Statistics Explained

Archive:Patent statistics background

Intellectual property rights and in particular patents provide a link between innovation, inventions and the marketplace. Applying for a patent makes an invention public, but at the same time gives it protection. A count of patents is one measure of a country’s inventive activity and also shows its capacity to exploit knowledge and translate it into potential economic gains.

In this context, indicators based on patent statistics are widely used to assess the inventive and innovative performance of a country. This article sketches the context and methodology of statistics on patent applications and grants in the European Union (EU) and some other European countries.

An analysis of the most recent EU data can be found in the Statistics explained dedicated to patent statistics.

Data sources and availability

From 2007 onwards, Eurostat’s production of European Patent Office (EPO) and United States Patent and Trademark Office (USPTO) data has been based almost exclusively on the EPO Worldwide Statistical Patent Database (PATSTAT). This was developed by the EPO in 2005, using their collection and knowledge of patent data.

European patent applications refer to applications filed directly under the European Patent Convention or to applications filed under the Patent Co-operation Treaty (PCT) and designated to the EPO (Euro-PCT), regardless of whether the patents are granted or not. For patent applications to the EPO all direct applications (EPO-direct) are taken into account, but among the PCT applications (applications following the procedure laid down by the PCT) made to the EPO, only those that have entered into the regional phase are counted. Patent applications are counted according to the priority date (the year in which they were filed anywhere in the world) at the EPO and are broken down according to the International patent classification (IPC). Applications are assigned to a country according to the inventor’s place of residence, using fractional counting if there are multiple inventors to avoid double counting. To normalise the data, the total number of applications at the EPO is divided by the national population and expressed in terms of patent applications per million inhabitants.

A patent application to the EPO can be valid in several countries and at most in all of the Contracting States of the European Patent Convention. In July 2009, the Convention was in force in 36 countries (all EU Member States plus Switzerland, Iceland, Liechtenstein, Norway, Monaco, San Marino, Croatia, the former Yugoslav Republic of Macedonia and Turkey). In addition to the Contracting States, three other countries (Albania, Serbia and Bosnia and Herzegovina) have concluded an ‘extension agreement’ with the EPO, by which these states can also be designated in a European patent application.

The falling trend in patent applications in the last years is linked to the length of patenting procedures and should not be interpreted as a real decline in patenting activity. For this reason the 2009 figures in Eurostat’s reference database are flagged as provisional or as estimates.

Patent indicators provide a measure of the innovative performance at country, firm or region level. Nevertheless, indicators are criticized as being “outdated”. This is due to the fact that information on patent applications is disclosed to the public 18 months or more after priority date. This issue is known as “timeliness”. In order to overcome this, Eurostat carried out a study where nowcasting methods have been discussed and developed.

The main purpose of this study it was to do the presentation of the existing methods for nowcasting of patent data to the European Patent Office (EPO) and the proposal of improved methods. An evaluation of these methods has been performed and conclusions regarding the most adequate methods for most of the countries have been drawn.

Moreover, an attempt to apply econometric models for nowcasting patent applications to the EPO was attempted as well as a comparison analysis. The existing econometric models are presented and 6 new models are detailed. A comparison analysis is performed and the strengths and weaknesses of each model is outlined.

In contrast to EPO data, the USPTO data refers to patents granted. Patents are allocated to the country of the inventor, using fractional counting in the case of multiple inventor countries. Comparisons between EPO and USPTO patents data should be interpreted with caution.

What is new?

KUL / EUROSTAT

Patent Statistics at Eurostat: Methods for Regionalisation, Sector Allocation and Name Harmonisation

Over the last decade, Eurostat regularly increased and improved the dissemination of patent data with the production of new indicators on economic activities. The core of the work is linked to the European Research Area (ERA), and to the priorities of the Europe 2020 strategy. There is therefore a need to enhance the information available in patent databases.

In that context Eurostat, in collaboration with the Katholieke Universiteit Leuven (K.U. Leuven), developed and implemented several innovative methodologies to be applied on PATSTAT data. They relate to regionalisation of patentee and inventors addresses (according to the NUTS classification), as well as sector allocation and name harmonisation of applicants.

This compendium of methodologies developed by Eurostat in the field of patent statistics contributes to the further development of indicators that are instrumental for analysis and policy development. It presents several methodological enhancements to deal with limitations in patent data sources. First, until now, no exhaustive sector allocation was available for identifying the nature of the applicant: individual, firm, university, public research organisation… Eurostat has bridged this gap by developing an exhaustive sector allocation methodology that is now made available for research and policy analysis.

Second, regarding applicant names in existing patent data sources, non-uniformity is the rule rather than the exception. Therefore, name harmonisation algorithms have been developed with a considerable impact in terms of coverage, resulting in highly improved indicator accuracy. Third, regionalization methodologies have been developed to better capture the regional dimension of technology development within the European Research Area (at EU-27 level).

These enhancements allow greater efficiency and accuracy in patent indicator extractions at regional,sectoral and institutional level, and are hence a considerable step forward in monitoring innovation systems in terms of technological activities.


REGIONALISATION OF PATENT DATA

Until recently, economic geography has played only a minor role in economic theory, despite the obvious fact that economic activities are not equally distributed over space. Relatively little empirical attention has been paid to the emergence and growth of regional clusters of technological activities. The existing evidence is mostly based on case study research, while large-scale empirical evidence or verification is rather scarce (Lecocq, 20103). One reason for this lack of large-scale empirical evidence on the phenomenon of technology clusters is the low availability of quantitative data at the region-technology level, covering regions worldwide over longer periods. Patent data, which provide information on the date and geographic location of technological development and on the organisations and institutions involved, have become increasingly available at regional level. However, in order to be able to construct patent indicators from them, addresses of inventors and patentees need to be allocated to regions. This section outlines a methodology for achieving this.

Regional patent statistics build on the allocation of inventor and patentee addresses to regions. This allocation or ―regionalisation exercise requires first of all an exhaustive list of postcodes and city names and their respective regions. Within Europe, the NUTS classification (Nomenclature of Territorial Units for Statistics) is a hierarchical system used to divide the economic territory of the EU4. It is used in the collection, development and harmonisation of EU regional statistics; in socio-economic analyses of the regions; and in the framing of EU regional policies.


SECTOR ALLOCATION

A corollary of this conception of innovation dynamics is the need for refinements in patent indicators. Sector assignment — i.e. identifying whether patentees are companies (private business enterprise), universities and higher education institutions, or governmental agencies — becomes a necessary condition for further analysis of the dynamics underlying technological performance.

This section outlines an updated version of the sector allocation methodology that was developed in 2006 (Van Looy, du Plessis & Magerman, 2006). It starts with an overview of previous efforts in sector assignment of patentees, indicating the relevance of additional development efforts. After that, the currently developed methodology and its outcomes are outlined. Conclusions are drawn on the performance of the current sector allocation methodology, and future avenues for further improvement are delineated.


NAME HARMONISATION

The development of patent indicators on the micro-level of specific entities like companies, universities and individual inventors is faced with specific concerns stemming from the heterogeneity of patentee names that appear in patent documents within and across patent systems. Whereas this poses no challenge to the functioning of the patent system itself, it does complicate analyses at patentee level: the analyst is confronted with inconsistencies such as spelling mistakes, typographical errors and name variants, which often also reflect idiosyncrasies in the organisation of R&D and/or IPR activities within a single organisation.

With the objective of reconciling completeness and accuracy, a comprehensive methodology was developed to obtain harmonised patentee names in an automated way. The methodology consists of several harmonisation layers. In a first layer, which emphasised accuracy or precision‘, the number of unique patentee names was reduced by approximately 20 % and the average number of patents per patentee increased from 5.5 before to 6.8 after harmonisation. In a second layer, emphasis was placed on ‗recall‘(a high coverage in terms of patent volumes). This layer covers the top 500 most active patentees, as well as university patentees. For the top 500 patentees, this additional harmonisation layer resulted in allocating over 30 000 patentee names to the top organisations, raising their aggregated patent volume by almost 70 %.


Method for harmonizing applicants' names and PATSTAT harmonized PERSON table

The K.U.Leuven/EUROSTAT method for harmonized patent applicant’s names is a comprehensive method to achieve harmonization of patentee names in an automated way. This method was applied on all applicant names in the PERSON table of EPO’s Worldwide Patent Statistical Database (PATSTAT), edition April 2009, resulting in a new PERSON table with harmonized applicant names. The developed method is based on the contents of the name and country address of the applicant name, no other information is used in the harmonization process. All names are processed by a step-wise validation process based on rules: character cleaning; punctuation cleaning; legal from indication treatment; common company word removal; spelling variation harmonization; condensing; umlaut harmonization. About 4 000 search and replace rules are executed for every step to handle the particular issue.

See: 'Data production methods for harmonized patent statistics: Patentee name harmonization'

A first version of the method was applied on EPO and USPTO applicant names in 2006 (EUROSTAT Working Paper - Magerman, T., Van Looy, B. & Song, X. (2006) Data Production Methods for Harmonized Patent Indicators: Patentee Name Harmonization.

An extended methodological outline describing more in details all aspects of: The Name harmonization versus legal entity harmonization and consolidation; the Automated rule-based system; the Trade-off between precision/accuracy and recall/completeness and the possibilities for Extending the method and for applying it on the EPO’s Worldwide Patent Statistical Database can be found under the following link: 'Harmonizing harmonized patentee names: an exploratory assessment of top patentees'


Data production methods for harmonized patent statistics: Patentee sector allocation

In addition to the ‘Method for harmonizing applicant's names’ the K.U. Leuven and Eurostat published in 2006 also a working paper on a method for sector allocation (Van Looy, B., du Plessis, M. & Magerman, T. (2006) Data Production Methods for Harmonized Patent Indicators: Assignee sector allocation. Eurostat Working Paper and Studies, Luxembourg). This method on assignee sector allocation has been reviewed and improved recently.
The latest results are available in the working document 'Evolution of innovation actors and the influence of legislation’.

In applying numerous rules based on keywords and/or clues patent applications are classified to five different institutional sectors:

More than 85% of all patent applications to the EPO are made by the business enterprises sector. For around 9% of the patent applications individual applicants can be identified. Whereas the patenting activity of the hospital sector is very small the patent applications of the remaining two sectors make up respectively about 2%. However one should bear in mind that these percentages are calculated for all patent applications to the EPO. This implies that the results may vary significantly across countries.

The working document focuses on the higher education sector and studies the role of legislation on the patenting activity in this sector.

The full methodology is described in the recent Eurostat working document 'Data production methods for harmonized patent statistics: Patentee sector allocation (2009)'.


Eurostat quarterly review of literature on Patent Statistics - December 2011

'This report' has been prepared in the framework of Eurostat‘s work on patent statistics. It consists of an extensive targeted review of the existing and currently developed methodological, analytical and scientific material and of the most recent relevant literature in the following domains:

  • PATSTAT data quality
  • Name harmonisation and gender;
  • Existing information and databases at institutional or company level or business surveys: comparison and possible integration and matching of data for analytical purposes;
  • Development of methods on patent data based on technical fields: biotechnology, nanotechnology, environmental technologies, energy, measurement of climate change, eco-innovation, technological innovation, other emerging fields;
  • Development of methods for compiling information on patent families and citations;
  • Regionalisation.

The report essentially consists of a collection of electronic articles, working papers and other publications available free of charge. An update of the report will be made available every three months.

Conferences on Patent Statistics

Flyer: Conference on Patent statistics for Decision Makers 2009

A conference on Patent Statistics for decision makers has taken place in Alexandria (USA), on 16 - 17 November 2011. The event it was organised by United States Patent and Trademark Office (USPTO) in co-operation with the Organisation for Economic Co-operation and Development(OECD), the European Patent Office (EPO), 'Eurostat', Japan Patent Office (JPO), US National Science Foundation (NSF), and World Intellectual Property Organization (WIPO).

Patent data and patent statistics are one important way of measuring innovation and science and technology-related activities, but they can also reveal important information about the patent system and its performance.

The goal of this conference was to discuss advances in the analysis of patent information in the context of the current economic landscape, and in light of changes within the patent system and its use by applicants. The conference brought together the latest thinking on issues relevant to companies and policy-makers. In particular, the conference emphasized how patent data can be used in decision-making about R & D investments, in assessing the competitive environment of certain technologies and in formulating policy reform for the system in order to adapt it to a changing economic environment.

The conference attracted around 130 people. Participating target groups included:

  • experts, statisticians and analysts working in the fields of innovation, patents, and science and technology policies, including academics and researchers using patent data for economic and policy analysis;
  • policy analysts dealing with innovation and patent policies;
  • national, regional and international patent institutions;
  • business professionals specializing in patent statistics and valuation;
  • business and intellectual property analysts;
  • national statistical institutes (and Eurostat).
Flyer: Patent Statistics for Decision makers 2010

Another conference on the same topic will take place in Vienna, on 17-18 November 2010. For more information: Patent Statistics for Decision Makers conference website.

Context

Patents are generally used to protect R & D results, but they are also significant as a source of technical information, which may prevent re-inventing and re-developing ideas because of a lack of information. However, the use of patents is relatively restricted within the EU. There are a number of possible reasons for this including:

  • their relative cost;
  • the overlap between national and European procedures;
  • the need for translation into foreign languages.

Most studies in this area show that innovative enterprises tend to make more use of intellectual property protection than companies that do not innovate. Enterprise size and the economic sector in which an enterprise operates are also likely to play an important role in determining whether an enterprise chooses to protect its intellectual property.

The European Council held in Lisbon in March 2000 called for the creation of a Community patent system to address shortcomings in the legal protection of inventions, while providing an incentive for investments in R & D and contributing to the competitiveness of the economy as a whole. In July 2000 the European Commission made a first proposal for the creation of a Community patent. This was discussed at various levels and despite a number of proposals and amendments for a Council Regulation on the Community patent during 2003 and 2004 no legal basis was forthcoming. In April 2007 the European Commission released a Communication ’Enhancing the patent system in Europe’. This demonstrates that the European patent system is more expensive, uncertain and unattractive compared to patent systems in competitor states. The report also underlined the European Commission's belief that a more competitive and attractive Community patent system can be achieved, based upon the creation of a unified and specialised patent judiciary, with competence for litigation on European patents and future Community patents.

Further Eurostat information

External links

See also

Retrieved from "https://ec.europa.eu/eurostat/statistics-explained/index.php?title=Archive:Patent_statistics_background&oldid=81725"