Knowledge flows through patent citation data

Hugo Pillu

Description

Sous la direction de Paul Zagame
Thèse soutenue le 18 décembre 2009: Ecole centrale Paris
Dans cette thèse, nous analysons les différents aspects des externalités de connaissance et la façon dont les citations de brevet peuvent être utilisées comme un indicateur de ces flux. La première partie de cette thèse examine la littérature traditionnelle sur les externalités de connaissance, et cela d'un point de vue à la fois qualitatif et quantitatif (la forme quantitative est réalisée grâce à une méta-analyse). Nous insistons sur les conséquences résultant de l'utilisation de différents canaux de mesure de ces externalités, précisément nous nous attardons sur les hypothèses sous-jacentes et sur leurs implications en termes d'estimations empiriques. Ce point est important car ces canaux sont la principale source d'hétérogénéité des résultats empiriques. Dans la seconde partie, nous explorons des données de brevets et de citations de brevet encore peu étudiées (ces données sont extraites de la base de données Patstat pour les offices de brevets du G5, de l'OEB et de l'OMPI). Cette analyse est à nouveau réalisée à la fois en termes qualitatifs et quantitatifs. La troisième partie, dans un premier temps, examine de façon empirique les caractéristiques des flux de connaissance entre et au sein des inventeurs des pays du G5 et cela pour 13 secteurs industriels. Dans un deuxième temps, cette partie propose et valide la création d'un indicateur de stocks de connaissance qui prend en compte les externalités de connaissance internationales. Cet indicateur se révèle particulièrement utile puisque les indicateurs traditionnels ne sont pas toujours disponibles (comme les indicateurs basés sur les dépenses de R&D). Enfin, l'indicateur précédemment créé sera appliqué à une étude de cas consacrée à l'analyse des déterminants de l'innovation pour les technologies énergétiques efficientes.
-Base de donnée sur les brevets
-Externalité de connaissance
-Diffusion de connaissance
-Innovation induite
In this dissertation, we analyze the different aspects of knowledge spillovers with special emphasis on the use of patent citations as an indicator of knowledge flows. The first part of the thesis reviews the traditional knowledge spillovers literature both with a qualitative and a quantitative view (the quantitative form is done through a meta-analysis). A particular focus is placed on the consequences of using the different channels reflecting different weighted functions, i.e. about the underlying assumptions and the implications in terms of estimates. This point is important since these channels are the main source of heterogeneity in the analysis of knowledge spillovers. In the second part, we shed some light on patent and patent citation data that have not been much studied (taken from the Patstat database for the G5 patent offices plus the EPO and the WIPO). Once again this analysis is conducted both in qualitative and quantitative terms. The third part, in a first time, investigates empirically the diffusion pattern of knowledge between and within the G5 inventors for 13 manufacturing sectors. In a second time, it proposes and validates the creation of an input-based indicator of knowledge that takes into account international knowledge spillovers. This indicator can be very helpful since traditional indicators are not always available (such as R&D indicators). Finally, the previously created indicator will be applied in a case study dedicated to analyze the determinants of innovation for energy efficient technologies.
-Patents
-Knowledge diffusion
-Induce innovation
Source: http://www.theses.fr/2009ECAP0048/document

Informations

• Publié par: Thesee
• Nombre de lectures: 88
• Langue: English
• Poids de l'ouvrage: 10 Mo

Extrait

ÉCOLE CENTRALE DES ARTS
ET MANUFACTURES
« ÉCOLE CENTRALE PARIS »



THÈSE

présentée et soutenue par :

Hugo PILLU

Le 18/12/2009

pour l’obtention du grade de

DOCTEUR


Spécialité : Sciences Economiques

Laboratoire d’accueil : ERASME

TITRE :

Knowledge Flows through Patent Citation Data




Directeur de Thèse :

Paul ZAGAME, Professeur à l’Université Paris 1 – Panthéon Sorbonne


Membre du Jury :

Dominique FORAY, Professeur à l’Ecole Polytechnique Fédérale de Lausanne, rapporteur

Jean-Luc GAFFARD, Professeur à l’Université de Nice - Sophia Antipolis, rapporteur

Gilles KOLEDA, Directeur des Etudes à Coe-Rexecode

Bruno van POTTELSBERGHE, Professeur à l’Université Libre de Bruxelles, président


Numéro d’ordre : 2009ECAP0048
tel-00458678, version 1 - 22 Feb 2010
tel-00458678, version 1 - 22 Feb 2010Knowledge ows through patent citation data
Hugo Pillu
tel-00458678, version 1 - 22 Feb 2010ERASME Ecole Centrale Paris, Grande voie des vignes, 92295 Ch^ atenay Malabry cedex, FRANCE
Phone: [phone] Fax: [phone] Email: [e-mail]
2
tel-00458678, version 1 - 22 Feb 2010Acknowledgement
First I would like to express my deep and sincere gratitude to my supervisor, Professor Paul Zagame,
Head of ERASME Laboratory and professor at the University Paris 1 - Pantheon Sorbonne, for his
supervision, advises, and guidance from the early stages of this research. His vast knowledge and
his logical way of thinking have been of great value to me. Above all and most needed, he gave me
unwavering encouragement and support in various ways. His scientist intuitions made him a constant
oasis of ideas and passions in science, which exceptionally inspired me and enriched my improvement
as the student, the researcher and the scientist I want to be. I am indebted to him more than he
knows.
A special thank goes to Gilles Koleda, who is most responsible for the completion of my rst paper
and the challenging research that lies behind it. He taught me how to write academic papers, helped
me become a researcher, and highlighted my most signicant ideas. Without his encouragement and
constant guidance in the rst period of my Ph.D., I could not have nished this dissertation. His
involvement and his originality triggered and nourished my intellectual maturity. A maturity that I
will enjoy for a long time to come.
I gratefully thank Professor Dominique Foray and Professor Jean-Luc Ga ard for agreeing to be
the o cial referees of this thesis and for their detailed review. I also wish to thank Professor Bruno
van Pottelsberghe, I am honored he has agreed to participate in the jury of this thesis.
During this work in the ERASME Laboratory at Ecole Centrale Paris, I collaborated with many
colleagues for whom I have great regard, and I wish to extend my warmest thanks to everyone who
helped me in my work. I thank Pascal Da Costa and Florent Pratlong for their intelligent comments
on my work, I also thank Arnaud Fougeyrollas and Pierre Le Mouel for their useful comments and
the fact they introduced me to one of the most challenging science that is modeling ; and Danielle
Schirmann Duclos for her energy and for her help during these years. Many thanks also go to Baptiste
Boitier, Gabriel Galland, Oualid Gharbi, and Boris Le Hir for giving me a good time when working
with them. A special thanks goes to Boris, who taught me how to y a helicopter.
The research of this thesis was sponsored by the Ecole Centrale Paris for three years (2006 to
2009). I am also grateful to the Research D.G. of the European Commission for their additional
nancial support of my thesis and of several conferences.
It is also a pleasure to pay tribute to Marion Senant, Marine Rabreau and Rudy Bezir for reading
over my manuscript and for their comments on the latter, it was a great help for me.
3
tel-00458678, version 1 - 22 Feb 2010I thank my family: my mother Nathalie Leconte, my father Pascal Pillu and my stepfather Francoi s
Coutin, who taught me the joy of intellectual pursuit, provided me moral advises and supported me
ever since I was a child.
I have been fortunate to come across many funny and good friends, without whom life would be
bleak. It would be di cult to list them all because it would be dangerous to forget some.
Last but not least, I wish to thank Charlotte for accompanying me during these years, for putting
up with me and just be present.
4
tel-00458678, version 1 - 22 Feb 2010