO
X
regulation, Zhu et al. (2006) argued that other
aspects played a role in the timing of the development of the new three-way
catalyst in 1977. For instance, the three-way catalytic converter was preceded
by the development of the Compound Vortex Controlled Combustion (CVCC)
Figure 2.11. Evolution of patent applications at the GPTO, 1975-2001
1993 1997 1995 1999 2001 1983 1987 1985 1989 1975 1979 1977 1981 1991
1 400
1 200
1 000
800
600
400
200
0
Patent applications Germany EU: PM diesel
EU: CO diesel /gasoline
EU: CO, PM,
HC + NO
x
EU: HC, PM,
NO
x
, CO diesel
OBD
Priority year
EU: HC + NOx
EU: member states offer
unleaded gasoline
Introduction unleaded
gasoline in Germany
80
engine. This passed US EPA certification in December 1972 (Zhu et al., 2006).
This is a clear example of an innovation that was developed at least in part in
response to foreign regulation. Despite its potential, the CVCC engine ultimately
did not become the dominant technology because it could not effectively cope
with fuel economy and engine power problems. Moreover, the engine
encountered difficulties in meeting the tight 1976 US standards for HC, CO and
NO
X
. The relative failure of CVCC technology also triggered Japanese competitors
to come up with a substitute technology that would be able to meet existing
regulatory requirements.
Surprisingly, this particular technological achievement is not reflected in
the number of patent applications, either in Japan or in the US. In fact, catalytic
converter patents filed for protection in Japan actually decreased in the year of
its introduction (around 1977), and continue to exhibit a decreasing trend over
the entire period 1975-1984. It was not until 1985/1987 that Japanese patent
applications started to grow again, probably because of the introduction of CO
and HC standards for diesel-driven passenger cars.
In the period 1975-1982, there was a decreasing trend for US patent
applications related to catalytic converters. However, in 1979 and 1980, there
was a slight increase in the number of patent applications related to catalytic
converters, increasing from a total of 72 applications in 1978, to 78 in 1979 and
84 filings in 1980. Only in the EU patents was there an increasing trend in the
number of patent filings over the whole period, starting in 1977.
Figure 2.12. Evolution of patent applications at the JPO, 1975-2001
1993 1997 1995 1999 2001 1983 1987 1985 1989 1975 1979 1977 1981 1991
3 500
3 000
2 500
2 000
1 500
1 000
500
0
Patent applications Japan
PM diesel
Introduction CO and
HC diesel standards
US: OBD-ll CO, HC, NO
x
gasoline
Priority year
NO
x
gasoline NO
x
diesel
81
With respect to trends in overall Japanese patent applications (Figure 2.12), it
can be seen that after the implementation of the tighter NO
X
standard in 1978
engine re-design patents increased significantly and started to level off in 1983-
1984. Around 1985, when maximum CO and HC emissions for diesel vehicles
are introduced, there was a very sharp decrease in engine re-design patents,
followed by a very steep increase in 1986, and a strong decrease in corresponding
patents in 1987. Thus, inventors filing in Japan appeared to respond quite
dramatically in terms of innovative output in the mid-1980s.
The next Japanese policy shock occurred almost ten years later, with the
introduction of PM standards in 1994. In the period before 1994, both post-
combustion patents and integrated technology patents showed an increase,
which started around 1987. In contrast, after 1990 engine re-design patents
decreased until 1994. In 2000, Japan also imposed stricter CO, HC and NO
X
limits for gasoline cars. As of 1996, both engine re-design and integrated
patents increased towards 2000; patents related to post-combustion devices
exhibited a more stable and constant development over this period.
Recall that although Japan mandated the implementation of OBD-II systems
as of October 2000, this system was already installed by Japanese automobile
manufacturers in their 1996 vehicles. The Japanese patent data shown in
Figure 2.12 illustrates that, in the period 1995-2001, engine re-design patents and
integrated technology patents increased by 44% and 208%, respectively.
Finally, patents deposited at the EPO (Figure 2.13) can also be examined.
14
Following Germanys lead, the EU mandated that all Member States had to
offer unleaded gasoline as of 1989. In 1989, the number of patents filed for
engine re-design technologies (as well as post-combustion technologies)
increased. However, immediately after 1989, there was a decrease in engine
re-design patent filings, which lasted until 1992, when CO, combined
HC + NO
X
and PM standards became more stringent. After 1992, engine re-
design patents showed a very sharp increase, particularly after 1996, the year
in which the stringency of mandated levels for CO, HC + NO
X
and PM was
increased. Since 1997, integrated patents and post-combustion patents also
tended to increase relatively more quickly, in comparison with the period
before 1996.
Figures 2.10-2.13 reveal some common developments. First, engine re-
design patents were more common than patents for post-combustion devices.
Second, all three technology classes had a positive trend, but the trend was
most pronounced for engine re-design patents. Third, patents related to
integrated approaches started to became much more common in the
late 1980s and early 1990s. Fourth, for US, German, and EPO patents, the growth
in innovative output was strongest in the period 1992-2000. For Japanese patents,
the growth was less pronounced and started a bit later (around 1994).
82
Within the engine re-design group, patenting is mainly concentrated on
electronic control systems and plasma-based technologies, and fuel injection
systems. However, as can be seen in Figure 2.14 there are differences across
countries. Compared to the US, Germany, and the EU, there was a greater
concentration in Japan on the development of electronic control systems, and
less on fuel injection. Within the overall post-combustion technology group (not
contained in Figure 2.15), innovation was largely focused on the development of
Figure 2.13. Evolution of patent applications at the EPO, 1975-2001
Figure 2.14. Technology shares within engine re-design group
at different patent offices (1975-2001)
1993 1997 1995 1999 2001 1983 1987 1985 1989 1975 1979 1977 1981 1991
1 000
900
800
700
600
500
400
300
200
100
0
Patent applications EU
PM diesel
CO diesel /gasoline
Member states have to
offer unleaded gasoline
CO, PM,
HC + NO
x
HC, PM, NO
x
CO diesel
Priority year
HC + NO
x
%
70
60
50
40
30
20
10
0
United States Germany Japan EU

A
i
r

f
u
e
l

A
i
r

e
l
e
c
.

C
r
a
n
k
c
a
s
e

E
G
R

E
l
e
c
.

c
o
n
t
r
o
l

F
u
e
l

i
n
j
e
c
t
.

S
e
n
s
o
r
s
83
catalytic converters (about 95% in all four regions); the rest of the innovation
was in the field of particulate filters.
15
3.3.2. Source countries
Figure 2.15 provides information on the source countries of USPTO, GPTO,
JPO and EPO patents by type of technology. The first four columns on the left-
hand-side of the first panel give the share of domestic (successful) US inventors
that filed in the US, disaggregated according to the three different technological
classifications, as well as in total. Surprisingly, more of the patents granted in the
US came from Japanese inventors than from US inventors. German inventors,
and others followed. For the different types of technology, domestic US
Figure 2.15. Source countries for patents (1975-2001)
60
50
40
30
20
10
0 0
10
20
30
40
50
60
70
% %
USA JPN DEU USA JPN DEU
100
80
60
40
20
0 0
10
20
30
40
50
% %
USA JPN DEU USA JPN DEU
Engine re-design Post-combustion Integrated Total
Source countries Unites States patents Source countries German patents
Source countries Japanese patents Source countries EPO patents
Rest Rest
Rest Rest
84
inventors took the lead with respect to post-combustion technologies, while
Japanese inventors were granted more US patents in the fields of engine re-
design and integrated technology approaches.
Concentrating on Germany, the upper-right panel of Figure 2.16 illustrates
that most of the patented innovations (almost 60%) in all three technology
domains (as well as in aggregate) came from domestic inventors. Japanese
inventors filed the next most frequently, followed by the other category. On
average, US inventors filed least often for patent protection in Germany. The
domestic bias was even more pronounced for Japanese patents (lower-left
panel). More than 80% of patents filed in Japan come from Japanese inventors
filing in their home country in all three technology domains.
European patents exhibited a somewhat different picture in this respect.
The number of patent applications that came from different countries is more
uniform. Overall, EPO patents consisted mainly of German inventions, followed
by Japanese. However, compared to Germany, Japanese inventors filed more
frequently for patents for integrated technologies. The US did the least
patenting in this technology domain in Europe, but stood more-or-less on an
equal footing with Japan regarding post-combustion patents.
Patent data also provides some additional insight into the pattern of
inventive knowledge flows across countries. International patent families are
based on the original invention, which are filed in other countries, hence
creating protection of the original invention in these countries. After the
initial application, inventors can file for patent protection within one year.
Figure 2.16. Average patent family size by country and year
1975 1977 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2001
4.5
4.0
3.5
3.0
1.5
1.0
2.5
2.0
0.5
0
Priority year
Average family size
United States Germany Japan
85
Since it is costly to file in multiple countries, international patent families can
serve as an indicator of the relative international market potential of the
invention.
Figure 2.16 shows the evolution of the average international patent
family size of domestic patents across both engine re-design and post-
combustion technologies during the period 1975-2001. Japanese patents had
the smallest patent family size. This implies that Japanese inventions were
essentially only filed in Japan. This confirms the result that is reflected in the
lower-left panel of Figure 2.15, namely that most of the patents filed in Japan
came from domestic inventors across all technology domains. US and German
inventors applied for patent protection in overseas markets to a greater
extent.
In addition, the variability in the size of the average Japanese patent
family has been relatively less over time, with a small increase in the average
family size after 1977. This might be due to the Japanese development of the
three-way catalytic converter at that particular time. The average family size
of US and German patents fluctuated more over time, but has remained
largely in the 3-4 range, following more-or-less the same pattern. Both the US
and German patents showed a decreasing trend in family size as of 1997.
3.3.3. Technology transfer
Although Figure 2.16 provides information on the average size of the
international patent family thereby illustrating the propensity for domestic
inventors to file abroad it does not show which country files where for patent
protection, and when. The latter information is summarised in Figures 2.17-2.20,
which shows the change in patent filings over time, grouped by inventor
country. The Figures provide a specific indication of the knowledge flows
across the different countries for both engine re-design patents and post-
combustion patents.
The top panel of Figure 2.17 provides data on the number of US patents
for engine re-design disaggregated by source country domestic, German, and
Japanese inventors and a rest of world category referred to as other. The
lower panel presents the same information for post-combustion patents. The
main policy shocks are also indicated in the different graphs. The top-panel
reveals that, over time, most of the US patents in engine re-design came from
Japanese inventors.
Focusing on the OBD-I and OBD-II regulations in 1988 and 1996 shows
that prior to 1988 the number of Japanese inventors filing in the US grew only
marginally. One year before the introduction of OBD-I German inventors
started to apply for patent protection more frequently. There also appears to
have been some anticipation of the introduction of OBD-II on the part of
86
Japanese and German inventors. However, Japanese patenting in the US
declined after the actual introduction of OBD-II regulation. Surprisingly, two
years before OBD-II, the number of US patents that came from domestic
inventors actually decreased, but started to increase significantly following
the implementation of OBD-II. As can be seen in the lower panel, until 1986
the level of innovative output with respect to post-combustions technologies
was relatively stable. However, it started to increase significantly after 1986.
With respect to engine re-design, most of the patents that were filed in
Germany came from domestic inventors (Figure 2.18). Until 1992, there was
a slight increasing trend, but the rate of growth in the number of patent
Figure 2.17. Source countries of USPTO engine
re-design and post-combustion patents
1975 1977 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2001
700
600
500
400
100
300
200
0
1975 1977 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2001
160
140
120
100
40
20
80
60
0
Priority year
Patent applications
United States
NO
x
CO
NO
x
HC
NO
x
HC OBD-l OBD-ll
NO
x
CO
NO
x
HC
NO
x
HC OBD-l OBD-ll
Germany Japan Other
United States Engine (re)design
Priority year
Patent applications
United States Post-combustion
87
applications became significantly higher after 1992. Although the first EU-wide
installation of OBD systems was only required on 1 October 2005, these
provisions are included in Directive 1999/96/EC, which is indicated by the
vertical line in the top-panel of Figure 2.18. There was probably some
anticipation on the part of German inventors with respect to EU policy, but it is
difficult to trace the length of the anticipation period. Foreign patenting in
Germany remained largely constant over time, perhaps with a slightly increasing
trend.
Figure 2.18. Source countries of German engine
re-design and post-combustion patents
1975 1977 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2001
1 000
900
800
700
600
500
400
100
300
200
0
1975 1977 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2001
250
200
150
100
50
0
Priority year
Patent applications
United States
EU: OBD
EU: CO, PM,
HC + NO
x
Introduction unleaded
gasoline in Germany
EU member states have
to offer unleaded gasoline
EU: CO diesel /
gasoline
EU: PM diesel
Germany Japan Other
Germany Engine (re)design
Priority year
Patent applications
Germany Post-combustion
88
German post-combustion patents, as highlighted in the lower panel of
Figure 2.18, followed a similar pattern. That is, domestic inventors were the
prominent source of the patent filings in Germany. The degree of patent
output increases over time, particularly after 1982. However, from2000 there
was a sharp decline. There appears to have been some anticipation in the
development of post-combustion patents (mainly catalytic converters) prior to
the German introduction of unleaded gasoline in 1985. Thus, German inventors
mainly developed catalytic converters that could deal with lead by themselves
without importing them from the US, which already had developed that specific
Figure 2.19. Source countries of JPO Engine re-design
and post-combustion patents
1975 1977 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2001
600
500
400
300
200
100
0
1975 1977 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2001
80
70
60
50
40
30
20
10
0
Priority year
Patent applications
United States
CO, HC, NO
x
gasoline
Development Japanese
three-way catalytic converter
Introduction CO and HC
diesel standards
PM diesel
Germany Other
Japan Engine (re)design
Priority year
Patent applications
Japan Post-combustion
89
technology earlier (see e.g. Von Storch et al., 2002). As Figure 2.18 illustrates, this
German landmark law brought the level of inventive activity in the field of
post-combustion technologies to a higher level. Since then, the number of
patent applications in Germany has increased more steadily, probably (partly)
induced by more stringent EU regulations regarding CO, PM and combined
HC+NO
X
standards. The effects of EU policy measures on foreign patenting are
more difficult to disentangle.
Figures 2.19 and 2.20 provide the same information for Japan. Foreign
applicants are included in a separate graph (Figure 2.19) because of scaling,
since the number of Japanese patent filings (Figure 2.20) is considerably
greater. German inventors filed more frequently in Japan than US and other
inventors. Moreover, the number of German patent applications seeking
protection in Japan increased considerably after 1996. This difference is
mainly due to engine re-design patents. While German inventors have filed
significantly more over time in Japan in the field of engine re-design technologies
compared to the US and other, there is no difference in post-combustion
technologies. All three sources grew rapidly from the mid-1980s, the year in
which Japan introduced CO and HC standards for diesel passenger cars. The
variation in post-combustion patents filed by US, German and other inventors
is likely to have been largely driven by the adjustment of the corresponding
stringency levels of the standards for diesel engines, since the standards of CO,
HC and NO
X
emissions for gasoline-driven passenger cars were not adjusted
during 1978-1999.
Figure 2.20. JPO engine re-design and post-combustion patents
by domestic inventors
1975 1977 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2001
4 000
3 000
2 500
3 500
2 000
1 500
1 000
500
0
Priority year
Patent applications
Japan: Engine (re)design Japan: Post-combustion
Japan Engine (re)design and post-combustion
CO, HC, NO
x
gasoline
Development Japanese
three-way catalytic converter
Introduction CO and HC
diesel standards
PM diesel
90
Figure 2.20 provides information on patent filings in Japan by domestic
inventors. By the end of the 1970s, the degree of domestic patent applications
in engine re-design technologies started to diverge significantly from the level
of domestic post-combustion patents. Further, where the latter form of
innovation showed a rather gradual increase over time in the number of
patent applications, the pattern of engine re-design patents featured more
variation, including a sharp peak in 1986. It is not certain if this sharp fluctuation
arose from the introduction of the aforementioned CO and HC standards for
diesel engines. This seems unlikely given, that the implementation of emission
standards is more oriented towards the technological advancement of post-
combustion techniques.
4. Empirical analysis and results
The limited size of the sample and the difficulty in modeling policy
measures in a commensurable manner across countries restricts the issues
that can be addressed empirically. However, based on the data collected, some
preliminary analysis was undertaken on the relative importance of
environmental policies and other factors on both post-combustion and engine
re-design patents. Given that the latter are likely to result in private non-
environmental benefits, it is expected that factors other than environmental
policy (i.e. fuel prices) are likely to play a more important role than in the
former case.
An econometric model was, therefore, estimated to evaluate the impact
of environmental policy and market drivers as determinants of patenting
activity in automotive emission control technologies. A dataset consisting of a
panel of three countries (US, Japan, and Germany) over a period of 24 years
(1978-2001) was used to estimate the regression models. The full model is
specified as:
where i = 1, 2, 3 indexes the country (US, Japan and Germany) and t = 1978, ,
2001 indexes the year.
The dependent variable, a patent count (PATENTS), is the total number of
(successful or unsuccessful) patent filings deposited, distinguished by country
(US, Japan, and Germany), year (1978-2001), and technology group (post-
combustion devices vs. engine re-design technologies). This distinction is
made because it is expected that environmental policy and market drivers
have different impacts on innovation in the respective technological groups.
The proxy variable for environmental policy is based on the introduction
of OBD requirements in the US. As noted above, the introduction of
( )
, , , _ ,
it it it it it i it
PATENTS f EnvPolicy ENPRICE VA PAT TOTAL = +
91
regulations mandating OBD systems in the US appears to have had important
cross-border implications for Japanese and European manufacturers. The
early introduction of such regulations in the US, coupled with the size and
importance of the US car market, was an important driver of innovations
overseas. In this paper, the environmental policy shock is thus represented by
the introduction of OBD-I and OBD-II requirements in the US (DUM1_USOBD,
DUM2_USOBD) two dummy variables that have the value zero for all years
until a given OBD system was mandated in the US, and the value of one
thereafter.
The fuel price (ENPRICE) variable is constructed as the after-tax price of
unleaded gasoline, and represents the market driver of innovations in the
automobile industry.
16
For instance, a recent study by Crabb and Johnson
(2007) found that higher energy (oil) prices stimulate innovation in the field of
energy-efficient automotive technology.
17
Value added by the automobile
industry (VA) serves as an indicator of the scope of technological opportunities, as
well as the potential financing of expenditures for research and development in
the industry.
18
Total patents (PAT_TOTAL), measured as the total number of patent filings
deposited with the three national patent offices (the US, Japan, and European
patent office) in any technological area, serves as a proxy for the propensity to
patent which varies across countries and over time. Fixed effects (
i
) are
introduced to capture unobservable country-specific heterogeneity. All the
residual variation is captured by the error term (
it
), which is assumed to be
normally distributed.
Empirical results for the two technology groups are presented below. The
results for re-design technologies (Table 2.2) indicate that fuel price (ENPRICE)
has a positive and statistically significant (at the 1% level) effect on patenting
activity in engine re-design technologies. Conversely, neither of the
coefficients of the two policy dummies (DUM1_USOBD, DUM2_USOBD) are
statistically significant. The results also suggest that higher patenting activity
in general (PAT_TOTAL), as well as higher industry value added (VA) both yield
more innovations in the specific field of motor vehicle emission control.
Contrary to the above, the estimates given in Table 2.3 suggest that
innovation for post-combustion technologies is primarily driven by
environmental policy shocks. The coefficients of the two dummy variables
(DUM1_USOBD, DUM2_USOBD) are positive and statistically significant at the
1% and 5% levels, respectively. Conversely, fuel prices (ENPRICE) have a
statistically insignificant effect on patenting activity with respect to post-
combustion devices. The results also suggest that, while higher industry value
added (VA) is associated with higher levels of patenting activity in automotive
emission control, this is not the case for patenting activity in general
92
(PAT_TOTAL). This is also consistent with expectations, because regulation-
induced patenting activity may be unrelated to general determinants of
innovation.
In terms of environmental policy design, the results of the two models
would suggest that fuel taxes would be more likely to induce innovation with
respect to engine re-design, while regulatory measures induce post-combustion
innovation. These results are consistent with expectations. In effect, the
Table 2.2. Fixed-effects model estimates for engine re-design technologies
Variable Coefficient Std. error t-statistic Probability
C 5.103 0.219 23.279 0.000
VA 0.006 0.003 2.374 0.021
ENPRICE 1.287 0.249 5.163 0.000
PAT_TOTAL/1000 0.061 0.014 4.309 0.000
DUM1_USOBD 0.067 0.078 0.864 0.391
DUM2_USOBD 0.056 0.077 0.731 0.468
Fixed Effects
_USC 0.096
_DEC 0.476
_JPC 0.573
Observations 72
R-squared 0.920
Adjusted R-squared 0.911
Durbin-Watson 1.239
Notes: Dependent variable is a log of patent count for engine re-design and integrated technologies.
Table 2.3. Fixed-effects model estimates for post-combustion devices
Variable Coefficient Std. error t-statistic Probability
C 2.812 0.381 7.385 0.000
VA 0.019 0.005 3.937 0.000
ENPRICE 0.346 0.466 0.743 0.460
PAT_TOTAL/1000 0.093 0.021 4.490 0.000
Fixed Effects
_USC 0.484
_DEC 0.209
_JPC 0.693
Observations 72
R-squared 0.905
Adjusted R-squared 0.895
Durbin-Watson 1.246
Notes: Dependent variable is a log of patent count for post-combustion devices.
93
development of post-combustion technologies appears to be an activity which is
separate from general innovative activity, while for engine re-design technologies
which reduce pollution, this is not the case.
Given the small size of the sample, as well as the simplicity of the model
being estimated, these results should be interpreted with caution. For instance,
with the potential for lagged impacts of these policies on innovation, it is difficult
to be certain of the precise effect of a specific policy on patent activity.
5. Conclusions
This report has examined the links between environmental policy and
technological innovation, drawing on patent data. The report distinguishes
between: a) different types of policy measures; and b) patents which relate to
post-combustion technologies (such as catalytic converters) and more
fundamental changes (such as engine re-design). While the evidence is
preliminary, some interesting conclusions emerge:
a) Environmental policy shocks do appear to generate innovations. However,
the continuous increase in both policy stringency and patent activity
makes it difficult to disentangle the precise effects of individual policies
with such a small sample;
b) The share of engine re-design patents has been growing much faster than
patents for post-combustion abatement technologies. Whether this is
attributable to environmental policy design warrants further study;
c) Foreign regulatory pressures appear to have an influence on domestic
innovation. For instance, Japanese inventors have played a lead role in the
development of catalytic converter patents, even though the regulatory
shock initially came from the US;
d) There seems to be a distinct difference in the propensity for foreign
inventions to be patented in the three markets. In the US, many of the
patents came from foreign inventors, whereas in Japan, the vast majority of
patents came from Japanese inventors. The role of policy design in explaining
this could also be assessed in further work;
e) Within the engine re-design category, Japanese inventors have been
particularly strong with respect to electronic controls, whereas in Germany
(and in Europe more generally), fuel injection patents were more common.
The US had a strong comparative advantage in air-fuel ratio devices.
f) It appears that fuel prices and general trends in innovation have played a
stronger role in the development of engine re-design technologies than in
post-combustion technologies, while regulatory pressures have been
more important for the latter category.
94
These latter results indicate that more flexible policies which encourage
integrated solutions to environmental problems are more likely to allow for
the realisation of economies of scope between general R&D expenditures on
product design, and those which relate specifically to abatement control.
Notes
1. In the US, Europe and Japan, the contribution of road transport to CO emissions is
61%, 42% and 36%, respectively; the corresponding figures for NO
X
emissions are
38%, 40% and 28% (OECD, 2007).
2. In the US, Europe and Japan, the contribution of road transport to HC emissions is
25%, 20% and 6%, respectively; the corresponding figures for PM emissions in the
US and Europe are 3% and 17% (OECD, 2007).
3. Crabb and Johnson (2007) examined the effects of motor fuel prices on patent
activity in automotive energy-efficient technologies. While this is related to this
Chapter the effects of environmental policy on technologies which are explicitly
emission-reducing are not addressed.
4. Carbon monoxide standards have been included in a separate chart, due to
differences in scale.
5. The NO
X
standards for diesel engines were the same as the standards that applied
to gasoline until 1994. From that time on, diesel standards became a bit more
lenient (1.0 g/mile for diesel versus 0.4 g/mile for gasoline). As of 1994, PM standards
for gasoline have been set at 0.08 g/mile.
6. Note that for 2005, the value of the CO standard (1.15 g/km) is not included in
Figure 2.3, because it is higher than the 0.67 g/km that holds for the period 2000-2004,
implying a break in the time-series. This is due to changes in the test cycle (JAMA,
2005, p. 48).
7. Thus, the HC + NO
X
standard contained in Figure 2.6 is also included in Figure 2.9.
8. On the other hand, prior to reaching the standards that were in place as of 1975,
Japan gradually set more stringent CO, HC and NO
X
standards for gasoline vehicles
in the period 1965-1975.
9. This would bring together the insights from this Chapter and that of Crabb and
Johnson (2007), which focuses on fuel efficiency. Such an approach would allow for
a richer understanding of optimal policy design in the presence of pollutants
which are emitted jointly in combustion.
10. Under perfect combustion, all carbon is converted into CO
2
, leaving no CO behind.
11. An important source of information about these technologies is the various
publications and websites of automobile and parts manufacturers. Searches on
websites of major environmental research institutes, as well as government
(environment agency) websites were also used to develop this section. Faiz et al.
(1996) contains much of the description of technologies that were developed prior
to the first half of the 1990s.
12. www.wipo.int/classifications/ipc/ipc8/?lang=en.
13. Until the late 1980s, each Japanese patent application could only include one claim.
Thus, a patent with four claims in the US would be four separate patents in Japan, for
example. Even today Japanese patents tend to have fewer claims, as some applicants
are still in the habit of having one (or at least just a few) claims on a single patent.
95
14. Note that in the figure patents deposited at the EPO do not include patents filed in
Germany.
15. Many air-fuel ratio devices use electronic control, and as such a specific category
(AIR ELEC) was created, including both air fuel ratio devices in general, as well as
electronic controls.
16. Data was obtained from the IEA Energy Prices and Taxes Database (www.iea.org/
Textbase/stats/index.asp).
17. See also Popp (2002) for a general empirical exploration of the induced innovation
hypothesis in energy-efficient technologies and energy prices.
18. Data was obtained from the OECD Structural Analysis (STAN) database
(www.oecd.org/document/15/0,3343,en_2649_201185_1895503_1_1_1_1,00.html).
References
B&B Electronics (2005), OBD-II Background Information, The OBD-II Home Page
(www.obdii.com/background.html ).
Bresnahan, T.F. and D.A. YAO (1985), The Nonpecuniary Costs of Automobile Emissions
Standards, RAND Journal of Economics 16, pp. 437-455.
Brunnermeier, S.B. and M.A. Cohen (2003), Determinants of Environmental Innovation in
the US Manufacturing Industries, Journal of Environmental Economics and Management
45, pp. 278-293.
Center for Automotive Research (CAR) Publication, April 2004 (www.cargroup.org/pdfs/
AICEFinalReport.PDF ).
Concawe (Conservation of Clean Air and Water in Europe) (1994), Motor Vehicle Emission
Regulations and Fuel Inspections 1994 Update, Report 4/94, Brussels.
Corning, website of components manufacturers. www.corning.com/environmental
technologies/products-applications/.
Crabb, Joseph M. and D.K.N. Johnson (2007), Fueling the Innovation Process: Oil Prices
and Induced Innovation in Automotive Energy-Efficient Technology, Working
Paper, Colorado Department of Economics and Business, May 2007.
Environmental Protection Agency (2002), OBD: Frequently Asked Questions, Office of
Transportation and Air Quality (www.epa.gov/otaq/regs/im/obd/f02014.pdf).
EnvironmentaL Protection Agency (2004), On-Board Diagnostics Basic Information
(www.epa.gov/orcdizux/regs/im/obd/basic.htm).
Environmental Protection Agency (2007), Particulate Matter (www.epa.gov/oar/
particlepollution/).
European Union (1999), Directive 1999/96/EC of the European Parliament and of the
Council, Article 4.
Faiz, A., C.S. Weaver and M.P. Walsh (1996), Air Pollution from Motor Vehicles: Standards
and Technologies for Controlling Emissions, The World Bank, Washington, D.C.
Grard, D. and L.B. Lave (2005), Implementing Technology-Forcing Policies: The 1970
Clean Air Act Amendments and the Introduction of Advanced Automotive
Emissions Controls in the United States, Technological Forecasting and Social
Change 72, pp. 761-778.
96
Griliches, Z. (1990), Patent Statistics as Economic Indicators: A Survey, Journal of
Grsslinger, E., K. Radomsky and M. Ritter (1996), Corinair 1990 Summary Report,
under supervision of the European Environment Agency, Copenhagen.
Jaffe, A.B. and K. Palmer (1997), Environmental Regulation and Innovation: A Panel
Data Study, Review of Economics and Statistics 79, pp. 610-619.
Jaffe, A.B., R.G. Newell and R.N. Stavins (2003), Technological Change and the
Environment, pp. 461-516 in K.-G. Mler and J. R. Vincent (eds.), Handbook of
Environmental Economics, Vol. I, Elsevier, Amsterdam.
Jama (Japan Automobile Manufacturers Association), Annual Reports 2000-2005
(www.jama.org/library/brochures.htm).
Kerr, S and R.G. Newell (2003), Policy-Induced Technology Adoption: Evidence from
the US Lead Phasedown, Journal of Industrial Economics, Vol. 51, No. 3, pp. 317-343.
Lanjouw, J.O. and A. Mody (1996), Innovation and the International Diffusion of
Environmentally Responsive Technology, Research Policy, Vol. 25, pp. 549-571.
Manufacturers of Emission Controls Association (MECA): Publication #: EPA 400-F-92-007,
Fact Sheet OMS-5 (August 1994): US EPA Office of Mobile Source (www.meca.org/
page.ww?name=Who+We+Are&section=Organization+Info )
Matthey, J., Website of components manufacturers (http://ect.jmcatalysts.com/
technologies).
Michaelowa, A. (1997), Phasing out lead in gasoline How developing countries can
learn from the experiences of the industrialized world, in: A. Faircloiugh (ed.),
World Development Aid and Joint Venture Finance 1997/98, pp. 268-272.
OECD (2004), Can Cars Come Clean? Strategies for Low-Emission Vehicles, OECD, Paris.
Popp, D. (2002), Induced Innovation and Energy Prices, American Economic Review,
Vol. 92, No. 1, pp. 160-180.
Popp, D. (2006), International Innovation and Diffusion of Air Pollution Control
X
and SO
2
Regulation in the US, Japan and Germany,
Journal of Environmental Economics and Management, Vol. 51, pp. 46-71.
Von Storch, H., C. Hagner, F. Feser, M. Costa-Cabral, J. Pacyna and S. Kolb (2002),
Curbing the Omnipresence of Lead in the European Environment since the 1970s
A successful example of efficient environmental policy, Working Paper of The
GKSS Lead Project (http://w3g.gkss.de/staff/blei/).
Schwartz, J. (2003), No Way Back: Why Air Pollution Will Continue to Decline, AEI
Press, Washington, D.C.
Taylor, M.R., E.S. Rubin and D.D. Hounshell (2003), Effect of Government Actions on
Technological Innovations for SO
2
Control, Environmental Science and Technology,
Vol. 37, pp. 4527-4534.
Vollebergh, H. (2006), Impacts of Environmental Policy Instruments on Technological
Change, OECD COM/ENV/EPOC/CTPA/CFA(2006)36/FINAL.
Zhu, Y., A. Takeishi and S. Yonekura (2006), The Timing of Technological Innovation:
The Case of Automotive Emission Control in the 1970s, IIR Working Paper WP
No. 076-05, Institute of Innovation Research, Hitotsubashi University, Tokyo,
Japan.
97
ANNEX 2.A1
Overview of Technologies
and Corresponding Patent Classes
A systematic overview is given below of the IPC classifications related to
automotive emission control technologies. For example, class F015/
00 pertains to patents that deal with exhaust or silencing apparatuses.
These are followed by various sub-classifications (N01N5/02 etc.), listed in
descending order of precedence. Before each of the IPC classifications, a short
outline of the main technology is provided.
Exhaust emission control technologies: Improved engine design
or engine redesign
Air-fuel ratio devices
Air-fuel ratio has an important effect on engine power, efficiency, and
emissions. The ratio of air to fuel in the combustion mixture is a crucial design
parameter for spark-ignition engines. A very early emissions control system
the Air Injection Reactor (AIR) reduces the products of incomplete
combustion (HC and CO), by injecting fresh air into the exhaust manifolds of
the engine. In the presence of this oxygen-laden air, further combustion
occurs in the manifold and exhaust pipe. Generally, the air is driven through
an engine-driven smog-pipe and air tubing into the manifolds. This
technology was introduced in 1966 in California, and was in use for several
decades. With cleaner burning engines and better catalytic converters, this
technology is no longer as prevalent. However, having an air-fuel mixture that
has the exact amount of air to burn the fuel without either air or fuel leftover
is stoichiometric and has a normalized ratio of 1. Mixtures with more air
than fuel are therefore lean, while those with more fuel are not. Lean
mixtures are more efficient than stoichiometric ones, and are termed lean-
burn engines.
98
F01N3/05 Mechanical engineering; lighting; heating; blasting engines or
pumps. Machines or engines in general. Gas-flow silencers or
exhaust system apparatus for machines or engines in general;
gas-flow silencers or exhaust apparatus for internal combustion
engines. Exhaust or silencing apparatus having means for
purifying, rendering innocuous, or otherwise treating exhaust.
[N: by means of air e.g. by mixing exhaust with air, in tailpipes.]
F02M67/00 Apparatus in which fuel-injection is effected by means of high-
pressure gas, the gas carrying the fuel into working cylinders of
the engine, e.g. air-injection type [N: fuel-gas mixture being
compressed in a pump for subsequent injection into the engine].
F02M 23/00 Supplying combustion engines in general, with combustible
mixtures or constituents thereof: engine pertinent apparatus for
feeding, or treating before their admission to engine,
combustion-air, fuel, or fuel-air mixture. Apparatus for adding
secondary fuel-air mixture.
F02M 25/00 Supplying combustion engines in general, with combustible
mixtures or constituents thereof (charging such engines F02B).
Engine-pertinent apparatus for feeding, or treating before their
admission to engine, combustion-air, fuel, or fuel-air mixture [N:
treatment by admission of activating fluids]; Engine-pertinent
apparatus for adding non-fuel substances or small quantities of
secondary fuel to combustion-air, main fuel, or fuel-air mixture
F02M43/00 takes precedence; adding secondary air to fuel-air
mixture F02M23/00.
F02 M3/00 Mechanical engineering: combustion engines, carburetors
combined with low-pressure fuel-injection apparatus. Idling
devices, with means for facilitating idling below operational
temperatures.
F02 M3/02 Preventing flow of idling fuel.
F02 M3/04 Under conditions where engine is driven instead of driving, e.g.
driven by vehicle running downhill.
Exhaust Gas Recirculation (EGR) valves
Engines produced after the 1973 model year in the US were fitted with
exhaust gas recirculation valves on the intake manifold. The main purpose of
these valves was to reduce NO
x
emissions, by re-introducing exhaust gases
into the fuel mixture and lowering exhaust temperatures.
99
F01N5/00 Exhaust or silencing apparatus combined or associated with
devices profiting by exhaust energy (using kinetic or wave
energy of exhaust gases in exhaust systems for charging F02B;
predominant aspects of such devices, see the relevant classes
for the devices).
F01N5/02 The devices using heat.
F01N5/04 The devices using kinetic energy.
Oxygen, NO
X
and temperature sensors
Various sensors have been developed to provide feedback and increase
competence of computerized controls. For instance, temperature sensors
were initially used to gauge coolant temperature to send signals of radiator
fluid temperature, to warn and thus prevent overheating. However, modern-
day sensors for temperatures monitors can determine critical factors like air-
fuel mixtures and timing advances. These sensors also need to be more
durable and reliable, as they essentially send signals to a computerized system
of controls, rather than just acting as a warning for the driver. Durable and
precision sensors are critical to on-board diagnostic systems. Similarly,
precision sensors are installed in order to detect levels of oxygen and NOx.
Therefore, these sensors are tailor-made for different makes and models, to
maximize the sensitivity and accuracy of readings.
F02D41/14 Mechanical engineering; lighting; heating; weapons; blasting
engines or pumps. Combustion engines (cyclically operating
valves thereof, lubricating, exhausting, or silencing engines).
Electrical control of supply of combustible mixture or its
constituents. [N: characterised by the type of sensor used.]
Electronic control systems and plasma-based technologies
For easier functioning of stoichiometric engines, electronic control
systems (also called On-board Diagnostic (OBD) systems) were designed to
control air-fuel mixtures. These systems can both measure the air-fuel ratio in
the exhaust, and adjust the air-fuel mixture going into the engine. Japan has
incorporated such computers for such control systems since 1978 and they
have been in use in US engines since 1981. European cars manufactured in the
late 1980s are also fitted with computerized control systems (Faiz et al., 1996).
These computerized versions are also capable of handling the control of other
features like spark timing, exhaust gas recirculation, idle speed, air injection
systems and purging of evaporative canisters. These systems simultaneously
inform the driver of malfunctions, and are more resistant to tampering and
maladjustments than mechanical controls.
100
OBD systems are being installed in all modern vehicles, to monitor the
malfunctioning of various critical processes. These devices are sensitive to
sulphur content in fuels; plasma-based technologies are being introduced to
overcome this drawback. Advances in engine and vehicle technology
continually reduce the amount of pollutants generated, but this is generally
considered to be insufficient to meet emissions goals. Therefore, technologies
that react with (and clean up) the remaining emissions are an important part
of emissions control.
F02D41/00 Electrical control of combustion engines; electrical control of
supply of combustible mixture or its constituents.
F02D41/02 By changing the composition of the exhaust gas, e.g. for
exothermic reaction on exhaust gas treating apparatus.
F02D41/04 Electrical control of supply of combustible mixture or its
constituents; introducing corrections for particular operating
conditions (F02D41/14 take precedence).
F02D41/06 For engine starting or warming-up.
F02D41/10 For acceleration.
F02D41/12 For deceleration.
F02D41/14 Introducing closed-loop corrections.
F02D41/16 For idling.
F02D41/18 By measuring intake air flow measuring flow.
F02D41/20 Output circuits, e.g. for controlling currents in command coils
(current control in inductive loads.
F02D41/22 Safety or indicating devices for abnormal conditions.
F02D41/24 Characterised by the use of digital means.
F02D41/26 Using computer, e.g. microprocessor.
F02D41/28 Interface circuits.
F02D41/30 Controlling fuel injection.
F02D41/32 Of the low pressure type.
F02D41/34 With means for controlling injection timing or duration (ignition
timing F02P 5/00).
F02D41/36 With means for controlling distribution (arrangement of ignition
F02P 7/00).
F02D41/38 Of the high pressure type.
F02D41/40 With means for controlling injection timing or duration.
F02D43/00 Conjoint electrical control of two or more functions, e.g. ignition,
fuel-air mixture, recirculation, supercharging, exhaust-gas
treatment (electrical control of exhaust gas treating apparatus
per se F01N 9/00).
101
F02D43/02 Using only analogue means.
F02D43/04 Using only digital means.
F02D45/00 Electrical control not provided for in groups F02D41/00 to
F02D43/00 (electrical control of exhaust gas treating apparatus
F01N9/00; electrical control of one of the functions; ignition,
lubricating, cooling, starting, intake-heating, see relevant
subclasses for such functions).
F02M51/00 Mechanical engineering; lighting; heating; weapons; blasting
engines or pumps. Combustion engines. Supplying combustion
engines in general, with combustible mixtures or constituents
thereof. Fuel injection apparatus characterised by being
operated electrically.
F01N9/00 Mechanical engineering; lighting; heating; weapons; blasting
engines or pumps. Machines or engines in general; gas flow
silencers or exhaust apparatus for internal combustion engines
[evacuation of the fumes from the area where they are produced
B08B15/00; arrangements in connection with gas exhaust of
propulsion units in vehicles B60K13/00. Electric control of
exhaust gas treating apparatus; [N: electrical control of supply of
combustible mixture or its constituents in relation with the state
of exhaust treating apparatus F02D41/02C4].
Fuel injection systems
Fuel injection systems were first widely marketed by Robert Bosch AG.
These systems injected fuel continuously through nozzles at each intake port.
The rate of injection in turn was controlled by varying the pressure supplied to
the nozzles by an electric fuel pump. These fuel injection systems are now
fully computerized. There are two types of injection systems those with one
or two centrally located fuel injectors; and those with multi-port fuel-injection
systems. Electronic multi-port fuel injection systems either fire all the fuel
injectors at once, or each injector is fired sequentially at the optimal time
during engine rotation. Sequential systems allow for better air-fuel mixtures
and therefore better performance and fewer emissions.
F02M39/00 Mechanical Engineering; Lighting; Heating; Weapons; Blasting
Engines or Pumps. Combustion Engines. Supplying combustion
engines in general, with combustible mixtures or constituents
thereof. Fuel injection apparatus with respect to engines; pump
drives adapted top such arrangements.
F02M39/02 Arrangements of fuel-injection apparatus to facilitate the
driving of pumps.
102
F02M41/00 Fuel-injection apparatus with two or more injectors fed from a
common pressure-source sequentially by means of a distributor.
F02M43/00 Fuel-injection apparatus operating simultaneously on two or
more fuels or on a liquid fuel and another liquid, e.g. the other
liquid being an anti-knock additive.
liquid being an anti-knock additive: Pumps peculiar thereto.
liquid being an anti-knock additive: Injectors peculiar thereto.
F02M45/00 Fuel-injection apparatus characterised by having a cyclic
delivery of specific time/pressure or time/quantity relationship.
F02M47/00 Fuel-injection apparatus operated cyclically with fuel-injection
valves actuated by fluid pressure.
F02M49/00 Fuel-injection apparatus in which injection pumps are driven, or
injectors are actuated, by the pressure in engine working
cylinders, or by impact of engine working piston.
F02M53/00 Fuel-injection apparatus characterised by having heating,
cooling, or thermally-insulating means.
F02M55/00 Fuel-injection apparatus characterised by their fuel conduits or
their venting means.
F02M57/00 Fuel injectors combined or associated with other devices.
F02M59/00 Pumps specially adapted for fuel-injection and not provided for
in groups F02M39/00 to F02M57/00.
F02M61/00 Fuel injection not provided for in groups F02M39/00 to F02M57/00.
F02M69/00 Low-pressure fuel-injection apparatus.
F02M71/00 Combinations of carburetors and low-pressure fuel-injection
apparatus.
Crankcase emissions and control
The compressed gases that blow-by the piston rings in the crankcase
mostly consist of unburned or partly burned hydrocarbons. Prior to
regulations, such blow-by gases were vented into the atmosphere. To control
these emissions, the crankcase vent port requires closing the vent port and
venting the crankcase emissions back into the air-intake system. This is done
by using a check valve.
103
F01M13/04 Mechanical engineering; lighting; heating; weapons; blasting
engines or pumps. Machines or engines in general; engine
plants in general; steam engines. Lubricating of machines or
engines in general. Lubricating internal combustion engines;
crankcase ventilating or breathing [N: having means of purifying
air before leaving crankcase, e.g. removing oil].
Post-combustion devices
Catalytic converters, lean NO
x
catalysts, NO
x
adsorbers,
Catalytic Regeneration Technology (CAT)
A catalytic converter is a device that is placed in the exhaust pipe and is
capable of converting various types of emissions into less harmful ones,
depending on the catalyst used. The catalytic converter comprises a ceramic
support with a wash-coat of (usually) aluminum-oxide, providing a large
surface area. This comprises the substrate, which is then layered with the
catalyst. Substrates are sometimes honeycomb-like structures with
thousands of parallel channels. The walls of these channels provide the
surface for precious-metal catalysts. The catalysts used are made of noble
metals, like platinum, palladium, and rhodium, either singly or in
combination. They convert noxious emissions into carbon dioxide, nitrogen
and water vapor. Catalytic converters have greatly improved over time, and
can be easily retrofitted onto exhaust systems, in order to reduce tailpipe
emissions. An additional beneficial effect on pollution levels was that these
catalysts are rendered inactive by lead. This incompatibility further required
use of lead-free fuel in vehicles fitted with catalytic-converters, leading to a
gradual phasing out of leaded gasoline. In addition, sulphur and phosphorus
in fuels can also poison the catalyst, requiring further cleaning of fuels.
Automobile engines are generally fitted with one of two common types of
catalytic converters. These are the oxidation or two-catalysts and the
oxidation-reduction or three way catalysts.
1. Two-Way Catalysts: In most gas steams, CO and HC can be removed by
combining them with oxygen, using an oxidation catalyst. Oxidation
catalysts use platinum or palladium or a combination of both. These
catalysts increase the reaction rate between oxygen, hydro-carbons and
carbon-monoxide present in the exhaust. Without catalysts, this chemical
reaction is very slow, especially for HCs like methane and ethane in fuels,
which are linked with shorter chains and are therefore more difficult to
oxidize.
2. Three-Way Catalysts and NO
x
-adsorbers: Also known as non-selective
catalytic reduction (NSCR) catalysts, these generally use a combination of
platinum, palladium and rhodium. They are designed to simultaneously
104
convert three pollutants to harmless emissions, namely CO, HC and NO
x
to
carbon-dioxide, water, nitrogen and oxygen, respectively. Thus, these
catalysts not only promote oxidation of HCs and carbon-monoxide, but
when combined with rhodium, help to reduce nitric oxide (NO) to its
harmless components nitrogen and oxygen.
Under lean burn conditions, conventional three-way catalysts are not
effective in reducing nitrous oxides into its components. This drawback had
restricted the use of lean-burn engines in passenger cars. However, lean-burn
engines are still desirable, because they are more fuel-efficient. Recent
research has identified zeolite catalytic materials as being effective in
reducing about 50% of nitrous oxide in lean-burn engines. These catalysts
were first introduced in Japan (Faiz et al., 1996). Later improvements to this
technology (under richer conditions) not only adsorbs the NO
x
with platinum
and palladium, but first releases the NO
x
, and reduces it to nitrogen using the
rhodium.
F01N3/035 Gas-flow silencers or exhaust apparatus for machines or engines
in general; gas-flow silencers or exhaust apparatus for internal
combustion engines. Exhaust or silencing apparatus having
means for purifying, rendering innocuous, or otherwise treating
exhaust. [N: with catalytic reactors.]
F01N3/10 By thermal or catalytic conversion of noxious components of
exhaust by using other chemical processes, chemical aspects of
catalytic conversion, e.g. using specified catalysts.
F01N3/20 Specifically adapted for catalytic conversion; methods of
operation or regulation of catalytic converters.
F01N3/28 Catalytic reactors combined or associated with other devices,
e.g. exhaust silencers or other exhaust purification devices.
B01D53/92 Of engine exhaust gases (exhaust apparatus having means for
purifying or otherwise treating exhaust gases F01N 3/00).
B01D53/94 By catalytic processes.
B01D53/96 Regeneration, reactivation or recycling of reactants.
B01J23/42 Performing operations; transporting. Physical or chemical
processes or apparatus in general. Chemical or physical
processes, e.g. catalysis, colloid chemistry; their relevant
apparatus. Catalysts comprising metals or metal oxides or
hydroxides, not provided for in group B01J21/00 [mixed oxides:
Platinum].
B01J23/44 Catalysts comprising metals or metal oxides or hydroxides, not
provided for in group B01J21/00 [mixed oxides: Palladium].
105
B01J23/46 Catalysts comprising metals or metal oxides or hydroxides, not
provided for in group B01J21/00 [mixed oxides: Rhodium].
Particulate filters and regeneration
F01N3/00 Mechanical engineering; lighting; heating; blasting engines or
pumps. Machines or engines in general. Gas-flow silencers or
exhaust apparatus for machines or engines in general; gas-flow
silencers or exhaust apparatus for internal combustion engines.
Exhaust or silencing apparatus having means for purifying,
rendering innocuous, or otherwise treating exhaust. [N:
collecting or removing exhaust gases of vehicle engines on
highways.]
F01N3/01 By means of electric or electrostatic separators.
F01N3/08 Exhaust or silencing apparatus having means for purifying,
rendering innocuous, or otherwise treating exhaust. [N: for
making innocuous using electric or electrostatic separators.]
Evaporative emission control technologies
Charcoal canisters are not found as a separate IPC class or subclass and are
included in categories covered in tailpipe emissions classifications F01N3/2.
ISBN 978-92-64-04681-8
OECD 2008
107
Chapter 3
Policy Versus Consumer Pressure:
Innovation and Diffusion of Alternative
Bleaching Technologies in the Pulp Industry
by
Popp and Tamara Hafner
(Maxwell School of Public Policy, Syracuse University)*
In the late 1980s and early 1990s, concern over dioxin in both paper
products and wastewater led to the development of techniques that
reduced the use of chlorine in the pulp industry. In this chapter, the
evolution of two completing bleaching technologies in five major
paper-producing countries is reviewed. By the end of the 1990s, nearly
all pulp production in these countries used one of these technologies.
Evidence is provided that substantial innovation occurred before
regulations were in place. Instead, pressure from consumers to reduce
the chlorine content of paper appeared to drive the first round of
innovation. However, while some companies chose to adopt these
technologies in response to consumer pressure, not all firms
differentiated their product in this way.
* The authors would like to thank Nick Johnstone and Ivan Hascic for helpful
comments and assistance in locating data for the project and Dave Halliburton,
Mimi Nameki and Grethe Torrissen for their assistance in clarifying regulations in
their countries.
3. POLICY VERSUS CONSUMER PRESSURE...
108
1. Introduction
This Chapter uses patent data to examine the evolution of elemental
chlorine free (ECF) and totally chlorine free (TCF) technologies used by the
pulp and paper industry. In both cases, the technologies reduce (or eliminate)
the use of chlorine in the bleaching stage of pulp production. Use of these
technologies grew rapidly during the 1990s, beginning in the Nordic countries
(Finland, Norway, and Sweden), then spreading to the US and Canada. One
advantage of studying innovation on these technologies is that they are process
technologies. Most previous studies of environmental innovation using
patents examine end-of-pipe pollution control technologies, because it is easier
to identify patents for specific end-of-pipe technologies than for modifications to
the production process.
1
However, for ECF and TCF technologies, there are well-
defined patent classes related to these processes. Therefore, this study is among
the first to study the evolution of a process environmental technology using
patent data.
Another advantage of studying ECF and TCF technologies is that they
offer a window into the effects of different policy regimes on innovation.
While there is a large theoretical literature on the different effects of various
policy instruments on innovation, few empirical studies compare innovation
under different types of policy incentives.
2
Here, at least three types of policies
are relevant. First, in each country, command and control regulation limits the
amount of chlorine releases from the pulp bleaching process. In the US and
Canada, national (or provincial, in the case of Canada) standards set the basic
limits on chlorine usage. There is some variation across plants, as each plant
operates under an environmental permit. In the Nordic countries, decisions
on allowable emissions are made on a plant-by-plant basis as part of a plant
permitting system. While national guidelines were eventually developed in
Sweden, these play less of a role than the binding regulatory limits which exist
in North America.
In addition to these two regulatory systems, some European mills also
chose to adopt TCF production because of consumer preferences in the
European market (Reinstaller, 2005). Much of the early demand for reductions
in chlorine came from consumers, rather than from regulators. Chlorine used
in bleaching not only affects wastewater released from mills, but also persists
in the final paper product (Galloway, Helminen, and Carter, 1989). Concerns
over chlorine in paper products led to increased demand for chlorine-free
109
paper in the late 1980s and early 1990s. Product labelling requirements allow
consumers to identify paper made without chlorine. A growing literature in
economics looks at the possibility of voluntary provision of environmental
quality by firms (see, for example, Lyon and Maxwell, 2002). One reason often
proposed for such behaviour is that firms are responding to consumer demand.
This Chapter considers the potential effects of such demand-side influences on
environmentally-friendly innovation, and whether these demand-side
influences are magnified by product labelling schemes that make information
about paper quality more readily available to consumers.
This Chapter also looks at the links between environmental regulation
and innovation across countries. In previous work (Popp, 2006), it was found
that innovations for sulphur dioxide and nitrogen oxide control at coal-fired
power plants responds primarily to domestic regulation. Looking at patents in
these fields, it was found that increases in patents assigned to own-country
inventors when a country passes or strengthens environmental regulations
for power plants, but little increase in innovations from other countries. An
important question is whether this finding is robust for other technologies, or
is unique to the electric power industry. One important difference between the
electric power industry and the pulp and paper industry is that there is little
trade in electric power. Regulations affecting air pollution from power
production focus on the location of production. In contrast, while regulations
addressing chlorine in pulp and paper production do focus on the location of
production, the final paper product is a traded commodity, implying that
consumer preferences for chlorine-free paper in trading partner countries
may influence innovation in producer countries.
Finally, this Chapter looks at who innovates. In the case of pollution
control at power plants, innovation is not done by the plants themselves, but
by firms that produce boilers and pollution abatement equipment. In the case
of a new process, such as the ECF and TCF technologies, it would be expected
that more innovation would be done by users of the process, rather than by a
third party firm. This can be studied using assignee data from the patents.
2. The pulp and paper industry
The pulp and paper industry consists of two main types of firms. Pulp
and paper mills process raw wood fibre or recycled fibres to make pulp and
paper. Converting facilities use these primary materials to manufacture
specialised products such as paperboard products, writing paper, and sanitary
paper (EPA, 2002). The focus is on pulp mills. Pulp mills are typically located
near where trees are harvested. Sixty-five per cent of the world pulp market is
supplied by the NORSCAN countries (the US, Canada, Sweden, Finland, and
Norway) (Reinstaller, 2005).
110
Table 3.1 summarises data for the main pulp producing nations in 2000,
sorted by chemical pulp production and percentage of value-added from the
pulp and paper industries.
3
The first panel shows the top producers using
chemical pulp methods. This panel includes countries from around the world.
The second panel shows the top countries based on value added in the pulp,
paper, paper products, and printing and publishing industry. Value-added data
have been obtained from the OECD STAN database, and have two weaknesses.
First, the data only include OECD countries. Second, the value-added data are
for the entire paper industry, not just for pulp production. For example, some
of the top 10 countries in terms of value-added have no chemical pulp
production. Nonetheless, note that four countries (US, Canada, Finland and
Sweden) appear in both panels. In addition, Japan comes close to being on
both lists, as it ranks 12th among OECD nations in percentage of value-added
Table 3.1. Pulp producers
Top countries: pulp production
Chemical pulp production
(1 000 metric tons)
% Value added
from Pulp and Paper
United States 48 198 2.25%
Canada 13 553 2.99%
Japan 9 792 1.73%
Sweden 7 979 3.52%
Finland 7 100 6.06%
Brazil 6 689 N/A
Russian Federation 4 195 N/A
Indonesia 3 626 N/A
Chile 2 220 N/A
France 1 817 1.50%
Top countries: % total value added from pulp, paper, paper products, printing and publishing
Chemical pulp production
(1 000 metric tons)
Percentage
Finland 7 100 6.06%
Ireland 0 3.82%
Sweden 7 979 3.52%
Canada 13 553 2.99%
New Zealand 754 2.43%
United Kingdom 0 2.35%
United States 48 198 2.25%
Austria 1 190 2.02%
Netherlands 0 2.01%
Portugal 1 774 1.84%
Source: Pulp production from FAOSTAT (2006). Value-added percentages are the authors calculations,
based on data from the OECD STAN database.
111
from the pulp and paper industry. As it is also an important source of patents
for ECF and TCF technologies, it is included as well.
4
Because consumer pressure played an important role in the reduction of
chlorine bleaching, it is also important to consider where paper produced in
these countries is sold. Table 3.2 shows the percentage of exports going to
each of the countries in the study, along with Germany, other EU countries,
and the rest of the world. Germany is included separately because, as shown
in the next Section, consumer pressure in that country played an important
role in the diffusion of chlorine-free paper. These data show that much trade
in paper products is regional. Most exports from Sweden and Finland go to
other European countries. Most exports from Canada and the US flow between
the two countries. Japanese paper production is primarily exported to other
countries. Given that consumer pressure varied across the world, its effects
are also likely to vary by region.
2.1. Pulp and paper manufacturing
Paper manufacturing first requires the preparation of cellulose-based
fibres in a dilute slurry of about 0.5% solids. This ensures good separation of
the fibres from one another, and extensive inter-fibre bonding when water is
removed in the paper-making stage. Water is removed when the slurry is fed
Table 3.2. Percentage of exports to each country: paper and paperboard
Exports to:
Exporter Year Canada Finland Japan Sweden USA Germany Other EU Other
Canada 1988 0.0% 2.0% 0.0% 82.3% 1.2% 4.7% 9.8%
Finland 1988 0.9% 2.9% 2.9% 6.3% 12.4% 50.1% 24.5%
Japan 1988 1.5% 0.8% 1.1% 18.2% 4.5% 8.6% 65.3%
Sweden 1988 0.3% 1.7% 0.8% 4.6% 17.1% 58.7% 16.9%
USA 1988 19.3% 0.1% 12.4% 0.4% 2.8% 12.3% 52.9%
Canada 1993 0.0% 2.7% 0.0% 81.9% 0.8% 5.3% 9.3%
Finland 1993 0.5% 2.2% 2.6% 7.7% 14.5% 51.9% 20.6%
Japan 1993 0.7% 0.1% 0.2% 14.8% 1.7% 5.6% 76.8%
Sweden 1993 0.1% 2.2% 0.4% 2.0% 19.3% 57.6% 18.4%
USA 1993 27.5% 0.0% 9.2% 0.3% 2.4% 9.6% 51.0%
Canada 1998 0.0% 1.6% 0.0% 87.1% 0.5% 3.1% 7.6%
Finland 1998 0.8% 1.7% 3.0% 7.3% 16.1% 49.1% 21.9%
Japan 1998 0.8% 0.1% 0.1% 21.4% 1.7% 5.9% 70.0%
Sweden 1998 0.1% 2.5% 0.3% 2.0% 19.5% 60.3% 15.3%
USA 1998 30.5% 0.1% 6.5% 0.1% 2.1% 10.5% 50.2%
Source: Authors calculations, using data from Comtrade (http://comtrade.un.org). Includes exports in
SITC2 categories 641 (Paper and paperboard) and 642 (paper and paperboard, precut and articles of
paper or paperboard).
112
to paper machines, drawing the fibres together, and forming a sheet.
Combinations of suction, pressing and drying are used. Pulp fibres are obtained
from wood, other fibrous materials and recovered paper products. The intended
use of the paper products influences the types of fibre used.
Wood can be turned into fibres by either chemical or mechanical means.
For mechanical pulping, physical force breaks the wood into fine fibres with a
composition similar to the incoming wood. In contrast, chemical pulping
selectively attacks lignin polymers that hold the cellulose fibres together in
the wood. Bleached chemical pulps are often referred to as wood free since
all the lignin has been removed and the fibres have a different composition to
the wood. In contrast, mechanical pulps closely resemble wood owing to their
high lignin content. Certain use of paper requires high brightness and long
life. This is provided by bleached chemical pulps which account for a large
portion of world output. The absence of lignin and the pulps pure cellulose
nature ensures brightness and permanence.
Bright cellulose pulps are made using a two-stage process. The dominant
pulping process is kraft which uses a mixture of sodium sulphide and sodium
hydroxide to selectively dissolve away most of the lignin. Some lignin is
retained following pulping to minimise damage to the cellulose fibres by the
strong pulping solutions. The remaining lignin is removed by milder bleaching
stages. Prior to the discovery of dioxins discovery, treatment with chlorine gas
was the traditional approach used to solubilise most of the remaining lignin.
This was then removed by subsequent sodium hydroxide washing. The small
amounts of lignin remaining after hydroxide washing was removed by
treatment with a substance known as chlorine dioxide ClO2 and further
sodium hydroxide washing. Using C to represent elemental chlorine, D to
represent chlorine dioxide, and E to represent caustic soda, the predominant
five stage bleaching process in used could be denoted as CEDED (Norberg-
Bohm, 1998; Reinstaller, 2005; and EPA, 2002).
5
The dioxin-organochlorine issue emerged in the mid-1980s, following
earlier concerns in the 1970s that lead to measures to address suspended
solids and biochemical oxygen demand. This stimulated installation of
reasonably efficient washing of chemical pulps prior to bleaching, providing
good removal of spent cooking solutions. These were incinerated in recovery
boilers. Incineration destroyed all environmentally damaging materials in
pulping liquors, and allowed for the recovery of valuable pulping chemicals,
providing energy to generate power and heat to operate the mill at the same
time.
Due to the chlorine content of bleach-plant effluents and their
corrosiveness to the recovery systems, bleach-plant wastes remained the
major effluent source in the 1980s. Primary and in most cases secondary
113
treatment had emerged as a means to minimise the environmental impacts of
effluent from the bleach plant and other areas of the mills. The adverse impacts
of bleach-plant wastes were not recognised until advances in chemical analysis
methods and environmental science enabled detection of low concentrations of
pollutants in the wastes. The dioxin and organochlorine of the 1980s stemmed
from such advances.
As such, in seeking to address both the dioxin issue, as well as
organochlorines, many mills initiated pollution prevention based programs
directed at avoidance of both dioxin and organochlorine formation. The
formation of dioxins and furans presented a major product quality concern
owing to risks attributed to use of the products. Ex ante prevention (rather than
ex post treatment) was the only way to alleviate this. In essence, remedy of
dioxin-furan in mill effluents was indirectly addressed by the common
measures applied to prevent formation during manufacturing.
However, the resolution of the organochlorine issue presented a more
fundamental challenge to the sector as knowledge was lacking regarding the
composition and specific effects on biota of organo-chlorine substances in the
bleach-plant effluents. Over time Adsorbable Organic Halide (AOX), emerged
as a frequently applied regulatory measure to address organochlorines.
However the reliability of AOX as a measure of effluent toxicity was unclear.
For example, the test did not differentiate between chlorine attached to low
molecular weight and high molecular weight compounds, which had vastly
different toxicities. Considerable scientific debates arose in the 1990s concerning
the value of AOX as the primary target of regulatory standards (Harrison, 2002).
2.2. Elemental chlorine free bleaching
Elemental chlorine free bleaching replaces the use of chlorine gas by
chlorine dioxide in the first bleaching stage. Its use prevents formation of
chlorinated dioxins and furans and results in a mix of organochlorine
compounds that have lower toxicities than those produced using chlorine
since the different molecular structure of chlorine dioxide greatly reduces
levels of AOX. At many mills technologies such as oxygen bleaching and
extended cooking were added to reduce levels of lignin in pulp prior to bleaching.
Thus, the typical bleaching process changed from OCEDED to ODEDED. This
reduced the amount of chlorine dioxide required and provided lower AOX
discharge per tonne of pulp produced.
2.3. Total chlorine free bleaching
Total chlorine free bleaching differs from ECF in that no chlorine
compounds are used and use of elemental chlorine (such as chlorine dioxide
as well as molecular chlorine as chlorine gases) are displaced. Hydrogen
114
peroxide (H
2
O
2
) and ozone (O
3
) are used as substitute bleaching agents.
Denoting peroxide as P, ozone as Z, and chelation as Q,
6
a typical TCF bleaching
sequence is OQPZP. Discharges of AOX from TCF processes are lower and the only
AOX present in theory would come from trace amounts of organically bound
material naturally present in mill process waters and incoming raw materials.
TCF processing provides the ability to eliminate releases of all but the
naturally present AOX and dioxin furans. However use of the process comes at
the expense of reduced pulp brightness and poorer runnability, owing to
slower drainage of water on the paper machine. Additionally the energy
demand per tonne of TCF pulp made is higher than in the case of ECF and the
yield of pulp per tonne of wood fed is lower. Nonetheless, despite these
disadvantages, TCF technologies were adopted by some firms, particularly in
the Nordic paper producing countries (Reinstaller, 2005).
3. Pollution and the pulp and paper industry
As noted earlier, increased awareness of the links between chlorine and
dioxin led to dramatic reductions in chlorine use by the pulp and paper
industry. One notable feature of this reduction was that consumer demand
played a critical role in both the innovation and adoption of chlorine-free
technologies. A series of environmental concerns led to increased environmental
awareness among consumers. Increasing consumer awareness and activism
regarding the environment spurred governments into action, creating regulatory
pressure on the industry to respond. More importantly, competition among
major industry players to satisfy consumer demand for chlorine-free paper
played a vital role. Companies took competitive advantage of the demand for
environmentally-friendly products, through the use of environmental
labelling. This in turn led to the diffusion of more environmentally-friendly
technologies, even before regulations limiting chlorine bleaching were put in
place, as companies tried to maintain (or increase) their share of the global
market.
Pulp and paper mill effluents have been of particular concern to
environmental activists. Initially the concern was focused on discharges with
high concentrations of biochemical oxygen demanding (BOD) substances
(Burton, 1989). In the 1980s, the emphasis shifted to the release of halogenated
organic compounds in effluent. Chlorinated organic compounds such as
dioxins and furans (which are both recalcitrant and bioaccumulative) are by-
products of the bleaching process when chlorine gas is used as the bleaching
agent. In 1980, the US EPA discovered furans and dioxins in paper mill waste;
in 1983 dioxins were found in fish living downstream from pulp and paper
mills. These studies were first reported to the public by Greenpeace in
August 1987 (Gray, Lowther, and Todd, 1987). In addition, Greenpeace publicised
115
studies finding trace amounts of dioxin in consumer products such as diapers,
milk cartons and coffee filters (Collins, 1992), creating consumer awareness of
the environmental impacts caused by the pulp and paper industry.
Some of the most publicised research on the accumulation and adverse
effects of chlorinated organic compounds in the environment emerged from
Sweden, and to a lesser extent Finland, in the mid-1980s. In Sweden, concerns
about the discharge of chlorinated organic compounds in pulp mill effluents
emerged in the late 1970s and reached very high levels by the mid-1980s
(Galloway, Helminen, and Carter, 1989). In 1982, the Environment/Cellulose I
project was initiated in Sweden to investigate possible links between chlorinated
organics and possible adverse environmental effects. A 1983 report suggested
that chlorinated organic compounds in the effluents of pulp and paper mills
were to be blamed for the declining health of coastal waters. Government-
sponsored scientists did a more comprehensive assessment at the Norrsundet
kraft mill at the Gulf of Bothnia in 1984. Researchers found altered fish
populations with acute skeletal deformities and other adverse effects living in
waters receiving mill discharges (Larsson, Andersson, Frlin, and Hrdig, 1988;
Thulin, Hglund, and Lindesj, 1988). Elevated levels of chlorinated organics
were also found in sediments, but these effects were not found at non-kraft
mills. Finally, in 1987 the discovery of dioxin in diapers prompted a call for the
ban of chlorine in the bleaching of pulp to be used in the manufacture of
disposable diapers (Anonymous, 1987). These findings led regulators to
reassess discharge limits within the context of possible toxicological and
bioaccumulative effects and pressure pulp and paper mills to address the
problems associated with chlorine bleaching (Smith and Rajotte, 2001).
Two other key events occurred in the German market, which was a major
market for paper producers from Sweden and Finland. In 1989, a leading toilet
tissue manufacturer, announced plans to abandon chlorine and ECF pulp
altogether. All its competitors in Germany, Austria and Switzerland took
similar steps within 3 months, switching to TCF or deinked secondary fibres.
Arguably the most influential action occurred in 1991 when Greenpeace
provided information on TCF technology and its benefits, including reply cards
to the publishers of Der Spiegel, requesting that future issues be printed on TCF
paper. This led to many publishers requesting TCF paper from their suppliers
(Smith and Rajotte, 2001).
In 1992, another influential campaign was started when a Swedish firm
started its Z pulp campaign. This campaign publicised the companys
discussions with Greenpeace and embraced the goal of zero discharge. Most
influentially, it adopted the stance that brilliant white paper might be poisoning
its users (Smith and Rajotte, 2001). The importance of firms exploiting a perceived
market niche, has therefore aided in the diffusion of new bleaching technologies
throughout the global industry.
116
3.1. Environmental labelling
Linking chlorine to contamination in everyday consumer products helped
drive consumer demand for paper products produced using chlorine-free
technologies. However, for this demand to have an impact on production
processes, it was important that consumers be informed as to the production
processes used. For this, environmental labelling (ecolabelling) emerged as an
early policy instrument to address increasing consumer demand for
environmentally friendly products. Environmental labelling promotes more
environmentally-friendly consumption for the consumer, in addition to acting as
an economic instrument for the industry, which can tap into a perceived market
niche for green products (Salzman, 1991). Environmental labelling is based
mainly on the impact of the product on environment, and less so on the
impact on the consumer. This type of labelling therefore represented a major
change in product labelling and reflected the emergence of growing awareness
of environmental issues among consumers, as well as their strong desire to
participate in the protection of the environment through their consumption
choices. Regarding pulp and paper products, most labelling schemes have
tend to emphasize recycled fibre content, underscoring the perception that
recycling is the best way to minimise the environmental impacts of paper
products (Webb, 1994). Table 3.3 summarises the major labelling schemes,
along with noting the limits on chlorine content for paper products for each
ecolabel.
Most labelling schemes began in the late 1980s. One prominent exception
is Blue Angel. The earliest labelling scheme, Blue Angel was launched in
Germany in 1978, with 200 labels across 33 product categories (Sammarco,
1997). The Blue Angel scheme did not cover chlorine usage in paper until
February 1992, when a new category for newsprint was introduced.
Many countries adopted similar schemes in the late 1980s, although
requirements on chlorine content for paper typically came later. The
Canadian government introduced Environmental Choice in 1988. Criterion for
paper products were first proposed in 1991, stipulating limits for total absorbable
organohalides (AOX), biological oxygen demand (BOD), and total suspended
solids (TSS) in wastewater discharge, as well as requiring that bleached paper
products not produce measurable concentrations of chlorinated dioxins in the
wastewater or have an effect on rainbow trout (Webb, 1993). However, these
proposed limits were postponed (Webb, 1994), so it was not until the second
iteration of Environmental Choice in 1998 that dioxins were addressed.
7
The Nordic Council launched the Nordic White Swan in 1989, with the
first paper standards beginning in November 1991. Halogenated and aromatic
hydrocarbon cleaning solvents and fluorescent brightening agents were
prohibited in fine papers. While a product could be made from virgin and/or
117
recycled fibres, chlorine-based bleaching agents were banned for recycled
fibres. For pulp and paper production in general, the Nordic Swan limited AOX
releases to 0.5 kg/ton. A 1994 revision of the criteria for the Nordic White Swan
label removed the ban of chlorine bleaching of recycled fibres and increased
the constraint on the use of chemicals with a general ban on the use of
chemicals containing more than 1% of any substance that has been classified
as harmful to the environment by the EU (Webb, 1994). This new limit allows
the use of ECF, rather than TCF bleaching, and had been criticised by
environmental groups as being too lenient. Additionally, Sodr Cell and the
Swedish Society for Nature Conservation have argued that too many products
qualify for the label, thereby decreasing its value in encouraging companies to
improve their environmental performances (Sammarco, 1997). In the UK,
Ecocheck was introduced in 1991 (Webb, 1993), and included limits for chemical
oxygen demand (COD), AOX, BOD, and TSS in wastewater discharge; as well as
sulfur dioxide and nitrogen oxides emission into the atmosphere (Webb, 1996).
The US launched Green Seal in 1989. Green Seal is administered by a
private US environmental labelling agency. Its first paper standards covered
bathroom and facial tissues, and were introduced in 1992. Chlorine bleaching
was allowed until 1996, if the wastewater AOX was below 1 kg/ton pulp, but
banned thereafter. It also stipulated that the tissues be made of 100% waste
paper with 10-20% post-consumer waste content. Standards for printing and
writing paper were established in 1993, which included a complete ban on
chlorine-containing bleaches (Webb, 1994).
Table 3.3. Summary of Ecolabel programs
related to pulp and paper manufacturing
Blue Angel (Germany), begins 1978
1992: new category for newsprint: no halogenated bleaches
Environmental Choice (Canada), begins 1988
1991: limits for AOX
Nordic White Swan, begins 1989
1991: Chlorine bleaching prohibited
1994: Revised to allow ECF
Ecocheck (UK), begins 1991
1991: Included limits for COD, AOX, BOD and TSS in wastewater
Green Seal (US), begins 1989
1992: Bathroom and facial tissue standards limit AOX to 1 kg/ton pulp
1993: Standards for printing and writing paper prohibit chlorine bleaching
1996: Chlorine bleaching prohibited for bathroom and facial tissues
Eco-label (EU), begins 1993
1996: Criteria for copy papers includes pass/fail system in four areas, including AOX
Ecomark (Japan), begins 1989
2004: Chlorine gas not to be used in bleaching process
118
Japan introduced the EcoMark label in 1989. However, limitations on the
use of chlorine were not part of the criteria until 2004, when the use of chlorine
bleaching was not allowed for products receiving this label. Documentation on
the EcoMark label notes that chlorine content were not considered because
dioxin pollution had already been addressed by all the relevant emitters.
8
The Eco-label launched by the EU in mid-1993 has experienced many
difficulties. Forest management criteria were a major sticking point because
they required all wood to be sourced from managed forests following the
criteria established under the Helsinki Guidelines (Webb, 1996). Furthermore,
this label was criticised as being biased towards recycled paper grades because
manufacturers of tissue paper made from 100% virgin fibres were concerned
that their tissues would not meet the AOX limits. At the global level, the Eco-
label was viewed as a form of protectionism against imported virgin pulp
products, and therefore as a trade barrier (Sammarco, 1997). The criteria for
toilet tissues and kitchen towels were finalised in 1994, but the first paper
label was only awarded in December 1995. By mid-1996, agreement was
reached on the criteria for copy papers based on a pass/fail system in four
areas: COD and AOX content in wastewater discharge, sulfur-compounds air
emissions, energy consumption (Webb, 1996).
4. Regulatory responses
Increased awareness of the links between chlorine and dioxin also
spurred governments into action, creating regulatory pressure on the industry
to reduce the use of chlorine in pulp production. In the Nordic countries,
regulation is done on a plant-by-plant basis.
9
In Sweden, permits, which are
reviewed every ten years, are issued by the National Licensing Board for
Environmental Protection. Regulatory authority regarding pulp mill effluents
comes from the Environmental Protection Act of 1969, which emphasized
prevention (instead of control) of pollution. Two key principles played a role in
innovation and adoption of pulp and paper technologies in Sweden. First,
regulators adopted a substitution principle, which urged the industry to
replace harsh chemicals with less benign ones wherever possible. Second,
best available technologies were used as the starting point for approving
permit applications (Smith and Rajotte, 2001). Whereas other countries relied
on the industry to control effluent parameters, Sweden aimed at preventing
discharge (Burton, 1989). This was done through in-process changes such as
adopting the oxygen delignification process to recycle the waste stream. Such
changes gave the industry experience with mill retrofitting and sourcing the
best available technology and made chlorine-free technologies available when
consumer demand for chlorine-free pulp reached its peak in the early 1990s
(Smith and Rajotte, 2001).
119
Because of the use of plant-by-plant licensing, Sweden has not imposed
national discharge standards. However, legislative action has been used to set
national goals for environmental performance. In 1987, the Swedish
Environmental Protection Agency (NV) established goals for organochlorine
substances. In 1992, Parliament established more stringent national goals,
stating that the pulp and paper industry should work to attain no noticeable
effect of effluents by the end of the century (OECD, 1999a). Recommended limits
for AOX releases from kraft pulp mills were just 0.1-0.2 kg/t (OECD, 1999a).
The final requirements of each permit are developed after negotiation
with each plant. The focus is on application of the best available technology
(BAT), which the NV defines as the best technology used on a commercial
scale at a similar plant anywhere in the world (OECD, 1999b, p. 176). By 1990,
ECF was considered to be BAT. The Licensing Board also considers technical,
environmental, and economic factors. Typically, economic considerations
focus on effects on the industry as a whole, rather than on a specific plant.
Under special circumstances, a plant may be given more time to implement
needed upgrades (OECD, 1999b).
In contrast to Sweden, Finland moved more slowly towards chlorine
reduction. Like Sweden, Finland has no regulation specifically limiting AOX
emissions, as it issues permits to plants on a case-by-case basis. In 1988,
Sweden proposed new discharge limits for chlorinated organics for Nordic
states at a meeting of Nordic Ministries (Smith and Rajotte, 2001). Finland
disagreed with these limits, opting to set less stringent targets for the kraft
pulp industry in 1989, limiting AOX releases to 1.4 kg/ADt by 1994.
10
Finland
later accepted more stringent performance targets developed by a Nordic
Working Group in 1993. These targets limited AOX releases to 0.2-0.4 kg/t for
bleached kraft mills, with the more stringent guidelines applying to new mills
(OECD, 1999b).
Unlike the Nordic countries, both the US and Canada have binding
regulations limiting AOX emissions. Individual permits are still needed for
each plant, but the national performance standards must be met or exceeded.
In the US, it was not until the Cluster Rules of 1997 that a binding limit on AOX
releases took effect. Prior to this (1984), the EPA established an ambient water
quality standard for dioxin of 0.013 ppq. However, pulp mills were not covered
as they were not a known source of dioxin at that time. A follow-up study
completed in 1989 confirmed that pulp mills were an important source of
dioxin. The EPA responded by initially requiring pulp mills to meet the 0.013 ppq
ambient standard in their wastewater, although this requirement was later eased
to 1.2 ppq (Norberg-Bohm and Rossi, 1998).
The EPA first proposed regulations for AOX releases in 1993. These standards
could not have been met using existing ECF technology, suggesting that TCF
120
would be considered the BAT in the US. The final rules
11
established ECF as the
BAT, with less stringent AOX limits than the Nordic countries. The monthly
average of AOX emissions from existing sources cannot exceed 0.62 kg/ton
pulp. For new sources, the monthly average cannot exceed 0.27 kg/ton pulp.
Daily discharges cannot exceed 0.95 kg/ton pulp and 0.48 kg/ton pulp,
respectively. In addition, for bleached sulfite mills, both existing and new sources
require TCF technology (Webb, 1998). Mills had until 2001 to comply. To encourage
TCF usage, mills that voluntarily chose to install TCF technology are given an
additional three years to comply (OECD, 1999b).
Similarly, regulation in Canada addressed dioxins in the 1990s. Federal
and provincial governments share responsibility for water pollution control.
The federal role in water pollution control is outlined in the Fisheries Act
of 1970. Pulp and Paper Effluent Regulations were first introduced in 1971. In
May 1992, the federal government introduced new regulations for the pulp
and paper industry, which included new standards for dioxins, but did not
include specific limits for AOX (OECD, 1999b). However, several Provinces have
established limits for AOX emissions.
The first to do so was British Columbia, which in 1990 set a limit of
1.5 AOX kg/ADt, to be met by 1995. The 1990 legislation called for eliminating
AOX emissions by 2002 (OECD, 1999b). However, this was later repealed after
review by a panel of scientific experts. New standards now limit the monthly
average AOX releases to 0.6 kg/ADt.
12
Quebec established AOX limits in 1992;
followed by Ontario in 1993. In both cases, the standards were phased in
gradually, with monthly AOX averages needing to fall to 0.8 kg/ADt by 2000. In
addition, new mills in Quebec face a more stringent standard of 0.25 kg/ADt
(OECD, 1999b).
During the period when dioxin and furan issues unfolded two new bleached
kraft mills were built in Alberta, as well as the modernization and expansion of
two existing bleached kraft mills. The Provincial Government conducted
extensive assessments of technology and environmental requirements, setting
case by case limits in each mills permit. In 2002 its mills had daily AOX limits in
the range of 1 to 3.0 kg/t and monthly limits in the range of 0.55 to 1.5 kg/t.
13
In Japan, regulation was slower to develop than voluntary industry
measures. In 1991, the Japanese pulp and paper industry proposed that AOX
levels be limited to 1.5 kg/metric ton by the end of 1993, and recommended
the use of oxygen delignification equipment and chlorine dioxide substitution
to meet this goal (Management Institute for Environment and Business, 1994).
The first law pertaining to dioxin took effect in 2000, limiting dioxins in
wastewater to 1 pg/l.
14
This was a general regulation applying to all industries.
No specific limits apply to the pulp and paper industry.
15
121
Table 3.4 summarises the key regulations in each country. Finland first
regulated in 1988, and Sweden in 1991. While Canadian provinces also passed
regulations in the early 1990s, these did not take effect immediately. The
standards (0.6-0.8 kg/ADt) were less stringent than the permitting guidelines
used in Sweden and Finland (0.1-0.4 kg/t). While the United States attempted
to establish TCF as best available technology in 1993, the final Cluster Rule
published in 1997 did not include this standard. The monthly average AOX
releases from existing sources must be below 0.62 kg/t for existing sources a
standard comparable to British Columbia. And finally, Japan relied voluntary
compliance until 2000.
5. Data
Using IPC classes, it is possible to identify patents specifically pertaining to
ECF and TCF technology, as well as the source of each patent. One advantage of
looking at ECF and TCF technologies is that there are specific patent classes
pertaining to these technologies. (These are listed in Annex 3.A1.) Within these
classes, the first, D21C 9/14, pertains to ECF production, as it covers the use of
chlorine dioxide. Chlorine dioxide was used in the later stages of bleaching (even
before the switch to ECF technologies), so there will be patents in this even class
before regulations were in place. However, an increase in innovation would be
expected once the shift from elemental chlorine to ECF began. The second class,
Table 3.4. Summary of key regulations
Sweden
1991: Environmental legislation establishes strict guidelines for AOX (0.1-0.2 kg/t).
Enforcement is through plant-by-plant permitting.
Finland
1987: Issues first guidelines for AOX (1.4 kg/ADT), to be met by 1994.
1993: Accepts Nordic Working Group performance standards for AOX (0.2-0.4 kg/t).
Canada
1990: British Columbia sets AOX limits of 1.5 kg/ADt, to be met by 1995. Since lowered to 0.6 kg/ADt.
1992: Quebec passes AOX limits that are phased in gradually. AOX limit of 0.8 kg/ADt by 2000.
New mills limited to 0.25 kg/ADt.
1993: Ontario passes AOX limits that are phased in gradually. AOX limit of 0.8 kg/ADt by 2000.
United States
1993: Proposed Cluster Rule suggests TCF as best available technology. Never took effect.
1997: Revised Cluster Rule limits monthly average AOX releases to 0.62 kg/t pulp for existing sources,
and 0.27 kg/t pulp for new sources. Mills have until 2001 to comply.
Japan
1991: Pulp and paper industry proposes voluntary AOX limit of 1.5 kg/metric ton by end of 1993.
2000: First law limiting dioxins in wastewater (1 pg/l). No specific limit for AOX or for the pulp and paper industry.
122
D21C 9/147, includes bleaching processes using oxygen. These can be either ECF
or TCF. The last two classes, D21C 9/153 and D21C 9/16 cover bleaching using
ozone or per compounds (e.g. hydrogen peroxide). Although these also can be
used in both ECF and TCF bleaching, they are particularly important for TCF, as
these chemicals substitute for chlorine dioxide in the TCF process.
Data was collected for each of the five countries, using the Delphion
patent database.
16
Within each country, patent counts were generated for the
home country of the inventor.
17
The focus was primarily on patents by
domestic inventors, which shows how inventors are reacting to incentives in
that country. Patents were sorted by the priority year.
In the case of Sweden and Finland, inventors might also choose to file
patents through the European Patent Office (EPO), rather than through the
national patent offices. EPO patent applicants may designate as many of the
32 EPO member-states for protection as desired. The application is examined
by the EPO. If granted, the patent is transferred to the individual national
patent offices designated for protection. Because EPO applications are more
expensive, European inventors typically first file a patent application in their
home country, and then apply to the EPO if they desire protection in multiple
European countries. Thus, most Swedish and Finnish inventors will first file
an application in their home country. However, inventors from other European
countries are likely to use an EPO patent if they desire protection in Sweden
and Finland. Thus, data on European Patents that designate Sweden and/or
Finland for protection are also included.
6. Analysis
6.1. Patent trends
Domestic patent applications in each country are first reviewed. This allows
for a comparison of incentives for invention in each of the countries in the study.
Figure 3.1 shows patents granted in each country to domestic inventors, sorted
by the first priority year. What is most notable from this Figure is the role that
perceived public pressure (in response to initial reports of dioxin in waterways)
appears to have played in driving innovation with respect to ECF and TCF
technologies. With the exception of Canada, every country experienced an
increase in ECF and TCF patents that began after release of the Greenpeace report
in 1987. While there was some regulation at this time, the initial regulations were
not very strict. Sweden, the first country to pass stringent AOX guidelines, did so
in 1992. While the US did announce plans for strict regulations that would declare
TCF to be best available technology in 1993, the lack of innovative response from
US inventors after this announcement suggests that this initial proposal was not
perceived as credible (this proposal was eventually withdrawn, and replaced by
the weaker Cluster Rules in 1998).
18
123
The finding that patenting increased before regulations were put in place,
rather than in response to regulation, and that these increases occurred even in
countries that did not pass early regulation, suggests that increased public
scrutiny played an important role in influencing this first wave of
innovation.
19
The American Paper Institute noted as early as November 1988
that several mills where dioxin was detected downstream had begun process
modifications, even though no regulations were in place at the time (Chemical
Week, 1988). Similarly, the discharge limits adopted by the Nordic States
became redundant because green market demand had surpassed those limits
for more stringent measures (Smith and Rajotte, 2001, p. 146).
In addition, industry experts expressed concern over future regulation in
response to increased scrutiny. An April 1988 article in Chemical Week cited a
prediction from an industry source, that the bleaching process would soon be
regulated, and that this would have a heavy impact on purchasing over the
next 5-10 years (Agoos and Portnoy, 1988, p. 45). Another unnamed paper
producer noted at the time that there is likely to be a substantial change in
the way pulp is bleached (Agoos and Portnoy, op. cit.).
A second implication of these trends is that stringent regulation seems to
have spurred additional innovation. Here, it is worth comparing innovation in
the Nordic countries to innovation in the US and Canada. Both Swedish and
Finnish ECF and TCF patents peaked within two years of those countries
passing more stringent guideline limits.
20
By comparison, US and Canadian
mandatory standards were less stringent than the Nordic guidelines, and
Figure 3.1. Domestic ECF and TCF patents by country
The Figure shows patent applications of domestic inventors for each country in our sample. In all cases
except the US, both successful and unsuccessful applications are included. Only patents that were
subsequently granted are included in US applications.
1975
35
30
25
20
15
10
5
0
1977 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2001 2003
SE United States Finland Canada Japan
Number
Priority year
Greenpeace report
SE tightens guidelines
US Cluster Rule
New JP dioxin limits
F1 tightents
standards
124
typically made use of existing technologies. There is little evidence of innovation
occurring in response to US and Canadian regulations. Canadian patents
experience no notable increases; US patenting activity increased dramatically in
response to the initial news about dioxin, but not in response to new regulations.
Neither the proposed (and ultimately defeated) standards of 1993, nor the final
Cluster Rule of l998 led to additional patenting in the US.
Further evidence of the importance of the regulatory framework can be
seen by looking at the types of innovation that were taking place. Table 3.5
provides counts of patents using chlorine dioxide (ClO
2
) and chlorine substitutes
for selected years. Patents using ClO
2
could only be used for ECF processes,
whereas those using chlorine substitutes could be used in either ECF or TCF
processes. While it cannot be ascertained when research has shifted to TCF
processes, it is possible to identify some (but not all) of those cases where ECF was
clearly the main research goal, as reflected in the row labelled ClO
2
in Table 3.5.
While the other category can in theory reflect either ECF or TCF innovations,
in some countries it is likely to be almost exclusively ECF technologies.
21
US patents focused on chlorine substitutes. Sweden had fewer patents at
this time, and many used chlorine dioxide. However, after Sweden established
its strict AOX standards in 1992, nearly all Swedish patents were for technologies
which used chlorine substitutes. This was also true after Finland adopted
stringent standards in 1993. In contrast, not only did US patenting activity fall
after the adoption of the revised Cluster Rule in 1997, but the nature of innovation
shifts. Once the revised Cluster Rule established ECF as acceptable, most
Table 3.5. Number of domestic chlorine and non-chlorine patents,
selected years
Priority Year
1975 1980 1985 1988 1989 1990 1991 1992 1993 1998 2000 2002
Canada ClO
2
2 0 1 0 0 1 0 1 0 1 0 1
Other 1 0 2 1 2 2 2 5 2 3 2 2
Finland ClO
2
0 0 0 1 2 2 0 0 0 1 3 0
Other 0 0 2 0 2 1 3 1 7 3 0 1
Japan ClO
2
3 1 0 4 1 6 2 1 4 3 5 10
Other 2 7 0 8 15 7 14 22 25 11 14 21
Sweden ClO
2
1 2 0 1 3 6 0 0 1 0 2 2
Other 2 4 1 4 1 2 5 11 22 5 4 0
United States ClO
2
1 1 1 1 7 2 2 5 0 5 2 0
Other 3 0 3 4 11 15 27 13 21 0 2 0
The Table shows the number of domestic patent applications for selected years for technologies using
chlorine dioxide (ClO2) and using substitutes for chlorine (other). ClO2 patents correspond to IPC class
D21C 9/14, and other corresponds to the other three IPC classes. Note that ClO2 patents could only be
used in ECF processes, whereas the other patents could be used in both ECF and TCF processes.
125
US patents made use of chlorine dioxide, because there was little incentive for
plants to consider alternatives such as TCF bleaching.
Similarly, a review of who received patents in each country illustrates
differences in the types of innovations occurring across countries. Table 3.6
shows the domestic assignees receiving the most patents in each country,
along with the type of firm. Innovations came from pulp producers, pulping
equipment manufacturers, and from chemical companies. Chemical companies
were more prominent in the Nordic countries than in North America. While the
move to TCF appears to have required innovation from chemical suppliers,
chlorine dioxide had been used in the pulping process even before the move to
ECF. The switch to ECF simply required using more ClO
2
than before. Also of
note is that many of these patents come directly from the pulp and paper
industry.
The range of patent assignees found here contrasts with Popp (2006),
which examined innovations for NO
X
and SO
2
control. There, nearly all
patents came from equipment producers, rather than from the regulated
firms themselves. Plant-by-plant permitting might be expected to induce
more innovation when the regulated firms are the innovators (as they often
are here), as these firms know what they can do when they agree to a permit.
In contrast, outside suppliers should be more comfortable responding nationwide
regulations, since such regulations provide a demand for new (cleaner)
technologies. Unfortunately, it was not possible to pursue this hypothesis
further, since the regulations vary not only by type, but also by stringency
level.
Examination of patent family data provided further evidence of the
leading innovative role of Sweden and Finland. Patent protection is only valid
in the issuing country. To receive protection in multiple countries, an applicant
must obtain a patent in each country for which protection is desired. Additional
fees apply for each application. However, inventors are given a one-year window
after their first filing to decide to patent elsewhere. Only the most valuable
inventions are therefore filed in several countries. Moreover, filing a patent
application in a given country is a signal that the inventor expects the invention
to be profitable in that country. Because of this, researchers such as Lanjouw and
Schankerman (2004) have used data on patent families as proxies for the quality
of individual patents.
Figure 3.2 shows the average family size of domestic patents in each
country by year. Sweden, Finland, and the US have consistently produced the
largest family sizes, whereas most Japanese patents were filed only in Japan.
22
Moreover, family sizes were largest after the initial outcry over dioxin in the
late 1980s. This provides further evidence that global concern, rather than
domestic regulation, drove innovation. Also notable is that the family size of
126
Table 3.6. Top domestic patent assignees
Number % total Org. type
Canada
Pulp and Paper Research Institute of Canada 7 12.5% Pulp/Paper
MacMillan Bloedel Ltd. 6.5 11.6% Pulp/Paper
Individual 5 8.9% N/A
Erco Industries Ltd. 4 7.1% Chemicals
Unassigned 3 5.4% N/A
Finland
Kemira OY 27 35.1% Chemicals
Ahlstrom Machinery OY 23 29.9% Pulp/Paper
Enos-Gutzeit OY 6 7.8% Pulp/Paper
Rauma-Repola OY 3 3.9% Pulp/Paper
Valtion Teknillinen Tutkimuskeskus 3 3.9% Independent research org.
Japan
Mitsubishi Paper Mills Ltd. 51.5 15.7% Pulp/Paper
Oji Paper Co. Ltd. 47 14.3% Pulp/Paper
Mitsubishi Gas Chem Co. Inc. 35 10.7% Chemicals
Nippon Paer Industries Co. Ltd. 27 8.2% Pulp/Paper
Sanyo Chem Ind. Ltd. 10.5 3.2% Chemicals
Sumitomo Heavy Ind Ltd. 10 3.0% Chemicals
New Oji Paper Co. Ltd. 10 3.0% Pulp/Paper
Sweden
Kvaerner Pulping AB 32 22.4% Pulp and chemicals
Sunds Defibrator Industries AB 26 18.2% Pulping equipment
Eka Nobel AB 14 9.8% Chemicals
Mo Och Domsjoe AB 14 9.8% Pulp/Paper
SCA Development AB 7 4.9% Pulp/Paper
Kamyr AB 6 4.2% Pulping equipment
Metos Paper Inc. 5 3.5% Pulp/Paper
Valmet Fibretech AB 5 3.5% Pulping equipment
AGA AB 4 2.8% Chemicals
Stora Kopparbergs Bergslags AB 4 2.8% Pulp/Paper
United States
Union Camp Patent Holding, Inc. 31 15.9% Pulp/Paper
International Paper Company 20 10.3% Pulp/Paper
Wayne State University 13 6.7% University
Kamyr, Inc. 12 6.2% Pulping equipment
Beloit Technologies, Inc. 11.5 5.9% Pulping equipment
Individual 9 4.6% N/A
Air Products and Chemicals, Inc. 8 4.1% Chemicals
Weyerhaeuser Company 7 3.6% Pulp/Paper
Champion International Corporation 5 2.6% Pulp/Paper
E. I. Du Pont de Nemours and Company 4.5 2.3% Chemicals
The BOC Group, Inc. 4 2.1% Chemicals
The Dow Chemical Company 4 2.1% Chemicals
The Table shows the number of domestic patents granted to frequent assignees, by country. For firms,
Org. Type describes the primary industry of the firm. Fractional counts are used for patents with
multiple assignees. % total is the percentage of all domestic ECF and TCF patents assigned to the firm,
including firms not listed above.
127
Swedish and Finnish patents remained high throughout the 1990s. In contrast,
the average family size of US patents falls dramatically after 1996. A likely
explanation for this is the technology-following nature of US regulations. Because
the Nordic regulations were more stringent than those in the US, larger inventive
steps would have been needed to comply with Nordic regulations. US regulations,
in contrast, could be met with existing technology. As such, not only did the level
of innovation fall once it was clear that TCF technology would not be required in
the US, but so did the quality of innovation. Since the mid-1990s, most major
patents for ECF and TCF technologies came from the Nordic countries.
To better illustrate the flows of knowledge across countries, Figure 3.3
shows both domestic and foreign patents in Finland, Sweden, and the US.
Both domestic and foreign regulations appear to have influenced innovation.
For example, patents from US inventors peaked in 1990 in the US. However,
they peak in 1992 in Sweden and 1993 in Finland after passage of tighter
regulations in those countries.
23
Similarly, an increase in Swedish patents can
be observed (both in Sweden and the US) after the 1997 Cluster Rule in the
US.
24
This contrasts with the results of Popp (2006), which found that domestic
regulations were the primary drivers of innovation for air pollution control
devices for coal-fired power plants. One difference here is that the pulp and
paper industry is a global market, whereas most suppliers of pollution
abatement equipment in Popp (2006) were domestic companies.
Figure 3.2. Average patent family size by country and year
The Figure shows the average patent family size of patent applications of domestic inventors for each
country in our sample. In all cases except the US, both successful and unsuccessful applications are
included. Only patents subsequently granted are included in US applications.
16
1985
14
12
10
8
6
4
2
0
1987 1989 1991 1993 1995 1997 1999 2001 2003
Sweden United States Finland Canada Japan
Average family size
Priority year
128
Figure 3.3. ECF and TCF patent trends, selected countries
The Figures show all ECF and TCF patents in Finland, Sweden, and the United States, grouped by
inventor country.
1975
16
25
35
14
12
10
8
6
4
2
0
1977 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2001 2003
1975 1977 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2001 2003
20
15
10
5
0
30
25
20
15
10
5
0
1975 1977 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2001 2003
FI
A. Finland
B. Sweden
C. United States
Patent applications
Patent applications
Patent applications
SE United States Other European Other
Year
SE United States Other European Other
F1 tightens standards
Year
SE tightens guidelines
SE United States Other European Other CA
Year
Failed Cluster Rule
Revised Cluster Rule
Greenpeace report
129
6.2. Adoption of ECF and TCF technologies
Figure 3.3 suggests that innovation on ECF and TCF technologies came
early, in some cases preceding regulations. However, looking at innovation
does not present a complete picture, as it does not consider whether newly
developed technologies are actually put to use. Figures 3.4 and 3.5 show the
percentage of pulp production using ECF technology and ECF or TCF
technology respectively.
25
It is here where the influence of regulation becomes
clearer. By 1994, all pulp production in the Nordic countries used either ECF or
TCF technology. In contrast, North American usage increased more slowly. It
was not until the Cluster Rule deadline of 2001 that nearly 100% adoption was
achieved. One important difference in North America is that public pressure
for chlorine-free paper did not persist as long as it did in Europe. Moreover, the
US industry served primarily a domestic market, exporting just 10% of its
paper products (Norberg-Bohm and Rossi, 1998). While public pressure was
sufficient to jumpstart innovation on ECF and TCF technologies, as well as to
encourage some reductions of chlorine use, the adoption data make clear that
complete diffusion will not occur unless binding regulations are also in place.
Figure 3.4. Diffusion of ECF bleaching technologies
The Figure shows the percentage of chemical pulp production using ECF technology in The Nordic
countries, North America, and the rest of the world (ROW). Diffusion of ECF has been rapid in the
Nordic countries, due to both strong consumer demand and early regulation. In contrast, diffusion in
North America has been more gradual until 2001, the deadline for compliance with the US Cluster
Rules. Source: Authors calculation, using data from Alliance for Environmental Technology (2006).
1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002
100
90
80
70
60
50
40
30
20
10
0
World Nordic countries North America Row
% of pulp production
Year
130
6.3. Public Policy vs. Consumer Pressure
The combination of early innovation and delayed adoption in North
America provides some interesting lessons for the induced innovation literature.
First, the early influence of public pressure, particularly after Greenpeace
published EPA reports on dioxin, is striking. Typically, induced innovation
studies focus on the effect of regulation on innovation. Here, innovation
comes first. Regulation followed, made possible both by public pressure for
action and the availability of alternative technologies for pulp production. The
role of leading countries was also important. Sweden and Finland moved
quickly to reduce chlorine usage. As discussed earlier, these decisions were
made partly because the technologies that had been developed in response to
news on dioxin discharges were deemed acceptable. However, the data suggest
that additional research was needed in order to perfect these technologies, as
both Sweden and Finland experienced an increase in ECF and TCF patenting after
announcing stringent national guidelines in the early 1990s. In contrast, the US
and Canada delayed regulation, and appeared to have developed their regulations
in light of the availability of existing technologies. Early attempts to establish TCF
as the best available technology in the US did not come into force. When the
US finally adopted the Cluster Rule in 1998, ECF had been clearly established
as a viable technology. As such, while the rule served to increase adoption of
ECF technology, no further innovation was needed.
Figure 3.5. Diffusion of ECF and TCF bleaching technologies
The Figure shows the percentage of chemical pulp production using either ECF or TCF technology in
The Nordic countries, North America, and the rest of the world (ROW). Use of elemental chlorine has
been completely eliminated in The Nordic countries, and nearly eliminated in North America since the
US Cluster Rules took effect. Source: Authors calculation using data from Alliance for Environmental
Technology (2006).
1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002
100
90
80
70
60
50
40
30
20
10
0
2003 2004 2005
World Nordic countries North America Row
% of pulp production
Year
131
This differs from the conclusion reached by Popp (2006), which found that
even late adopters of coal-fired power plant regulations needed to innovate to
adapt technologies to local conditions. One difference was that domestic
innovation did occur in the US prior to the Cluster Rule, whereas in the case of
power plants, little domestic innovation occurred before regulations were
enacted. In the case of power plants, even if there had been public pressure to
reduce emissions, consumers only option before regulation would have been
to reduce electricity usage. Alternative, clean suppliers of electricity were not
available, as regulations for nitrogen dioxide and sulfur dioxide took effect
before the movement towards deregulated electricity markets in the late 1990s. In
contrast, some pulp manufacturers did face incentives to reduce chlorine usage
before regulations were in place, because this enabled these manufacturers to
differentiate their product and target environmentally-conscious consumers.
Finally, the importance of early publicity suggests a possible role for
labelling to encourage both innovation and diffusion of ECF and TCF
technologies. However, the first labelling requirement restricting chlorine was
the Nordic Swan in 1991. Most schemes did not address chlorine usage until
later in the 1990s. While they may have played a role in the diffusion of ECF
and TCF technology, labelling schemes appear to incorporate existing
technologies in their criteria, rather than serve as technology-forcing
standards.
26
Given that these labels are voluntary measures, this is not
surprising, as labelling programs do not offer consumer choices unless some
products qualify for the label.
7. Conclusions
This Chapter drew upon patent data to study the development of ECF and
TCF bleaching technologies in the pulp and paper industry across five OECD
countries. While most studies using patent data focus on end-of-the-pipe
solutions to environmental problems, this Chapter examined patenting for a
process technology. As in other studies of environmental innovation,
regulation does seem to play a predominant role in both the development and
diffusion of these technologies. However, it is not the only driver of innovation,
and perhaps not even the most important.
While the data used in this Chapter is insufficient to draw definitive
conclusions, one striking finding from the data presented is the apparent role
of public pressure. While ECF and TCF technologies are process technologies,
they affect the quality of the final product. In the late 1980s, studies linking
chlorine bleaching technologies to dioxin led to pressure from environmentalists
for reduced chlorine bleaching in paper production. Detection of trace amounts of
dioxin in products such as diapers and coffee filters led to increased awareness
of the issue, particularly in Europe. In response, the development of alternative
132
bleaching technologies increased rapidly in each of the countries examined.
Moreover, this increase occurred before new environmental regulations could be
put in place limiting chlorine use, suggesting that public pressure, rather than
regulation, was the primary driver of this first wave of innovation.
Nonetheless, public policy did play an important role. In response to
increased awareness of the problems of chlorine bleaching, both Sweden and
Finland enacted relatively strict regulations in the early 1990s, typically
through plant-by-plant permitting. In both countries, these regulations were
followed by both increased innovation and increased adoption of ECF and TCF
technologies. Moreover, these tighter regulations shaped the nature of
innovation, as firms focused on technologies relevant for TCF production. In
contrast, the US, Canada, and Japan all enacted weaker regulations that could be
satisfied using ECF technology. Moreover, in the US and Japan, these regulations
did not come until later in the decade. As such, patenting in the US did not
remain at high levels (although it did in Japan, presumably in response to
foreign regulations). Moreover, the majority of US patents at the end of the
decade focused on ECF, rather than TCF, technology.
Public policy also plays a role in diffusion. While the patent data show
that the US was an early innovator of ECF and TCF technologies, adoption of
these technologies was slower in the US than in Sweden or Finland. While
some US plants seem to have adopted ECF technology in response to consumer
pressure, it was not until regulations which required its use took effect in 2001
that near universal adoption of ECF or TCF technologies occurred. In contrast,
Sweden and Finland achieved 100 percent diffusion of ECF and TCF
technologies by 1994, due to earlier regulation requiring these technologies.
Finally, while this Chapter illustrates the effect of different policy regimes
on innovation, it says nothing about efficiency. TCF technology was more
costly and produced lower quality paper than ECF. While the more stringent
regulations in Sweden and Finland did hasten the development and diffusion
of TCF technology, it is beyond the scope of this Chapter to assess whether the
additional benefits from completely removing chlorine from the bleaching
process, compared to the partial reduction achieved by ECF, were worth the
additional costs of developing and using TCF technology.
Notes
1. Examples include Popp (2003, 2006), Taylor et al. (2003), and Lanjouw and Mody
(1996).
2. The theoretical literature includes Magat (1978), Milliman and Prince (1989), and
Fisher et al. (2003). These papers predict that market-based environmental policies,
such as a tax or permit trading, will induce more innovation than a comparable
command and control policy.
133
3. Pulp can be produced using chemical or mechanical methods, as described in the
following section. Most production, particularly in developed countries, uses
chemical methods.
4. Despite being a NORSCAN country, Norway appears on neither list, and has few
ECF or TCF patents. Thus, Norway is not included in the analysis.
5. These five stages may be preceded by oxygen delignification, denoted by the letter
O, so that the process is OCEDED.
6. Chelation is the addition of compounds to control the formation of free radicals to
retard decomposition of the hydrogen peroxide (Reinstaller, 2005).
7. Personal communication, Dave Halliburton, Environment Canada.
8. Personal Communication, Mimi Nameki, First Secretary, Permanent Delegation of
Japan to the OECD.
9. In addition, European Union integrated pollution and prevention control (IPPC)
regulations cover European pulp mills. The European Union reached agreement on
IPPC in 1996, and is based on similar legislation passed earlier in the United Kingdom
(Webb 1999). Directives for pulp and paper production took effect in 2001. For
bleached kraft pulp, the new AOX standard is < 0.25 kg/adt. Existing standards in both
Finland and Sweden already satisfy this requirement.
10. ADt represents an air dry ton of pulp product.
11. These are called the Cluster Rules, because the standards address multiple
pollutants, including both air and water, simultaneously.
12. Personal communication, Dave Halliburton, Environment Canada.
13. Personal communication, David Halliburton, Environment Canada.
14. There is no specific standard for AOX in Japan.
15. Personal communication, Mimi Nameki, First Secretary, Permanent Delegation of
Japan to the OECD.
16. This database was preferred over the OECD Triadic Patent Family database since
there would be little variation in the counts for the latter due to the restricted
nature of the technologies examined. However, as with TPF data, the Delphion
data does combine related applications and grants into families to ensure
commensurability.
17. In the case of multiple inventors from different countries, the home country of the
inventor is listed first.
18. In contrast, in the late 1980s, patenting activity for sulfur dioxide control
technologies increased in response to early attempts at modifying the US Clean
Air Act, suggesting that innovators expected that these attempts would eventually
lead to regulatory changes (Taylor et al., 2003).
19. At the same time, the implementation of the Toxic Release Inventory (TRI) in the
US in 1990 led firms to reduce releases of chemicals such as chlorine. However, the
finding that innovation is global, rather than just in the US, suggests that the
importance of consumer pressure was greater than the pressure from TRI. Of
particular importance, TRI data only reports releases from the production process.
It says nothing about the levels of chlorine that remain in paper. Consumer
pressure, particularly in Europe, focused on the chlorine content of paper, rather
than chlorine releases.
20. One interesting possibility is that case-by-case permitting may hasten the
development of new technologies. Bjlls (1999) notes that Swedens approach to
134
permitting encourages innovation by allowing individual plants to propose different
solutions. Moreover, the Swedish National Licensing Board can postpone decisions in
order to investigate new technologies. Conversely, mandatory standards such as
those applied in Canada and the United States may allow for greater certainty with
respect to the potential size of the market for any innovations. Unfortunately
sufficient patent data does not exist to allow testing of whether either of these
factors were important.
21. See, for instance, the review of softwood bleaching practices in Canada by Earl and
Pryke (2003).
22. Japans small family size is explained by Japan being a technology-follower in this
field. Most pulping equipment used in Japan comes from foreign sources
(Management Institute for Environment and Business, 1994).
23. Patents are sorted by their priority date, which is the date of the first patent filing
anywhere in the world. As such, these trends are not simply results of delays in
filing patent applications abroad.
24. Since the peaks are different for each destination country, at least some of the
variation comes from inventors choosing to file more of their patent applications
in other countries. However, the data does allow for conclusions concerning the
extent to which inventors were responding to regulation in other countries by
increasing innovative activity (e.g. creating more inventions) or simply by choosing
to file more of their patent applications in other countries.
25. The chart combines Canada and the US as part of North America, because
separate data are available for these countries only through 2001. Using separate
data reveals similar trends. Canada had slightly higher adoption rates than the US,
due to earlier regulation at the provincial level. The only deviation between the
two countries was that increased North American diffusion in 2001 was entirely
due to the US, which saw usage of ECF technology increase from 76% to 96% as the
deadline for compliance with the Cluster Rule regulations passed (Alliance for
Environmental Technology, 2002). Less than 1% of plants in the US and Canada use
TCF, since it is possible to comply with regulation using ECF technology.
26. Even the influence of labelling schemes to encourage adoption is hard to discern from
the data. Clearly, adoption of ECF and TCF technology increases in The Nordic
countries in the early 1990s, but this is not solely due to labelling, as regulations in
Sweden and Finland were also tightened at this time. However, the Nordic Swan
standards are more stringent than the AOX requirements in these countries, and may
help reduce emissions beyond what is required (personal communication,
Grethe Torrissen, Advisor, Sustainable Production and Consumption/IPP, the
Norwegian Ministry of the Environment). In contrast, the US Green Seal label allowed
chlorine, as long as releases were below 1 kg/ton, until 1996. This may be one
explanation for adoption of ECF technology prior to the Cluster Rule. However,
although Green Seal banned chlorine bleaching after 1996, there was no shift to TCF
technology in the US, with less than 1% of production using this technology (Alliance
for Environmental Technology, 2002).
References
Agnoos, Alice and Kristine PORTNOY (1988), Chemicals Alter Their Roles in
Papermaking, Chemical Week, p. 45.
Alliance for Environmental Technology (2002), Trends in World Bleached Chemical Pulp
Production: 1990-2001.
135
Alliance for Environmental Technology (2006), Trends in World Bleached Chemical Pulp
Production: 1990-2005.
Anonymous (1987), Dioxins in Diapers Stir Sweden, Chemical Week, December 23,
1987, p. 28.
Ratthew Auer, M. (1996), Negotiating Toxic Risks: a Case from the Nordic Countries,
Environmental Politics, Vol. 5, No. 4, pp. 687-699.
Bjlls, Ulf (1999), The Swedish Integrated Permitting System, Environmental
Requirements for Industrial Permitting, Vol. 2, OECD, Paris.
Burton John, (1989), A Global Leader Sweden Sets the Pace in Tackling
Environmental Issues, Financial Times, December 13, 1989, p. IV.
Chemical Week (1988), Paper Industry Fights Dioxin, p. 84.
Collins, Lyndhurst (1992), Environment versus Industry: a Case Study of how the Pulp
and Paper Industry is Responding to Changing Attitudes to the Environment,
Business Strategy and the Environment, Vol. 1, No. 4, pp. 29-36.
Earl, Paul F. and Douglas C. Pryke (2005), Softwood Bleaching Practices in Canada:
Analysis of the 2003 PAPTAC Bleaching Committee Best Practices Survey, paper
presented at the 2005 Engineering, Pulping and Environmental Conference, 2005, /
www.tappi.org/s_tappi/doc_bookstore.asp?CID=5021&DID=526517EPA Office of
Compliance (2002), Profile of The Pulp And Paper Industry, 2nd Edition, Washington, DC.
Faostat Data (2006), www.fao.org/faostat, Accessed May 2, 2006.
Fisher, Carolyn, Ian W.H. Parry, and William A. PISER (2003), Instrument Choice for
Environmental Protection when Environmental Protection is Endogenous, Journal
of Environmental Economics and Management, Vol. 45, pp. 523-345.
Gray, Malcolm, William LOWTHER and David TODD (1987), Alarm over paper goods,
Macleans, October 26, 1987, p. 57.
Griliches, ZVI (1990), Patent Statistics As Economic Indicators: A Survey, Journal of
Harrision, Kathryn (2002), Ideas and Environmental Standard-Setting: A Comparative
Study of Regulation of the Pulp and Paper Industry in Governance: An International
Journal of Policy, Administration and Institutions, Vol. 15, No. 1, pp. 65-96.
Lanjouw, Jean O. and Ashoka MODY (1996), Innovation and the International Diffusion of
Environmentally Responsive Technology, Research Policy, Vol. 25, pp. 549-571.
Lanjouw, Jean O. and Mark Schankerman (2004), Patent Quality and Research
Productivity: Measuring Innovation with Multiple Indicators, Economic Journal,
Vol. 114, No. 495, pp. 441-465.
Larsson, A., T. Andersson, L. Frlin and J. Hrdig (1988), Physiological Disturbances in
Fish Exposed to Bleached Kraft Mill Effluents, Water Science and Technology, Vol. 20,
No. 2, pp. 67-76.
Levin, Richard C., Alvin K. Klevorick, Richard R. Nelson, and Sidney Winter (1987),
Appropriating the Returns from Industrial Research and Development, Brookings
Papers on Economic Activity, Vol. 3, pp. 783-820.
Lockie, Mark (1998), Pulp Producers Hunt for Lower Cost Bleaching, Pulp and Paper
International, Vol. 40, No. 10, pp. 44-45.
Lyon, Thomas P. and John W. Maxwell (2002), Voluntary Approaches to Environmental
Regulation: A Survey, in Maurizio Franzini and Antonio Nicita (eds.), Economic
136
Institutions and Environmental Policy: Post Present and Future, Aldershot, Hampshire,
UK, Ashgate Publishing Ltd.
Magat, A. Wesley (1978), Pollution Control and Technological Advance: a Dynamic Model
of the Firm, Journal of Environmental Economics and Management, Vol. 5, pp. 1-25.
Management Institute for Environment and Business (1994), Competitive Implications of
Environmental Regulation: A Study of Six Industries, US EPA, Washington, DC,
Avai l abl e onl i ne at http: //yosemi te. epa. gov/ee/epa/eerm. nsf /vwSER/
C6E46F0830E7B87 78525644D0053BE5A?OpenDocument.
Milliman, Scott R. and Raymond Prince (1989), Firm Incentives to Promote Technological
Change in Pollution Control, Journal of Environmental Economics and Management,
Vol. 17, pp. 247-165.
Norberg-Bohm, Vicki and Mark Rossi (1998), The Power of Incrementalism:
Environmental Regulation and Technological Change in Pulp and Paper Bleaching in
the US,Technology Analysis and Strategic Management, Vol. 10, No. 2, pp. 225-245.
OECD (2005), Environmental Exposure Assessment: Final Draft for Emission Scenario
Document on Kraft Pulp Mills, document ENV/JM/EEA(2005)4.
OECD (1999a), Environmental Requirements for Industrial Permitting: Case Study on
the Pulp and Paper Sector, Part One, document ENV/EPOC/PPC(99)8/FINAL/PART1.
OECD (1999b), Environmental Requirements for Industrial Permitting: Case Study on
the Pulp and Paper Sector, Part Two, document ENV/EPOC/PPC(99)8/FINAL/PART2.
Popp, David (2006), International Innovation and Diffusion of Air Pollution Control
X
and SO
2
Regulation in the US, Japan, and Germany,
Journal of Environmental Economics and Management, Vol. 51, No. 1, January 2006,
pp. 46-71.
Popp, David (2003), Pollution Control Innovations and the Clean Air Act of 1990,
Journal of Policy Analysis and Management, Vol. 22, pp. 641-660.
Reinstaller, Andreas (2005), Policy Entrepreneurship in The Co-Evolution of Institutions,
Preferences, and Technology. Comparing the Diffusion of Totally Chlorine Free Pulp
Bleaching Technologies in the US And Sweden, Research Policy, Vol. 34, pp. 1366-1384.
Salzman, Jim (1991), Environmental labelling in OECD countries, Paris, France, OECD.
Sammarco, L. (1997), Eco-Labelling Paper and Board Products: The Story so far and
What the Future Holds, Paper and Packaging Analyst, Vol. 29, pp. 37-51.
Smith, Adrian and Alain Rajotte (2001), When Markets Meet Sociopolitics: The
Introduction of Chlorine-Free Bleaching in the Swedish Pulp and Paper Industry,
in Rod Coombs, Ken Green, Albert Richards and Vivien Walsh (Eds.), Technology and
the market. Demand, users and innovation, Cheltenham, UK, Edward Elgar.
Taylor, Margaret, R., Edward S. Rubin, David H. Hounsell (2003), Effect of Government
Actions on Technological Innovation for SO
2
Control, Environmental Science and
Technology, Vol. 37, pp. 4527-4534.
Thulin, Jan, J. and E. Lindesj (1988), Diseases and Parasites of Fish in a Bleached
Kraft Mill Effluent, Water Science and Technology, Vol. 20, No. 2, pp. 179-180.
Webb, Leslie (1999), IPPC Steps up to BAT, Pulp and Paper International, Vol. 41, No. 4,
pp. 29-32.
Webb, Leslie (1998), Wastewater Treatment: Regulations, Bugs and Beds, Pulp and
Paper International, Vol. 40, No. 6, pp. 39-43.
137
Webb, Leslie (1996), Making the Labels Stick on a Complex Issue, Pulp and Paper
International, Vol. 38, No. 12, pp. 31-34.
Webb, Leslie (1994), Eco-labels Stuck on Search for Common Standards,Pulp and
Paper International, Vol. 36, No. 11, pp. 39-42.
Webb, Leslie (1993), How Green is your Labelling Scheme?, Pulp and Paper International,
Vol. 35, No. 5, pp. 32-35.
138
ANNEX 3.A1
Relevant Patent Classes
for Pulp Bleaching Technologies
D21C 9/14 Paper/Paper-Making; Production of Cellulose/Production of
Cellulose by Removing Non-cellulose Substances from
Cellulose-containing Materials; Regeneration of Pulping Liquors;
Apparatus Therefor/After-treatment of cellulose pulp, e.g. of
wood pulp, or cotton liners/Bleaching/with halogens or halogen-
containing compounds/with ClO
2
or chlorites
wood pulp, or cotton liners/Bleaching/with oxygen or its
allotropic modifications (9/16 takes precedence)
wood pulp, or cotton liners/Bleaching/with oxygen or its
allotropic modifications (9/16 takes precedence)/with ozone
wood pulp, or cotton liners/Bleaching/with per compounds
ISBN 978-92-64-04681-8
OECD 2008
139
Chapter 4
Renewable Energy Policies
and Technological Innovation:
Energy Source and Instrument Choice*
by
Nick Johnstone and Ivan Hascic (OECD Environment Directorate)
* The authors wish to thank Piotr Tulej formerly of the International Energy Agency for
the policy data used in this study. Thanks also to David Popp, Frans de Vries,
Paul Lanoie, Jeremy Luchetti, TomJones and Nils Axel Braathen for valuable comments
on a previous version.
This paper examines the effect of environmental policies on
technological innovation in the specific case of renewable energy.
Patent data is collected for wind, solar, wave/tide, biomass,
geothermal, and waste-to-energy technologies. Empirical analysis
is conducted using patent data on a panel of 25 countries over the
period 1978-2003. It is found that in addition to the influence of
market factors (i.e. electricity prices) and expenditures on R&D,
public policy plays a significant role in determining patent
applications. However, different types of policy instruments are
effective for different renewable energy sources.
4. RENEWABLE ENERGY POLICIES AND TECHNOLOGICAL INNOVATION...
140
1. Introduction
In this Chapter, an assessment is made of the effects on innovation of
different policies implemented to favour the development of renewable
energy. The Chapter reports on empirical analysis undertaken with respect to
a wide variety of policy types, using patent counts as an indicator of technological
innovation.
The first Section provides an overview of trends and technologies in the
area of renewable energy, as well as the policies implemented in different
OECD countries in support of its development. In the Second section, data on
patents with respect to renewable energy are presented. In the third Section,
the empirical model and empirical results are discussed. The Chapter concludes
with a brief summary of the main policy implications.
2. The renewable energy sector: trends, technologies and policies
Investment in renewable energy sources wind, solar, geothermal, ocean,
waste-to-energy, and biomass can contribute significantly to the realisation
of public environmental objectives. In addition, it is sometimes argued that
such investment contribute to other public policy objectives, such as increased
energy security, in the face of uncertain markets for fossil fuels.
The penetration of renewables, although increasing, remains limited. In
the absence of public intervention favouring their development, full life-cycle
costs (including development, investment and operating costs) remain higher
than for substitute fossil fuels. Governments are now introducing a wide
variety of policy instruments to seek to accelerate market penetration. As
such, the assessment of the role of public policy in inducing innovation in the
development of renewable energy technologies is an interesting context in
which to assess the effectiveness of different types of measures.
Three generations of renewable energy technologies can be distinguished
(IEA, 2006a):
First-generation technologies which have already reached maturity, such as
hydropower, biomass combustion, and geothermal energy.
Second-generation technologies which are undergoing rapid development
such as solar energy, wind power, and modern forms of bio-energy.
141
Third-generation technologies which are presently in developmental stages
such as concentrating solar power, ocean energy, improved geothermal, and
integrated bio-energy systems.
Figure 4.1 gives the growth rates for different renewable energy sources
amongst OECD countries and for the rest of the world between 1990 and 2004
(IEA, 2006b), compared with the rate of growth of total primary energy supply.
The rapid growth in wind power is evident. However, the relative importance
of slow-growing renewables, such as hydropower and geothermal, means that
the overall rate of growth (1.9%) has only been marginally above the growth
rate of total primary electricity supply (TPES) (1.8%).
In 2004, among the three main regions of the OECD, Europe had the
highest share of renewable use in total energy production (6.9%). In North
America (Canada, US and Mexico) the figure was 5.7%; for OECD Pacific (Japan,
Korea, Australia and New Zealand) it was 3.4%. Biomass was the single largest
source (44.6%), followed by hydropower (34.6%). Figure 4.2 gives a breakdown of
the contribution of different renewable energy sources in the OECD (IEA 2006b).
There are important differences within regions with respect to overall use
and the contribution of different sources. For instance, in 2003, over 15% of
total energy production (minus hydropower) was attributable to renewable
sources in Iceland (principally geothermal), and Denmark (wind). Conversely,
Figure 4.1. Annual growth rates for renewable energy in the world
and the OECD (1990-2004)
Source: IEA (2006b).
30
25
20
15
10
5
0

S
o
l
i
d

b
i
o
m
a
s
s

G
e
o
t
h
e
r
m
a
l

H
y
d
r
o

W
a
s
t
e
/
b
i
o
g
a
s
/
l
i
q
u
i
d

b
i
o
m
a
s
s

S
o
l
a
r

W
i
n
d
T
P
E
S
World OECD
142
in the Czech Republic, France, Hungary, Norway, Poland and Slovakia, less
than 1% of total energy production was attributable to renewable sources.
Table 4.1 gives figures for renewable energy use as a percentage of electricity
consumption for European countries.
As noted above many governments have recently sought to encourage
the further development of renewable energies (IEA, 2004). For instance, a
European Union Directive of 2001 (Directive 2001/77/EC) provides a framework
for the development of renewable energies in Europe. Figure 4.3 gives national
targets for renewable energy use in European countries for 2010 (CEC 2004). In
March 2007 EU Heads of State agreed to set a binding target of using renewable
energy to meet 20% of total EU energy needs by 2020.
To meet these targets a variety of policies have been introduced. These
vary according to the point of incidence, and IEA (2004) distinguishes between
policies which address supply or demand, and whether this targets generation
or capacity in either case. An example of a policy which affects supply and
capacity would be investment tax credits for investment in renewable energy
facilities (e.g. wind turbines). A supply-side policy which targets generation
could be obligations for electricity generators to use a minimum amount of
renewable energy in their fuel mix. Demand-side policies which relate to
capacity would target firms and households as end-users, such as through
sales tax rebates and exemptions or grants on small-scale capital equipment
(e.g. solar panels). And finally, demand-side policies which target generation
would include measures such as public procurement or green energy
pricing.
In addition to differences in terms of the point of incidence of a policy, it
is important to also distinguish policies according to the renewable energy
source(s) concerned, since specific incentives are often provided for specific
renewable types. Indeed, it is relatively rare for policies to be strictly neutral
Figure 4.2. Renewable energy sources in the OECD in 2004
Hydro, 34.6%
Solar/Tidal, 1.0%
Geothermal, 8.9%
Wind, 2.1% Liquid biomass, 3%
Solid biomass, 44.6%
Renewable municipal
waste, 3.2%
Gas from biomass, 2.6%
143
with respect to different types of renewable energy. It is also important to
distinguish between policy types according to the precise nature of the
instrument being applied, with the following criteria being potentially
important determinants of their impacts on innovation:
whether the measure is price-based (i.e. taxes, exemptions, subsidies, etc.)
or quantity-based (obligations, procurement, etc.);
whether the measure is mandatory or voluntary; and
Table 4.1. Share of electricity production from renewable sources
(excluding hydro)* (%) by country
1990 1995 2000 2005
Australia 0.4 0.4 0.5 1.2
Austria 2.3 3.3 2.9 5.5
Belgium 0.4 0.4 0.7 2.1
Canada 0.8 1.0 1.4 1.7
Czech Republic 0.7 0.7 0.9
Denmark 3.1 5.3 16.1 28.1
Finland 10.3 12.3 13.7
France 0.5 0.5 0.6 0.9
Germany 0.3 0.8 2.4 6.9
Greece 0.0 0.1 0.8 2.3
Hungary 0.1 0.2 0.2 4.7
Iceland 6.7 5.8 17.2 19.1
Ireland 0.1 1.4 4.8
Italy 1.5 1.5 2.4 4.2
Japan 1.4 1.5 1.6 1.9
Korea 0.0 0.1 0.0 0.1
Luxembourg 2.1 3.9 11.5 3.4
Mexico 4.1 3.6 3.1 4.2
Netherlands 1.0 1.6 3.2 7.4
New Zealand 8.1 7.7 9.3 9.7
Norway 0.2 0.2 0.2 0.6
Poland 0.0 0.0 0.2 1.1
Portugal 2.5 3.2 3.6 7.6
Slovak Republic 0.1
Spain 0.4 0.8 2.8 8.2
Sweden 1.3 1.7 3.1 5.3
Switzerland 0.8 0.9 1.3 1.8
Turkey 0.1 0.4 0.2 0.1
United Kingdom 0.2 0.6 1.3 3.0
United States 3.0 2.0 1.9 2.2
* Renewable sources include geothermal, solar thermal, solar PV, tide, wind, renewable municipal
waste, solid biomass, liquid biomass and biogas.
Source: IEA/OECD Renewables Statistics and IEA/OECD Energy Balances of OECD Countries.
144
whether a positive incentive is provided for renewables or a negative
incentive is imposed upon fossil fuel substitutes.
The Global Renewable Energy Policies and Measures Database (www.iea.org)
provides data on policies applied in over 100 countries in support of renewable
energy. Some examples of recent significant policy measures introduced in OECD
countries are provided in Table 4.2.
Figure 4.4 provides summary data on the point at which the first
significant example of given type of measure was introduced in a particular
country. This indicates clearly that different policy types have been introduced
with some temporal regularity. First, in the 1970s, a number of countries
introduced support for R&D. This was followed by investment incentives
(third-party financing, investment guarantees), taxes (exemptions, rebates),
and price-based policies (tariffs, guaranteed prices). More recently, a number
of countries have introduced quantity obligations, often followed by certificates
in which the obligations are tradable across generators. Figure 4.5 provides a
graphical representation of the introduction of different policy types in different
countries (IEA, 2004).
3. Patent applications for renewable energy
As with the other studies contained in this report, patents were used as a
measure of innovation. Relevant patents applications have identified using
the International Patent Classification (IPC) codes, developed at the World
Intellectual Property Organisation (www.wipo.org). Based on an extensive
Figure 4.3. Percentage of energy to be provided by renewable energy
90
80
70
60
50
40
30
20
10
0
%

A
u
s
t
r
i
a

B
e
l
g
i
u
m

D
e
n
m
a
r
k

F
i
n
l
a
n
d

F
r
a
n
c
e

G
e
r
m
a
n
y

G
r
e
e
c
e

I
r
e
l
a
n
d

I
t
a
l
y

L
u
x
e
m
b
o
u
r
g

N
e
t
h
e
r
l
a
n
d
s

P
o
r
t
u
g
a
l

S
p
a
i
n

S
w
e
d
e
n

U
n
i
t
e
d

K
i
n
g
d
o
m
145
literature of technology developments in the area of renewable energy, a set of
keywords were identified for this study. These were used to determine
appropriate IPC codes which relate directly to renewable energy in the areas of
wind, solar, geothermal, ocean, biomass, and waste. Table 4.3 gives the relevant
codes and their definitions. (A sample patent for technological development with
respect to wind energy is provided in Annex 4.A1. More information on the
sample patent is available at http://v3.espacenet.com/textdoc?DB=EPODOC&IDX=
WO9404820&F=0&QPN=WO9404820.)
The patent data used in this study were drawn from EPO/OECD PATSTAT
database. The count represents the number of patent applications to the EPO,
based on the priority date and inventor country, using fractional counts.
Table 4.2. Examples of policies aimed at supporting renewable energy
Year Description of policy Type
Belgium 2001 Green Certificates From Jan. 1, 2002 all electricity suppliers have
the obligation to buy a specified amount of green certificates from
electricity producers (6% in 2010).
Tradable certificates
Germany 2004 Renewable Energy Sources Act Prescribes fixed tariffs which grid
operators must pay for renewable energy.
Feed-in tariffs
Germany 2005 Fifth Energy Research Programme Sets the framework for public
RD&D support for the development of renewable energies.
Subsidies for RD&D
Austria 2006 Green Electricity Act 2006 Amendment. Provision of investment
subsidies for new renewable energy power plants.
Investment subsidies
Denmark 2001 New Rules for Payment of Green Electricity Fixed settlement price
per kWh for initial load-hours of energy from wind turbines.
Feed-in tariffs
Japan 2002 Renewable Portfolio Standards Annual obligation on electricity retailer
to use a certain amount of electricity generated from renewable sources.
Can be met in three ways: own generation; purchase from other generators;
and, purchase of certificates.
Obligations
and tradable
certificates
Canada 2006 Sustainable Technology Development Canada Subsidy payments
up to a maximum of 33% of project costs.
Subsidy for RD&D
and investment
United States 2004 Tax Incentives for Renewable Energy (Extension) Tax exemption for
financing of investments in renewable energy technologies.
Tax exemption
for investment
Italy 2000 Tax Credit for Geothermal and Biomass-Fuelled District Heating
Users connected to a geothermal or biomass fuelled district-heating
grid receive a tax credit.
Tax credits
for users
United Kingdom 2002 Renewables Obligation Order. Electricity supply companies are required
to source a percentage of electricity from renewable sources (7.9%
in 2007/2008). Can be met by either surrendering a certificate or paying
a contribution.
Obligations
Netherlands 2003 MEP: Environmental Quality of Electricity Production Subsidy paid to
domestic producers of electricity generated from renewable sources.
Production subsidy
Poland 2000 Obligation for Power Purchase from Renewable Sources: Distribution
companies obliged to provide a certain share of energy produced from
renewable sources (5% in 2008).
Obligations
Source: IEA/JRC Global Renewable Energy Policies and Measures Database (www.iea.org/textbase/pm/grbackground.htm)
accessed Nov. 22, 2007).
146
Counts were obtained for all important patenting countries, including non-
European countries. While the European market is significant, it is still
expected that there will be some bias toward applications from European
inventors (see Dernis and Guellec, 2001). In the empirical analysis undertaken
here this bias is addressed through the inclusion of both fixed effects and data
on total EPO applications to reflect different propensities to patent in general
across countries and through time.
Figure 4.5 gives data for total patent applications for six different renewable
energy sources. Geothermal applications fell off dramatically after the late 1970s,
while there has been continuous growth in patenting for solar power
technologies. Wind power and waste-to-energy exhibited even more rapid
growth, particularly since the mid-1990s. There were very few patents for
wave-tidal energy and geothermal, but they are also now increasing.
1
Figure 4.6 compares total patent applications for a selection of OECD
countries which exhibited significant levels of innovation. Germany had the
highest number of patents, but relative to the US and Japan, this partly reflects
the home bias in EPO applications. France and the UK both have at least
Figure 4.4. Introduction of policies by type for renewable energy
in OECD countries (1973-2003)
Source: An updated version of the table published in IEA (2004) was kindly provided by Piotr Tulej of the
International Energy Agency.
1970 1973 1976 1979 1982 1985 1988 1991 1994 1997 2000 2003
NE ITA
AUS H B SW
A
J
DK
CH
DK A AUS E
H D
L
FI
B
NE
ITA SW
CH
US
P
D
AUS C ITA
J L
US
E
DK P
D UK ITA
L A B F SW H K NE
CH GR IR NO C
CZ
US F T L GR
C NE DK
NO UK
IR ITA
SW
A P E
DK
US ITA T K GR
H
A P NE C UK
NZ
F D CH J SW NO
B L
D B
C NZ
IR
GR K T F L H
FI
US
A
AUS
P
DK
ITA
UK
J
NZ
IR
F
NO
UK
EU
GR
CZ
J
K
NO
NE
CH
SW
E
B CZ K
FI
AUS E
CZ
FI
D
FI K
NO
UK
Research and
Development
Investment
incentives
Taxes
Tariffs
Voluntary
programmes
Obligations
AUS : Australia C : Canada FI : Finland GR : Greece ITA : Italy L : Luxembourg
NO : Norway SW : Sweden UK : United Kingdom A : Austria CZ : Czech Republic
F : France H : Hungary J : Japan NE : Netherlands P : Portugal CH : Switzerland
US : United States
Tradable permits
147
Table 4.3. IPC Classifications for Renewable Energy
Class Sub-Classes
WIND
Wind motors with rotation axis substantially in wind direction F03D 1/00-06
Wind motors with rotation axis substantially at right angle to wind direction F03D 3/00-06
Other wind motors F03D 5/00-06
Controlling wind motors F03D 7/00-06
Adaptations of wind motors for special use; F03D 9/00-02
Details, component parts, or accessories not provided for in, or of interest apart
from, the other groups of this subclass
F03D 11/00-04
Electric propulsion with power supply from force of nature, e.g. sun, wind
B60L 8/00
Effecting propulsion by wind motors driving water-engaging propulsive
elements
B63H 13/00
SOLAR
Devices for producing mechanical power from solar energy F03G 6/00 08
Use of solar heat, e.g. solar heat collectors
F24J 2/00 54
Machine plant or systems using particular sources of energy sun F25B 27/00B
Drying solid materials or objects by processes involving the application of heat
by radiation e.g. sun
F26B 3/28
Semiconductor devices sensitive to infra-red radiation including a panel or
array of photoelectric cells, e.g. solar cells
H01L 31/042
Generators in which light radiation is directly converted into electrical energy H02N 6/00
Aspects of roofing for the collection of energy i.e. Solar panels
E04D 13/18
Electric propulsion with power supply from force of nature, e.g. sun, wind
B60L 8/00
GEOTHERMAL
Other production or use of heat, not derived from combustion using natural or
geothermal heat
F24J 3/00 08
Devices for producing mechanical power from geothermal energy F03G 4/00-06
Electric motors using thermal effects H02N 10/00
WAVE/TIDE
Adaptations of machines or engines for special use characterised by using
wave or tide energy
F03B 13/12-24
Mechanical-power producing mechanisms ocean thermal energy conversion F03G 7/05
Mechanical-power producing mechanisms using pressure differentials or
thermal differences
F03G 7/04
Water wheels F03B 7/00
BIOMASS
Solid fuels based on materials of non-mineral origin animal or vegetable C10L 5/42-44
Engines operating on gaseous fuels from solid fuel e.g. wood
F02B 43/08
Liquid carbonaceous fuels organic compounds C10L 1/14
Anion exchange use of materials, cellulose or wood B01J 41/16
WASTE
Solid fuels based on materials of non-material origin refuse or waste C10L 5/46-48
Machine plant or systems using particular sources of energy waste F25B 27/02
Hot gas or combustion Profiting from waste heat of exhaust gases F02G 5/00-04
Incineration of waste recuperation of heat F23G 5/46
Plants or engines characterised by use of industrial or other waste gases F012K 25/14
Prod. of combustible gases combined with waste heat boilers C10J 3.86
Incinerators or other apparatus consuming waste field organic waste F23G 7/10
Manufacture of fuel cells combined with treatment of residues H01M 8/06
148
200 patent applications over the period. In addition to these countries, there were
specific areas in which individual countries have been important innovators for
specific renewables.
In Table 4.4, the counts are averaged across all years. In addition to
Germany, Japan and the US (countries which are consistently important
for most renewables), other significant innovating countries for particular
sources have included Denmark (wind), Switzerland (solar, geothermal),
France (geothermal, biomass, waste), the UK (ocean, biomass and waste), Italy
Figure 4.5. Number of EPO patent applications by type of renewable
Source: OECD Patent Project (www.oecd.org/document/10/0,2340,en_2649_33703_1901066_1_1_1_1,00.html).
Figure 4.6. Number of EPO patent technologies for renewables by country
Source: OECD Patent Project (www.oecd.org/document/10/0,2340,en_2649_33703_1901066_1_1_1_1,00.html).
180
160
140
120
100
80
60
40
20
0
1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002
Wind
Wave-tide
Solar
Biomass
Geothermal
Waste
1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002
160
140
120
100
80
60
40
20
0
DE US JP GB FR
149
(ocean), Netherlands (wind), and Sweden (ocean). (Countries in the top five for
each renewable are indicated in bold face.)
In Table 4.5, the counts are weighted by the countrys GDP, to yield a
measure of patent intensity, relative to the size of the economy. On this basis,
a number of smaller countries, such as Denmark, Switzerland, Austria, and
Sweden achieve the highest innovation output per unit of GDP. Of the three
countries which have the highest absolute counts, only Germany continues to
rank consistently in the top five. However, Japan and the US remain first and
third among non-EPO countries, with Australia being second.
Table 4.4. Number of EPO patent filings in renewable energy technologies
(annual average 1978-2003, by inventor country)
Wind Solar
Geo-
thermal
Ocean Biomass Waste
All
renewables
1978-2003
Total
AT 0.46 1.19 1.23 0.15 0.27 0.92 4.23 110
AU 0.19 1.88 0.42 0.19 0.12 0.42 3.23 84
BE 0.92 0.50 0.42 0.04 0.15 0.31 2.35 59
BR 0.00 0.00 0.00 0.04 0.12 0.04 0.19 5
CA 0.58 0.54 0.23 0.08 0.12 1.15 2.65 66
CH 0.50 2.08 1.31 0.08 0.15 1.31 5.42 138
DE 13.88 12.81 7.00 0.69 3.54 11.77 49.46 1 285
DK 3.38 0.46 0.19 0.42 0.15 0.73 5.35 137
ES 0.96 0.81 0.08 0.46 0.00 0.08 2.38 61
FI 0.27 0.35 0.12 0.08 0.00 0.50 1.31 34
FR 1.81 1.92 2.85 0.35 1.62 1.88 10.38 267
GB 1.96 1.31 0.92 0.96 5.46 1.92 12.38 322
GR 0.19 0.19 0.00 0.08 0.00 0.08 0.54 14
HU 0.08 0.19 0.19 0.08 0.04 0.00 0.58 15
IE 0.19 0.15 0.00 0.19 0.00 0.00 0.54 14
IT 1.19 1.35 0.92 0.62 0.31 1.31 5.69 148
JP 1.73 7.19 1.73 0.42 0.77 13.42 25.27 656
KR 0.31 0.04 0.00 0.04 0.04 0.15 0.58 15
NL 2.00 1.58 0.96 0.19 0.35 1.19 6.19 161
NO 0.31 0.27 0.19 0.46 0.04 0.12 1.38 36
NZ 0.04 0.00 0.04 0.04 0.04 0.12 0.27 7
PL 0.04 0.08 0.08 0.00 0.04 0.04 0.27 7
PT 0.19 0.19 0.00 0.04 0.00 0.04 0.46 12
SE 1.35 0.62 1.12 0.62 0.15 0.42 4.27 109
TW 0.19 0.15 0.04 0.04 0.00 0.19 0.62 16
US 3.77 5.88 3.69 2.04 8.65 11.65 35.73 925
Total 942 1 079 616 216 566 1 285 4 702
Note: The Table provides the annual mean number of patent filings during 1978-2003, classified by
inventor country. The last column (row) gives the total number of patent filings in renewables in a
given country (technology).
150
As noted in the introductory Chapter of this volume, there can be differences
in the propensity to patent across countries. In order to control for this effect,
Table 4.6 normalises the patent counts with respect to renewable energy by total
EPO patents for all technological fields. In this case it is Denmark, Norway, Spain,
Australia and Austria which have the highest number of patents. The big three
have no particular revealed comparative advantage in renewable energy
patenting.
And finally, in order to address the concern that the inventive activity of
non-European countries may be biased downward, due to the fact that the
data used here are based on patent filings with the EPO, Figure 4.7 shows the
European and non-European countries separately.
Table 4.5. Number of EPO patent filings in renewable energy technologies
(annual average 1978-2003, per unit of GDP, by inventor country)
Wind Solar
Geo-
thermal
Ocean Biomass Waste
All
renewables
1978-2003
Total
AT 2.50 6.20 8.14 0.67 1.40 4.85 23.76 110
AU 0.41 4.41 1.48 0.43 0.38 0.78 7.89 84
BE 3.96 2.12 2.29 0.21 0.58 1.22 10.39 59
BR 0.00 0.00 0.00 0.00 0.00 0.00 0.00 5
CA 0.75 0.76 0.44 0.12 0.15 1.52 3.68 66
CH 2.72 10.36 7.96 0.45 0.81 6.83 29.14 138
DE 7.20 6.96 4.97 0.42 1.87 6.29 27.59 1 285
DK 23.10 3.28 1.89 2.99 1.48 5.81 38.56 137
ES 1.27 1.08 0.14 0.68 0.00 0.11 3.28 61
FI 2.63 3.10 1.35 0.56 0.00 4.27 11.90 34
FR 1.44 1.54 2.84 0.29 1.31 1.41 8.81 267
GB 1.58 1.06 1.00 0.84 4.30 1.53 10.18 322
GR 1.20 1.07 0.00 0.45 0.00 0.53 3.24 14
HU 0.39 1.95 0.37 0.00 0.39 0.00 3.11 15
IE 3.97 2.14 0.00 2.43 0.00 0.00 8.54 14
IT 0.91 1.03 0.89 0.49 0.23 0.98 4.52 148
JP 0.54 2.48 0.72 0.15 0.27 4.33 8.49 656
KR 0.44 0.05 0.00 0.07 0.05 0.21 0.81 15
NL 5.40 3.97 3.27 0.53 1.02 3.09 17.09 161
NO 2.12 1.99 1.76 3.41 0.25 0.70 10.24 36
NZ 0.54 0.00 0.54 0.60 0.60 1.45 3.73 7
PL 0.10 0.23 0.00 0.00 0.10 0.13 0.56 7
PT 1.59 1.17 0.00 0.22 0.00 0.26 3.25 12
SE 7.10 3.36 6.55 3.00 0.79 2.19 22.99 109
TW 0.00 0.00 0.00 0.00 0.00 0.00 0.00 16
US 0.53 0.77 0.64 0.31 1.15 1.43 4.84 925
Total 942 1 079 616 216 566 1 285 4 702
Note: The Table provides the annual mean number of patent filings during 1978-2003, classified by
inventor country, and weighted by countrys GDP (in trillions of US dollars, using PPP and 2000 prices).
151
Table 4.6. Number of EPO patent applications in renewable energy
technologies, normalised by overall patenting activity (1978-2003)
Wind Solar
Geo-
thermal
Ocean Biomass Waste
All
renewables
AT 0.67 1.75 1.76 0.22 0.39 1.33 6.13
AU 0.39 3.75 0.86 0.42 0.25 0.84 6.50
BE 1.31 0.69 0.61 0.06 0.17 0.44 3.28
CA 0.73 0.68 0.30 0.08 0.16 1.38 3.34
CH 0.29 1.15 0.75 0.03 0.08 0.74 3.03
DE 1.10 1.01 0.55 0.05 0.27 0.93 3.91
DK 7.65 1.03 0.44 0.92 0.35 1.64 12.04
ES 2.62 2.29 0.24 1.31 0.00 0.24 6.70
FI 0.47 0.60 0.20 0.13 0.00 0.85 2.25
FR 0.37 0.39 0.60 0.07 0.33 0.39 2.15
GB 0.51 0.35 0.24 0.25 1.46 0.50 3.32
IT 0.53 0.61 0.42 0.28 0.14 0.59 2.57
JP 0.16 0.65 0.16 0.04 0.07 1.21 2.29
KR 0.62 0.08 0.00 0.08 0.08 0.31 1.16
NL 1.11 0.88 0.55 0.10 0.20 0.68 3.52
NO 1.68 1.41 1.01 2.39 0.20 0.61 7.31
SE 1.05 0.49 0.90 0.49 0.11 0.31 3.36
TW 1.48 1.19 0.30 0.30 0.00 1.48 4.75
US 0.21 0.33 0.21 0.11 0.48 0.66 2.01
Figure 4.7. EPO patent filings in renewable energy technologies
(annual mean, per unit of GDP)
Note: The figure provides the annual mean number of patent filings during 1978-2003, classified by
inventor country, and weighted by countrys GDP (in trillions of US dollars, using PPP and 2000 prices).
40
35
30
25
20
15
10
5
0
DK CH DE AT SE NL FI BE NO GB FR IE IT ES PT GR HU PL
10
5
0
JP AU US NZ CA KR TWBR
152
Figure 4.8. Relationship between point of introduction of policies
and patent counts
1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002
1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002
1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002
150
125
100
75
50
25
0
20
16
18
12
14
8
10
4
2
6
0
100
80
60
40
20
0
Germany
Denmark
Japan
R-D INV TAR
INV TAR
VOL
TAX OBLIG
TAX OBLIG
INV
VOL
OBLIG
R-D
153
Clearly, natural conditions play a role in explaining national variations in
patent activity by renewable type in many cases, but this is not always evident.
Other factors are also at play. In order to get a first indication of the relative
importance of public policy factors on patenting, a comparison of total
renewable energy patent applications and the introduction of specific policy
tools is provided in Figure 4.8 for Germany, Denmark and Japan. There is no
obvious correlation between the introduction of different policies and spikes
in patent activity, except perhaps the introduction of tariffs in Germany (with
some lag), obligations and taxes in Denmark, and investment credits in Japan.
4. Empirical analysis
Using the patent count data obtained it is possible to assess the
determinants of patenting activity for renewable energy in a more formal
manner. In total, a panel of data for 26 countries and 26 years is available, but the
presence of missing observations for some of the variables (often R&D data)
reduces the size of the samples to between 400 and 500 in the models estimated.
Explanatory variables included in the model were the following:
4.1. Expenditures on R&D
Patent activity is clearly a result of scientific capacity. While it is difficult to
get a precise measure of national capacity for innovation, data was obtained from
the IEA on national public sector expenditures on R&D disaggregated by type of
renewable energy (IEA, Energy Technology Research and Development Database, 2006,
http://data.iea.org.) The sign on this variable is expected to be positive.
4.2. Electricity consumption
Returns on innovation are affected by the potential market for this
innovation. In the case of renewable energy, this is best reflected in trends for
electricity consumption. A growing market for electricity should increase
incentives to innovate with respect to renewable energy technologies. This
data was obtained from the IEA (IEA, Energy Balances of OECD Countries, 2006,
http://data.iea.org.).
4.3. Price of electricity
The commercial viability of renewable energy is dependent in large part on
the price of electricity. Since electricity production costs are often greater than for
fossil fuels, an increase in the price of electricity should increase incentives for
innovation in the area of renewable energies. Since renewable sources represent
a relatively small proportion of total electricity generation, it is assumed that this
is exogenous. This data was obtained from the IEA (IEA, Energy Prices and Taxes
Data Service, 2006, http://data.iea.org.), weighting price indices for residential and
industrial use by consumption levels. The sign is expected to be positive.
154
4.4. EPO applications
There will be a changing propensity to patent within a country across
time, both because different strategies may be adopted to capture the rents
from innovation, and because legal conditions may change through time. In
addition, there will be different propensities across countries to patent at a
specific office, due, for instance, to issues such as the home bias that was
discussed earlier. As such a variable was included reflecting overall EPO patent
applications in all areas (OECD, Triadic Patent Family Database, 2006).
2
The sign
is expected to be positive.
4.5. Public policies
Binary variables were also created for different policy types, such as
R&D support, taxes, investment incentives, differentiated tariffs, voluntary
programmes, obligations, and tradable certificates (see IEA 2004). This variable
takes on a value of 0 prior to introduction of the policy, and 1 thereafter. The
policies reflect the first introduction of significant policy measures as
determined by the delegates to the IEA Working Party on Renewable Energy.
While there are likely to be important differences between instruments in
terms of the stringency of the measures introduced, this shortcoming is
unavoidable for any cross-comparative analysis in which multiple instruments
are included.
3
Table 4.7 provides basic descriptive statistics for the explanatory
variables.
A negative binomial model was estimated to evaluate the effects of
environmental policy and market-related determinants on patenting activity in
renewable energy technologies. A dataset consisting of a panel of 26 countries
over the period 1978-2003 was used to estimate the regression models. The full
model is specified as:
where i = 1,,26 indexes the cross-sectional unit (country) and t = 1978, ,
2003 indexes time. In different specifications a variety of lag structures were
imposed to reflect the dynamic nature of innovation. This is further discussed
below.
The dependent variable, patenting activity, is measured by the number of
successful and unsuccessful patent applications in each of the technological
areas of renewable energy (wind, solar, geothermal, ocean, biomass, and waste).
Fixed effects (
i
) are introduced to capture unobservable country-specific
heterogeneity. All the residual variation is captured by the error term (
i
) , with
exp(
i
) assumed to be gamma-distributed with mean 0 and variance 1.
4
1 2 3
4 5
ln( ) . . & .
. .
it it it it
it it i it
PatentCount PRICE R D CONS
EPO POLICY

= + +
+ + + +
[1]
155
Several alternative specifications of the model were estimated. Table 4.8
presents the estimation results when all policy dummies are included in the
regressions, except the dummy for R&D programmes (due to correlation with the
intercept).
5
The coefficient of the electricity price has a positive sign in every
equation. It is statistically significant at the 1% and 5% levels in the solar and
biomass equations, respectively. This suggests that higher electricity prices
provide an incentive for increased patenting activity in the solar and biomass
technologies. The results also suggest that technology-specific R&D spending
is a significant determinant of patenting in renewable energy overall, and
especially in the case of wind and ocean technologies. The estimated coefficient
of electricity consumption is statistically significant only in the waste-to-
energy equation. The estimated coefficient of the total number of EPO filings
is statistically significant at the 1% level in every technological area, suggesting
that a part of the variation in patenting activity in renewable energies is due to
changes in the propensity to patent.
6
The results on the policy dummies suggest that public policy plays a
significant role in inducing innovations in renewable energies. For wind
technology and for renewable energies overall, tax measures, obligations, and
tradable certificates are statistically significant (at the 5% level and higher)
policy instruments. However, the efficacy of alternative policy instruments in
inducing innovations varies by the technological area. For example, providing
investment incentives is a statistically significant policy for solar energy
innovations. Investment incentives, as well as voluntary programmes, are
Table 4.7. Descriptive statistics of explanatory variables (1978-2003)
Variable Obs. Mean Std. dev.
Electricity price (US$/unit, using PPP) 583 0.0849 0.0345
Electricity consumption (millions GWh) 624 0.0158 0.0323
Total EPO patent filings (thousands) 673 2.3964 4.9912
Technology-specific R&D expenditures
Wind R&D (10E + 9 USD, 2005 prices and PPP) 478 0.0063 0.0140
Solar R&D (10E + 9 USD, 2005 prices and PPP) 479 0.0237 0.0702
Ocean R&D (10E + 9 USD, 2005 prices and PPP) 477 0.0016 0.0077
Bioenergy R&D (10E + 9 USD, 2005 prices and PPP) 478 0.0086 0.0157
Renewables R&D (10E + 9 USD, 2005 prices and PPP) 482 0.0481 0.1261
Policy dummies
R&D support 676 0.8432 0.3639
Investment incentives 676 0.4127 0.4927
Tax measures 676 0.2722 0.4454
Tariffs 676 0.3151 0.4649
Voluntary programmes 676 0.1050 0.3068
Obligations 676 0.2130 0.4097
Tradable certificates 676 0.0577 0.2333
156
statistically significant policy instruments for waste-to-energy incineration.
Finally, putting in place preferential tariff structures, and to a lesser extent tax
measures, are statistically significant policies with respect to biomass energy.
There are two concerns related to including all policy dummies in the
regressions. First, correlation among the dummies may cause multicollinearity
problems. In particular, dummies representing investment incentives, tax
measures, and tariffs are highly correlated Table 4.9).
7
Similarly, obligations and
tradable permits are highly correlated.
8
Second, it is possible that there are
interaction effects among alternative policy instruments (e.g. investment
Table 4.8. Estimated coefficients of the negative binomial fixed effects models
with individual policy variables
Wind Solar Ocean Biomass Waste All renewables
Electricity price 3.187 18.718
**
2.181 14.769
*
2.957 0.994
(0.488) (0.000) (0.737) (0.035) (0.469) (0.683)
Specific R&D expenditures 17.789
**
0.966 13.889
*
7.473
*
0.479 1.063
**
(0.000) (0.153) (0.038) (0.043) (0.249) (0.000)
Electricity consumption 9.630 8.060 15.200 15.900 13.600
**
5.030
(0.123) (0.141) (0.335) (0.115) (0.005) (0.162)
Total EPO filings 0.106
**
0.074
**
0.069 0.121
**
0.122
**
0.081
**
(0.001) (0.000) (0.188) (0.001) (0.000) (0.000)
Policy dummies
Investment incentives 0.214 0.626
**
0.097 0.176 0.723
**
0.145
(0.292) (0.000) (0.740) (0.481) (0.000) (0.146)
Tax measures 0.371
*
0.021 0.538 0.500
*
0.083 0.235
*
(0.040) (0.881) (0.089) (0.050) (0.578) (0.017)
Tariffs 0.434 0.116 0.015 0.783
**
0.192 0.043
(0.053) (0.547) (0.964) (0.000) (0.336) (0.717)
Voluntary programmes 0.089 0.020 0.066 0.240 0.334
*
0.119
(0.718) (0.898) (0.863) (0.307) (0.043) (0.318)
Obligations 1.157
**
0.181 0.472 0.212 0.045 0.384
**
(0.000) (0.214) (0.155) (0.372) (0.761) (0.001)
Tradable certificates 0.485
*
0.064 0.192 0.081 0.245 0.305
*
(0.034) (0.718) (0.597) (0.798) (0.159) (0.016)
Intercept 0.214 0.267 15.394 1.012 0.372 0.995
**
(0.598) (0.685) (0.992) (0.371) (0.509) (0.000)
Number of observations 452 427 450 334 441 463
Log-likelihood 477.65 488.20 238.56 289.98 482.30 926.40
Wald chi2 250.63 209.21 33.22 80.03 337.01 398.90
(Prob > chi2) (0.000) (0.000) (0.000) (0.000) (0.000) (0.000)
Notes: * and ** refer to 5% and 1% level of statistical significance. P-values are in parentheses. The dependent variable
is the natural log of patent counts (successful and unsuccessful applications) in a given technological area. Intercept
represents the average value of the country-specific fixed effects. Results for geothermal energy are not reported
because they represent a significant outlier. The R&D dummy is not included in the regressions because it is (almost
perfectly) correlated with the intercept (fixed effects). This is due to the fact that most countries introduced R&D
support programmes as early as 1970s, but the data here covers only the period after 1978.
157
incentives for capital goods may be accompanied by preferential tax rates for
final goods). In an alternative specification, the individual policy dummies
were included one-by-one in the regressions. The results (not reported)
suggest that the key qualitative findings remain unaffected. Policies which are
found to be statistically significant when all dummies are included in the
regression remain statistically significant, and with the same signs, when they
are included separately. Additional policy dummies are statistically significant
when included separately.
Thus, including all policy dummies may cause multicollinearity. However,
including policy dummies one-by-one may lead to incorrect conclusions (due to
omitted variables), and possible interaction effects among the different policies.
In order to address these issues: a) a composite policy variable was constructing
representing the number of policies in place; and b) clusters of policy variables
were developed by clustering similar policies in groups.
A composite policy variable is an index that reflects differences in the
richness of policy approaches across countries and over time. It is constructed
by summing across all seven policy dummies. It thus represents the (cumulative)
number of policies in place (including the R&D dummy in this case). The
disadvantage of this approach is that it does not distinguish among the individual
policy instruments. The advantage is that it implicitly deals with the correlation
and potential interaction effects among policy instruments. In addition, the
composite policy variable can be lagged, allowing analysis of dynamic issues.
Table 4.10 presents the estimation results using the composite policy
variable. Overall, the results remain robust. The estimated coefficient of the
policy variable is positive and statistically significant at the 5% level in every
equation. This suggests that public policy is a significant driver of innovative
activity in renewable energy overall, as well as in the specific technological
areas. Using the same model, the dynamic aspects of innovative activity were
also analysed by lagging the composite policy variable. Including the current
policy variable or its 1-year, 2-year, or 3-year lag separately yields qualitatively
identical results. When the policy variable and its first three lagged values are all
included in a regression, the current policy variable is positive and statistically
Table 4.9. Correlation coefficients between policy variables
Investment
incentives
Tax
measures
Tariffs
Voluntary
programmes
Obligations
Tax measures 0.572
Tariffs 0.531 0.524
Voluntary programmes 0.368 0.222 0.347
Obligations 0.442 0.369 0.470 0.343
Tradable certificates 0.283 0.264 0.205 0.197 0.484
158
significant,
9
while the lags are insignificant.
10
Finally, there could be a concern
that variables such as electricity consumption, EPO patent filings, and fixed
effects may all, to a certain extent, reflect the same tendencies. Including all
of these variables in a regression may cause over-fitting of the model.
However, even when country-specific fixed effects are dropped from the
regression, the key qualitative findings remain robust.
As an alternative to a simple summation of the policy types, hierarchical
cluster analysis can be used to identify clusters of policy instruments which
are then used as explanatory variables. Cluster analysis is a method that can
be applied in order to reduce a set of correlated variables into a smaller
number of cluster components, with little loss of information. It groups
variables in such a manner that variables within one group are correlated, but
uncorrelated with variables in the other groups. Figure 4.9 shows the estimated
tree diagram of hierarchical clusters (dendogram) for the six policy variables.
The choice of the number of clusters to retain must be made ad hoc. The
clustering of policy variables which yields three conceptually distinct groups
of policies was chosen: 1) price-based policy instruments (including investment
incentives, tax measures, and tariffs), 2) voluntary programmes; and 3) quantity-
based policy instruments (obligations and tradable certificates). The voluntary
programmes dummy is relatively equally correlated with either of the remaining
Table 4.10. Estimated coefficients of the negative binomial fixed effects models
with a composite policy variable
**
1.501 12.394 5.028 0.406
(0.304) (0.000) (0.811) (0.079) (0.188) (0.862)
**
0.971 13.999
*
7.304 0.496 1.110
**
(0.000) (0.144) (0.027) (0.059) (0.238) (0.000)
Electricity consumption 15.567
**
6.144 16.908 1.146 9.817
*
8.125
*
(0.005) (0.208) (0.253) (0.868) (0.020) (0.015)
**
0.064
**
0.055 0.063
*
0.119
**
0.089
**
(0.000) (0.000) (0.249) (0.025) (0.000) (0.000)
Composite policy variable 0.374
**
0.173
**
0.238
**
0.111
*
0.237
**
0.207
**
(0.000) (0.000) (0.000) (0.030) (0.000) (0.000)
Intercept 0.774
*
0.134 14.147 0.041 0.317 0.759
**
(0.036) (0.805) (0.985) (0.950) (0.487) (0.001)
Log-likelihood 496.61 494.88 240.15 299.02 487.99 931.35
Wald chi2 144.73 200.24 30.35 48.76 304.14 363.56
(Prob > chi2) (0.000) (0.000) (0.000) (0.000) (0.000) (0.000)
Notes: * and ** refer to the 5% and 1% level of statistical significance. P-values are in parentheses. The dependent
variable is the natural log of patent counts (successful and unsuccessful applications) in a given technological area. The
coefficient on the intercept represents the average value of the country-specific fixed effects. Results for geothermal
energy are not reported, because they represent a significant outlier.
159
two clusters. The estimated scoring coefficients were used to compute
component scores for each cluster.
Table 4.11 shows the regression estimates when policy variables are
divided into three clusters. The estimated results suggest that innovation
effects of alternative policy instruments differ by the type of renewable energy
technology. The evidence of differential policy effectiveness is particularly
straightforward for wind, solar, and waste-to-energy technologies. For wind
power, the coefficient of quantity-based policy instruments (cluster 3) is
positive and statistically significant at the 1% level. For solar and waste energy,
the coefficients of price-based policy instruments (cluster 1) are positive and
statistically significant at the 1% level. In addition, the coefficient of voluntary
programmes (cluster 2) in the waste equation is positive and statistically
significant at the 5% level or higher. For ocean energy, none of the policy
cluster variables are significant. For biomass, the coefficient of price-based
instruments (cluster 1) is positive and statistically significant at the 1% level.
Overall, for all renewables, the innovation effects of both price- and quantity-
based instruments are highly statistically significant. Voluntary approaches
seem to play a minor role.
In sum, the major finding of this Chapter is that different policies work
better for some technologies than for others. In particular, quantity-based
policy instruments such as obligations and tradable certificates are most
effective in inducing innovations in wind power technology. Price-based
instruments, such as investment incentives, tax measures and tariffs are most
effective in encouraging innovation in solar, biomass, and waste-to-energy
technologies. Voluntary programmes are not significant, except perhaps in the
case of waste. These findings are robust to alternative policy measures and
model specifications.
Figure 4.9. Dendogram of policy variable clustering
6.0 5.5 5.0 4.5 4.0 3.5 3.0 2.5 2.0 1.5 1.0
Investment
Tariff
Tax measures
Voluntary
Obligations
Tradable
Number of clusters
160
Interpretation of these results is complicated by the fact that it is difficult to
clearly distinguish the relevance of the policy variable by type of renewable. However,
there are good economic reasons to explain some of the findings. For instance, the
significance of investment incentives for solar energy is likely due to the fact that
many solar installations are the most capital intensive (in terms of investment costs
required per kW) among the studied renewable energy technologies (see, for
example, Dickson and Fanelli 2004). Waste-to-energy investments can also involve
significant up-front fixed costs, and the coefficient is significant and positive in this
case as well. In addition, the relative importance of voluntary programs for waste
may be explained by the importance of the public sector in waste management,
perhaps obviating the need for mandatory regulations. The significant and positive
coefficient on price-based measures (feed-in tariffs) for biomass may be explained by
the fact that this source is relatively mature and competitive, and thus changes in
relative prices are sufficient to have an effect on innovation. However, further work is
required to support (or reject) these conclusions.
Table 4.11. Estimated coefficients of the negative binomial
fixed effects models with clusters of policy variables
**
1.787 12.459 5.013 0.094
(0.547) (0.000) (0.775) (0.094) (0.190) (0.968)
**
1.100 15.028
*
6.705 0.490 1.069
**
(0.000) (0.091) (0.023) (0.067) (0.245) (0.000)
Electricity consumption 11.551 7.088 13.868 9.658 11.160
*
5.825
(0.073) (0.175) (0.361) (0.253) (0.011) (0.109)
**
0.075
**
0.064 0.094
**
0.116
**
0.087
**
(0.000) (0.000) (0.199) (0.004) (0.000) (0.000)
Policy clusters
Policy cluster 1 0.018 0.614
**
0.419 1.053
**
0.728
**
0.310
**
(incl. inv, tax, tar) (0.941) (0.000) (0.216) (0.000) (0.000) (0.005)
Policy cluster 2 0.006 0.030 0.075 0.127 0.366
*
0.072
(incl. vol) (0.980) (0.843) (0.842) (0.573) (0.031) (0.521)
Policy cluster 3 1.632
**
0.209 0.665 0.508 0.219 0.639
**
(incl. oblig, trad) (0.000) (0.184) (0.055) (0.074) (0.184) (0.000)
Intercept 0.132 0.006 13.727 0.346 0.027 1.029
**
(0.726) (0.991) (0.982) (0.646) (0.955) (0.000)
Log-likelihood 483.95 493.77 239.69 293.72 487.22 927.89
Wald chi2 216.29 200.81 31.89 64.08 297.49 393.14
(Prob > chi2) (0.000) (0.000) (0.000) (0.000) (0.000) (0.000)
Notes: * and ** refer to 5%, and 1% level of statistical significance. P-values are in parentheses. The dependent variable
is the natural log of patent counts (successful and unsuccessful applications) in a given technological area. The
coefficient on the intercept represents the average value of the country-specific fixed effects. Results for geothermal
energy are not reported because they represent a significant outlier. Policy cluster 1 includes investment incentives, tax
measures, and tariffs; Policy cluster 2 includes voluntary programmes; Policy cluster 3 includes obligations and
tradable certificates.
161
5. Conclusions
This Chapter has examined the effects of public policies on innovation in
the area of renewable energies in a cross-section of OECD countries over the
period 1978-2003. Patent counts were used as the most suitable proxy for
innovation, and the effects of a wide variety of different policy types were
assessed.
The descriptive data indicates rapid growth in wind and waste-to-energy
patent activity, particularly since the mid-1990s. There continues to be
innovation with respect to solar energy, perhaps reflecting the opportunities
presented by developments in concentrating solar power. Innovation with
respect to biomass and ocean are also growing, but from a very low base. And
finally, there appears to have been little innovation in the area of geothermal
energy since the 1970s.
At the same time there have been significant changes in the public policy
framework put in place to support renewable energy. Initially R&D programmes
were put in place in a number of countries. This was followed by investment
incentives, and later tax incentives and preferential tariffs. Later still, voluntary
programmes were developed. More recently, quantitative obligations, and
finally tradable certificates, have been applied.
The empirical results indicate that public policy has had a very significant
influence on the development of new technologies in the area of renewable
energy. Using the composite policy variable, statistical significance at the 1% level
is found for all renewable energy sources, except biomass (where it is significant
at the 5% level). However, the results suggest that instrument choice also matters.
With respect to patent activity in renewable energy overall, taxes, obligations and
tradable certificates are the only statistically significant policy instruments.
Interestingly however, source-specific models indicate that there is
variation in the effects of instrument type on different renewables. Broadly,
investment incentives are effective in supporting innovation in solar and
waste-to-energy technologies, tariff structures are important for biomass,
obligations and tradable certificates (which are closely related) support wind
technology, and voluntary programmes induce waste-to-energy innovations.
Overall, only investment and other tax incentives have wide influence on
innovation for a number of renewable energy sources.
While the results are interesting and robust, further work in the area
could be undertaken. This includes accounting for variation in natural conditions
as determinants of patenting in renewable energy technologies, and better
examination of dynamic issues, with a particular focus on addressing the
possible simultaneity of R&D expenditures and patenting activity. In addition,
there are a number of technologies which are particularly important for
electricity generation from renewable energies which could also be examined.
162
For instance, advanced batteries can overcome some of the losses associated
with intermittent energy supplies. Further work on assessing the effect of
innovation on the costs of generation and on the share of renewables in the
total energy mix is already underway.
Notes
1. Interestingly, EPO applications for patents in total increased approximately ten-
fold over the period in question, while patents for renewables increased just over
four-fold. However, in recent years the rate of growth in the area of renewables has
been higher than the rate of growth of total EPO applications.
2. The assistance of Hlne Dernis, OECD Directorate for Science, Technology and
Industry in the collection of the data is gratefully acknowledged.
3. This is further complicated by inclusion of a large number of countries in the
panel.
4. For further details on negative binomial models, see Cameron and Trivedi (1998).
5. Most countries introduced R&D support programmes as early as the 1970s. Since
the data begins in 1978, this policy variable is (almost perfectly) correlated with
the intercept (fixed effects), so it was dropped from the regression.
6. The coefficient is insignificant in the ocean equation. This is most likely due to the
low ocean patent counts.
7. The negative and statistically significant coefficient of tariffs in the wind energy
equation could be considered a consequence of multicollinearity. However, it
remains robust, even if the tax dummy is dropped.
8. In spite of this, a positive and significant coefficient is found, even for tradable
certificates. This suggests that allowing obligations to be traded provides a strong
incentive for innovation.
9. The exception is the ocean equation, where it is statistically insignificant.
10. The exception is the biomass equation, where the 1-year lag is negative and
significant.
References
Cameron, A. C. and P. K. Trivedi (1998), Regression Analysis of Count Data, Cambridge,
New York, Cambridge University Press.
Commission of the European Communities (2004), The Share of Renewable Energy in the
EU: Country Profiles (Brussels, COM).
Commission of the European Communities (2007), An Energy Policy for Europe (eur-
lex.europa.eu/LexUriServ/site/en/com/2007/com2007_0001en01.pdf) COM(2007)1 Final.
Dernis, Hlne and Dominique Guellec (2001), Using Patent Counts For Cross-Country
Comparisons Of Technology Output, STI mimeo, OECD, Paris (www.oecd.org/
dataoecd/26/11/21682515.pdf).
Dernis, Hlne and Mosahid Kahn (2004), Triadic Patent Families Methodology,STI
Working Paper 2004/2, OECD, Paris.
163
Dickson, Mary H. and Mario Fanelli (2004), What is Geothermal Energy? Instituto di
Geoscienze e Georisorse, CNR, Pisa, Italy.
International Energy Agency (2004), Renewable Energy Market and Policy Trends in IEA
Countries, IEA, Paris.
International Energy Agency/Johannesburg Renewable Energy Coalition (2005), Renewable
Energy Policies and Measures Database, OECD, Paris (www.iea.org/textbase/envissu/
pamsdb/index.html).
International Energy Agency (2006a), Renewable Energy: RD&D Priorities, Insights from the
IEA Technology Programmes, IEA, Paris.
International Energy Agency (2006b), Renewables Information: 2006, IEA, Paris.
OECD (2006a), Main Science and Technology Indicators, OECD, Paris (www.oecd.org/department/
0,2688,en_2649_34451_1_1_1_1_1,00.html).
OECD (2006b), Research and Development Statistics, OECD, Paris (www.oecd.org/department/
0,2688,en_2649_34451_1_1_1_1_1,00.html).
OECD Patent Project. Directorate for Science, Technology and Industry, OECD, Paris
(www.oecd.org/document/10/0,2340,en_2649_33703_1901066_1_1_1_1,00.html).
Popp, David (2005), Using the Triadic Patent Family Database to Study Environmental
Innovation, OECD Environment Directorate Working Paper ENV/EPOC/WPNEP/
RD(2005)2.
164
ANNEX 4.A1
First Page of Sample Patent Application
165
ISBN 978-92-64-04681-8
OECD 2008
167
Chapter 5
Policy Conclusions and Further Work
by
Nick Johnstone (OECD Environment Directorate)
A number of conclusions have emerged from the three case studies
that were undertaken as part of this study: 1) environmental policy
does have an effect on technological innovation; 2) general scientific
capacity is important in bringing about specifically environmental
innovations; 3) relative prices induce innovations in which commercial
and environmental benefits co-exist; 4) public concerns about the
environment appeared to encourage the development of some
technologies; 5) international diffusion of environmental innovation is
common; and 6) there is some limited evidence of a first-mover
advantage.
5. POLICY CONCLUSIONS AND FURTHER WORK
168
1. Policy conclusions
This report has examined the effects of environmental policy (and other
factors) on eco-innovation. A number of interesting conclusions have emerged
from the three case studies that were undertaken:
policy measures had a measurable impact on innovation, with both tax
measures and quota obligations being statistically significant determinants
of patent activity. However, the effect of different policies varied by the type
of renewable energy involved.
environment appeared to encourage the development of ECF and TCF
technologies, pre-dating the introduction of regulatory standards.
Interestingly, eco-labelling did not appear to have any influence on innovation
in this case.
renewable energy sources. In the case of motor vehicle emissions abatement,
there has been a shift from post-combustion technologies to integrated
technologies.
patent families (for some countries) were large, reflecting significant
169
technology transfer. In the case of motor vehicle emissions abatement, the
transfer of Japanese technologies to the US was striking.
1
There is some evidence of a first-mover advantage. For example, in the pulp
and paper sector, the early policy interventions introduced by Finland and
While not arising directly out of the results of the empirical work reported on
in this volume, a number of more general conclusions emerge from the work
reviewed as part of this project and summarised in the introductory Chapter.
These can be summarised as follows:
The general policy framework is of overarching importance. The factors which
drive eco-innovation are likely to be the same as those which drive innovation
in general. This includes macroeconomic factors, market regulation, and the
protection of intellectual property rights.
Environmental policy continuity is important. Investing in R&D is risky, and it is
important that the environmental policy framework not add to this risk. If
markets have difficulty efficiently dealing with commercial risk associated
with innovative activity, they will be even less likely to deal efficiently with
political risk. As such, stability in policy regimes is important.
Environmental and technology policy coordination is also key. Innovation and
environmental policy have different objectives. While the former is largely
concerned with internalising knowledge spillovers (and thus increasing
competitiveness and productivity), the latter is concerned with addressing
negative environmental externalities. While no single instrument is likely
to be able to address both market failures, co-ordination between the two is
vital, if both the rate and the direction of innovation are to be optimal.
Innovation arises out of a combination of both supply and demand factors. This is
no less true in the environmental sphere than it is elsewhere. As such, an
efficient environmental innovation policy framework must address both
sides of the market.
Policy incidence should encourage technology neutrality. Governments have
limited resources at their disposal, as well as limited information about
optimal technological trajectories. Moreover, with the potential for lock
in, it is important to develop policies which minimise the downside risks of
picking losers. In general, this means targeting environmental policies as
close as possible to the environmental objective itself, not some proxy.
Partnerships are important, but there should be separation of responsibilities
between public and private actors. This is clearest in the distinction between
basic and applied research. With the active role generally played by the
public sector in the environmental sphere, this issue is particularly important
here.
170
And finally, it is important to emphasise that this is an area where the
empirical evidence is limited, and a strong case can be made for the benefits
of further research. More importantly, efforts should be devoted to ensuring
that, in the process of policy development, the importance of ex post evaluation of
their innovation effects is recognised and accommodated i.e. through the
incorporation of ex ante data collection requirements.
2. Further work
The use of patent data to assess the extent, nature and determinants of
eco-innovation is a promising area of policy research. A number of different
areas could be explored, some of the most promising areas or which are
summarized below.
2.1. Development of robust indicators of eco-innovation based
on patent data
A key building block in assessing the role played by public policy factors
in the capacity of a country to benefit from environment-related innovation is
the development of an appropriate indicator of eco-innovation, which is
disaggregated by country, by year, and by type of environmental impact. As
has been noted, patent data has many advantages for the development of
such an indicator:
Patent classifications have been developed as detailed application-based
definitions. This means it is possible to focus on areas which are of direct
relevance to environmental concerns in a much more targeted manner than
is possible using sectoral or commodity classifications.
Unlike other commonly used data (i.e.environmental research and
development, scientific personnel, etc.) patents are an output-based
measure of innovation. The data generated is quantitative and
commensurable, allowing for the generation of indicators which are
comparable across time and across countries.
A time series for the indicators can be developed for all countries, stretching
back to 1977. In the event that users wish to modify the patent classifications
used in the indicator, the data can be modified at no cost, with no inputs from
government statistical agencies or other government officials.
Patent data can be married within other data sources, whether at the level
of the country, sector, or even enterprise. This allows for the empirical
analysis of a number of related policy-relevant questions.
While the existing work has focussed on specific areas of environmental
technologies, further work would allow for a broader perspective on eco-
innovation. The objective of such work would be to identify patent classes in
171
the International Patent Classification scheme, which are relevant for a broad
set of environmental technologies, including:
waste management and recycling;
wastewater treatment;
carbon capture and storage;
climate mitigation in selected sectors;
building energy efficiency;
air pollution abatement from stationary sources;
motor vehicle fuel efficiency;
environmental monitoring equipment;
green chemistry.
On the basis of this work an aggregate indicator of eco-innovation could
be developed, along with thematic indicators. This would be analogous to
work previously undertaken at the OECD, which has developed indices of
patent activity in the areas of biotechnology and information and
communications technology.
2
These indicators have been widely diffused and
applied. Experience in these two areas indicates that in addition to its value as
a contribution to measuring progress towards sustainable development,
general and thematic indicators of eco-innovation would serve as valuable
inputs to a wide variety of policy-oriented empirical studies.
2.2. Assessing the economic and environmental returns
on eco-innovation
While patents are an output from inventive activity, unlike other
measures such as investment in R&D and recruitment of scientific personnel,
they are intermediate outputs. As noted earlier, the value of patents can vary
widely. Moreover, it can be difficult to develop patent counts which accurately
reflect eco-innovation, particularly for technologies which represent
integrated changes in production processes and process design.
3
Therefore, it is important to assess the environmental and economic
benefits from patented inventions. For instance, in areas in which it is possible
to develop robust indicators of environmental impacts (e.g. emission levels) for
different environmentally-preferable technologies (and their conventional
substitutes), the environmental benefits of patented inventions could be
assessed. Areas such as motor vehicle emissions controls, green chemistry,
and wastewater treatment could be promising areas for such research.
In addition, to the environmental implications of eco-innovation, the
economic returns could be assessed. This would involve an assessment of the
links between patent activity, knowledge stocks, and economic activity. Work on
172
the links between renewable energy patents and the market share of renewable
energy is already underway. However, as noted in the Introduction, it is important
to take into account potential crowding out of inventive activity with respect to
environmental technologies on inventive activity in the economy in general. This
is an area which has not been previously addressed in the literature.
2.3. Eco-innovation in a globalizing world economy
The work reported on this volume, as well as that undertaken elsewhere
makes it clear that eco-innovation needs to be understood in the broader
context of a globalising world economy. In this vein, there are three other
pieces of information which can be obtained from patent applications and
which can be used to further refine our understanding of international economic
factors:
Patent citation data gives an indication of the international diffusion of
knowledge, by providing a trace of the geographical origins of relevant
previous innovations.
Data on foreign co-inventors gives an indication of the role played by
international collaboration in the invention process; and.
Patent family data gives an indication of the potential for technology
adoption, that is, the markets in which innovators feel they are likely to be
able to export their technologies.
With this information, the links between environmental innovation, and
issues such as technology transfer and international competitiveness could be
examined. Questions such as the benefits of international research co-
operation in environmental technologies, as well as the role of first-mover
advantage in environmental policy implementation could be assessed. The
relative openness of a country to overseas knowledge, research, and market
opportunities can also be affected by the nature of the environmental policy
instruments that are implemented. Conversely, some types of environmental
regulations may fragment markets for innovation, resulting in significant
inefficiency in technological development. Patent data is well-suited to the
assessment of these important policy questions.
Notes
1. Danish exports of wind power technologies to the US are also important, but
patent family data was not collected for this case.
2. www.oecd.org/dataoecd/54/57/33882780.pdf.
3. For a discussion of some of these issues see J. Labonne and N. Johnstone (2007)
Environmental Policy and Economies of Scope in Facility-Level Environmental
Practices (http://papers.ssrn.com/sol3/papers.cfm?abstract_id=1035661).
173
ANNEX 5.A1
Glossary of Relevant Patent
and Related Terms
Adoption: The point at which a technology is selected for use by an
individual or an organisation.
Applicant: The person or company that applies for the patent and
intends to work the invention (i.e. to manufacture or licence the technology).
In most countries the inventor(s) does not necessarily have to be the
applicant.
Application (or filing) date: The patent application date is the date on
which the patent office received the patent application.
Application for a patent: To obtain a patent, an application must be filed
with the authorised body (Patent Office) with all the necessary documents and
fees. The patent office will conduct an examination to decide whether to grant
or reject the application.
Assignee: The person(s) or corporate body to whom all or limited rights
under a patent are legally transferred. Assignment Transfer of all or limited
rights under a patent.
Bibliometrics: Study of the quantitative data of the publication patterns
of individual articles, journals, and books in order to analyze trends and make
comparisons within a body of literature.
Breadth (or scope): A measure of the extent of the invention covered by a
single patent application. For example, one patent application to the EPO
would generally more claims than an application to the JPO.
Citations: They comprise a list of references that are believed to be
relevant prior art and which may have contributed to the narrowing of the
original application. Citations may be made by the examiner or the applicant/
inventor.
174
Claim(s): These define the invention that the applicant wishes to protect.
A main claim will define the invention in its broadest form, by including its
essential technical features. Further dependant claims can then relate to
additional features of the invention.
Copyright: The legal right granted to an author, editor or publisher of an
article, chapter or complete work. Copyright applies to intellectual property in
a variety of artistic fields and attempts to be format-neutral.
Designated countries: Countries in which patent applicants wish to
protect their invention. This concept is specific to European patent
applications and international patent applications filed under the Patent
Cooperation Treaty (PCT).
Diffusion: The extent to which a technology spreads to general use and
application in the economy.
Duplicate: All patents relating to the same invention and sharing the
same priority, but filed at patent offices other than the priority office. The
count of such patents can be considered as the size of a simple patent
family.
ECLA: The European Patent Offices patent classification system. It is
based on the IPC Classification System, with greater disaggregation.
Equivalent: A patent that relates to the same invention and shares the
same priority application as a patent from a different issuing authority.
Espacenet: European Patent Office web site for searching, displaying and
downloading patent documents.
European Patent Convention (EPC): The Convention on the Grant of
European Patents (European Patent Convention, EPC) was signed in
Munich 1973 and entered into force in 1977. As a result of the EPC, the
European Patent Office (EPO) was created to grant European patents.
European Patent Office (EPO): The European Patent Office (a regional
patents office) was created by the EPC to grant European patents, based on a
centralised examination procedure. By filing a single European patent
application in one of the three official languages (English, French and
German), it is possible to obtain patent rights in all the EPC member and
extension countries by designating the countries in the EPO application. The
EPO is not an institution of the European Union.
European patent: A European patent can be obtained for all the EPC
countries by filing a single application at the EPO in one of the three official
languages (English, French or German). European patents granted by the EPO
have the same legal rights and are subject to the same conditions as national
patents (granted by the national patent office). It is important to note that a
granted European patent is a bundle of national patents, which must be
175
validated at the national patent office for it to be effective in member
countries.
Examiner: An employee of a patent office to whom an application is
assigned for handling prosecution.
Grant date: The date when the patent office issues a patent to the
applicant. On average it takes three years for a patent to be granted at the
USPTO and five years at the EPO.
Grant: A temporary right given by the authorised body for a limited time
period (normally 20 years) to prevent unauthorised use of the technology
outlined in the patent. A patent application does not automatically give the
applicant a temporary right against infringement. A patent has to be granted
for it to be effective and enforceable against infringement.
Home Bias: Propensity for the priority country to be the same as the
inventor or applicant country.
Infringement: Unauthorised use of a patented invention.
Innovation: The creation or introduction of something new, especially a
new product or a new way of producing something.
Intellectual property rights (IPR): IPR allows people to assert ownership
rights on the outcomes of their creativity and innovative activity in the same
way that they can own physical property. The four main types of intellectual
property rights are: patents, trademarks, design and copyrights.
International patent application: Patent applications filed under the
Patent Cooperation Treaty (PCT) are commonly referred to as international
patent applications. However, an international patent (PCT) application does
not result in the issuance of international patents, i.e. at present, there is no
global patent system that is responsible for granting international patents.
The decision of whether to grant or reject a patent application filed under the
PCT rests with the national or regional (e.g. EPO) patent offices.
International Patent Classification (IPC): The International Patent
Classification, which is commonly referred to as the IPC, is based on an
international multilateral treaty administered by WIPO. The IPC is an
internationally recognised patent classification system, which provides a
common classification for patents according to technology groups. IPC is
periodically revised in order to improve the system and to take account of
technical development. The current (eighth) edition of the IPC entered into
force on 1 January 2006.
Inventor country: Country of the residence of the inventor, which is
frequently used to count patents in order to measure inventive performance.
Inventor: Inventor names are recorded for all patents. These appear in
the standard last name-initial(s) format.
176
Japan Patent Office (JPO): The JPO administers the examination and
granting of patent rights in Japan. The JPO is an agency of the Ministry of
Economy, Trade and Industry (METI).
Kind Code: The letter, often with a further number, indicating the level of
publication of a patent. For example DE-A1 is the German Offenlegungsschrift
(application laid open for public inspection) while a DE-C1 is the German
Patentschrift (first publication of the granted patent).
Lapse: The date when a patent is no longer valid in a country or system
due to failure to pay renewal (maintenance) fees. Often the patent can be
reinstated within a limited period.
Learning by doing: Refers to the improvement in technology that takes
place in some industries, early in their history, as they learn by experience, so
that average cost falls as accumulated output rises. See infant industry protection,
dynamic economies of scale.
Learning curve: Relationship representing either average cost or average
product as a function of the accumulated output produced. Usually reflecting
learning by doing, the learning curve shows cost falling, or average product
rising.
Licence: The means by which the owner of a patent gives permission to
another person to carry out an action which, without such permission, would
infringe on the patent. A licence can thus allow another person to legitimately
manufacture, use or sell an invention protected by a patent. In return, the
patent owner will usually receive royalty payments. A license, which can be
exclusive or non-exclusive, does not transfer the ownership of the invention
to the licensee.
Novelty: If an application for a patent is to be successful, the invention
must be novel (new). The invention must never have been made public in any
way, anywhere, before the date on which the application for a patent is filed
(or before the priority date).
Obviousness: The concept that the claims defining an invention in a
patent application must involve an inventive step if, when compared with
what is already known (i.e.prior art), it would not be obvious to someone skilled
in the art.
OECD triadic patent families: The triadic patent families are defined at
the OECD as a set of patents taken at the European Patent Office (EPO), the
Japan Patent Office (JPO) and the US Patent and Trademark Office (USPTO) that
share one or more priorities. Triadic patent families data are consolidated to
eliminate double counting of patents filed at different offices (i.e. regrouping
all the interrelated priorities in EPO, JPO and USPTO patent documents).
177
Paris Convention: The Paris Convention for the Protection of Industrial
Property was established in 1883 and is generally referred to the Paris
Convention. The Paris Convention established the system of priority rights.
Under the priority rights, applicants have up to 12 months from first filing
their patent application (usually in their own country) in which to make
further applications in member countries and claim the original priority date.
Patent Cooperation Treaty (PCT): Signed in 1970, the PCT entered into
force in 1978. The PCT provides the possibility to seek patent rights in a large
number of countries by filing a single international application (PCT
application) with a single patent office (receiving office). The PCT procedure
consists of two main phases: a) an international phase; and b) a PCT
national/regional phase. PCT applications are administered by the World
Intellectual Property Organisation (WIPO).
Patent family: A patent family is a set of individual patents granted by
various countries. The patent family is all the equivalent patent applications
corresponding to a single invention, covering different geographical regions.
Patent family size is a measure of the geographical breadth for which
protection of the invention is sought.
Patent number: A patent number is a unique identifier of a patent. Patent
numbers are assigned to each patent document by the patent-issuing
authority. The first two letters designate the issuing patent office i.e. EP for EPO
patents and US for USPTO patents.
Patent: A patent is an intellectual property right issued by authorized
bodies to inventors to make use of, and exploit their inventions for a limited
period of time (generally 20 years). The patent holder has the legal authority to
exclude others from commercially exploiting the invention (for a limited time
period). In return for the ownership rights, the applicant must disclose the
invention for which protection is sought. The trade-off between the granting
of monopoly rights for a limited period and full disclosure of information is an
important aspect of the patenting system.
Patentability: Patentability is the ability of an invention to satisfy the
legal requirements for obtaining a patent. The basic conditions of
patentability, which an application must meet before a patent is granted, are
that the invention must be novel, contain an inventive step (or be non-
obvious), be capable of industrial application and not be in certain excluded
fields (e.g. scientific theories and mathematical methods are not regarded as
inventions and cannot be patented at the EPO).
PATSTAT: The EPOs World Patent Statistical Database.
Prior Art: Previously used or published technology that may be referred to
in a patent application or examination report. a) In a broad sense, technology
that is relevant to an invention and was publicly available (e.g. described in a
178
publication or offered for sale) at the time an invention was made. b) In a
narrow sense, any such technology which would invalidate a patent or limit its
scope. The process of prosecuting a patent or interpreting its claims largely
consists of identifying relevant prior art and distinguishing the claimed
invention from that prior art.
Priority country: Country where the patent is first filed before being
(possibly) extended to other countries.
Priority date: The priority date is the first date of filing of a patent
application, anywhere in the world (normally in the applicants domestic
patent office), to protect an invention. The priority date is used to determine
the novelty of the invention, which implies that it is an important concept in
patent procedures. For statistical purposes, the priority date is the closest date
to the date of invention.
Publication lag: In most countries, a patent application is published
18 months after the priority date. For example, all pending EPO and JPO patent
applications are published 18 months after the priority date. Prior to a change
in rules under the American Inventors Protection Act of 1999, USPTO patent
applications were held in confidence until a patent was granted. Patent
applications filed at the USPTO on or after 29 November 2000 are required to
be published 18 months after the priority date.
R&D expenditures: The basic measure of R&D expenditures is
intramural expenditures; i.e. all expenditures for R&D performed within a
statistical unit or sector of the economy.
R&D: Research and experimental development (R&D) comprises creative
work undertaken on a systematic basis in order to increase the stock of
knowledge, including knowledge of man, culture and society, and the use of
this stock of knowledge to devise new applications.
Renewal fees: Once a patent is granted, annual renewal fees are payable
to patent offices to keep the patent in force. In the USPTO these payments are
referred to as maintenance fees.
Rent: The premium that the owner of a resource receives over and above
its opportunity cost.
Reverse engineering: The process of learning how a product is made by
taking it apart and examining it.
Revocation: Termination of the protection given to a patent on one or
more grounds, e.g. lack of novelty.
Scientometrics: The quantitative study of the disciplines of science based
on published literature and communications. This could include identifying
emerging areas of scientific research, examining the development of research
over time, or geographic and organisational distributions of research.
179
Search report: The search report is a list of citations of all published prior
art documents which are relevant to the patent application. The search
process, conducted by a patent examiner, seeks to identify patent and non-
patent documents constituting the relevant prior art to be taken into account
in determining whether the invention is novel and includes an inventive step.
Technology Transfer: The communication or transmission of a technology
from one country to another. This may be accomplished in a variety of ways,
ranging from deliberate licensing to reverse engineering.
Term of patent: The maximum number of years that the monopoly rights
conferred by the grant of a patent may last.
Trade-Related Aspects of Intellectual Property Rights (TRIPS):
Agreement on trade-related aspects of intellectual property rights requires
members to comply with certain minimum standards for the protection of
IPR. But members may choose to implement laws which provide more
extensive protection than is required in the agreement, so long as the
additional protection does not contravene the provisions of the agreement.
The WTOs TRIPS agreement, negotiated in the 1986-94 Uruguay round,
introduced intellectual property rules into the multilateral trading system for
the first time.
United States Patent and Trademark Office (USPTO): The USPTO
administers the examination and granting of patent rights in the United
States. It falls under the jurisdiction of the US Department of Commerce.
Utility model: Also known as petty patent, these are available in some
countries (e.g. Japan). This type of patent involves a simpler inventive step
than that in a traditional patent and it is valid for a shorter time period.
World Intellectual Property Organization (WIPO): An intergovernmental
organisation responsible for the negotiation and administration of various
multilateral treaties dealing with the legal and administrative aspects of
intellectual property. In the patent area, the WIPO is notably in charge of
administering the Patent Cooperation Treaty (PCT) and the International
Patent Classification system (IPC).
Primary Sources:
OECD (2006), STI/EAS Division Glossary of Patent Terminology (2006)
(www.oecd.org/dataoecd/5/39/37569498.pdf).
Deardoffs Glossary of International Economic Terms
(www-personal.umich.edu/~alandear/glossary/).
OECD (2006), Economics Glossary: English-French (2006).
Thomson Scientific Glossary of Thomson Scientific Terminology
(http://scientific.thomson.com/support/patents/patinf/terms/).
OECD PUBLICATIONS, 2, rue Andr-Pascal, 75775 PARIS CEDEX 16
PRINTED IN FRANCE
(97 2008 06 1 P) ISBN 978-92-64-04681-8 No. 56339 2008
-:HSTCQE=UY[]V]:
ISBN 978-92-64-04681-8
97 2008 06 1 P
OECD Studies on Environmental
Innovation
and Patents
Environmental Policy, Technological
Innovation and Patents
Technological innovation can help realise environmental objectives in a less costly
manner than would otherwise be the case. Thus, understanding the role that
technological innovation can play in achieving environmental objectives is important for
policy debates.
However, the relationship between environmental policy and technological innovation
remains an area in which empirical evidence is scant. In an attempt to bridge this gap,
the OECD has examined the relevant issues, using patent activity as a measure of
Three case studies have been undertaken: abatement technologies for wastewater
efuent from pulp production, abatement of motor vehicle emissions, and development
of renewable energy technologies. On the basis of patent data, the nature, extent, and
causes of innovation in each of these areas have been explored. While a particular focus
has been placed on the role of environmental policy in bringing about the innovation
documented, it is recognised that other factors play a key role in inducing innovation
that has positive environmental implications.
The full text of this book is available on line via these links:
www.sourceoecd.org/environment/9789264046818
www.sourceoecd.org/scienceIT/9789264046818
Those with access to all OECD books on line should use this link:
www.sourceoecd.org/9789264046818
SourceOECD is the OECD online library of books, periodicals and statistical databases.
For more information about this award-winning service and free trials, ask your librarian, or write to
us at SourceOECD@oecd.org.
www.oecd.org/publishing
O
E
C
D

S
t
u
d
i
e
s

o
n

E
n
v
i
r
o
n
m
e
n
t
a
l

I
n
n
o
v
a
t
i
o
n

E
n
v
i
r
o
n
m
e
n
t
a
l

P
o
l
i
c
y
,

T
e
c
h
n
o
l
o
g
i
c
a
l

I
n
n
o
v
a
t
i
o
n

a
n
d

P
a
t
e
n
t
s

OECD Studies On Environmental Innovation Environmental Policy Technological Innovation and Patents OECD Studies On Environmental Innovation

Transféré par

Informations du document

Titre original

Copyright

Formats disponibles

Partager ce document

Partager ou intégrer le document

Options de partage

Avez-vous trouvé ce document utile ?

Ce contenu est-il inapproprié ?

Droits d'auteur :

Formats disponibles

OECD Studies On Environmental Innovation Environmental Policy Technological Innovation and Patents OECD Studies On Environmental Innovation

Transféré par

Droits d'auteur :

Formats disponibles

-:HSTCQE=UY[]V]:

Vous aimerez peut-être aussi