Technometrics

Patent analysis is one of the main research areas of ECOOM. The analysis of patent information is considered to be one of the best established, readily available and historically reliable methods for quantifying the output of science and technology systems. No other innovation indicator can be traced back over relatively long periods, and at the same time be disaggregated into low levels of analysis, relating to individual economic and technological units. Patent analysis also allows for accurate and precise mappings of the timing of innovation. This section provides an overview of patents and how they can be used for indicator development.

The following points are covered:

  • Defining patent
  • Patent statistics
  • The use of patent statistics for mapping technological activity

 

1. Defining patents

Definition

A patent can be defined as a document published by an authorized government body, granting the right to prohibit a person / institution to produce or use a specific new design, device or process for a certain number of years.

A patent is a property right that is based on an ‘officially sealed’ application. For the application to be recognized by competitors, all the details of the property right must be made public. The primary purpose of a patent system is to protect inventors. Without intellectual property, any technological knowledge that is developed would be public property, making it possible for competitors imitate without being penalized. Such a situation implies a market failure, as it leads to a suboptimal level of (private) investment in research and development.

 

Procedures and patent systems

Any inventor (individual, agency, or company) who wants to protect an invention in a specific country, files a patent application to the patent office of that country. Simultaneous protection in several countries is a possibility as well. The application contains one or more claims that show the extent to which the product or process is innovative. During the patent examination phase, these claims are further evaluated and compared to existing inventions: the so-called prior art search.

In most countries, an examination will take place to investigate novelty and potential (commercial) usefulness of the invention, before a patent is granted. The intellectual right associated with a patent may be assigned to another person / entity (e.g. the employer, another company,…). In addition, it may be sold / licensed to third parties.

Many different patent systems and offices exist worldwide. The main agencies are the European Patent Office (EPO) and the United States Patent and Trademark Office (USPTO), the World Intellectual Property Office (WIPO) and, increasingly, the Japanese Patent Office (JPO). In addition, most countries have a national patent office. Within Europe, the national agencies and the (overall) European Patent Office currently operate in parallel; but there has been an upsurge in the aspiration for a European patent. The major patent systems (especially the European and American one) differ mainly in terms of publication and award procedures. In the EPO system, a patent application is published after 18 months, whether or not the patent is or will be granted. Grants do not follow a strict timetable; and it may take up to 5 years before a patent becomes granted. In the USPTO system, until 2001, a patent was published only after it had been granted. Since 2001 however, the USPTO has been evolving to a similar publication policy as the one at EPO.

Such differences in patent procedures arise mainly from differences in the underlying philosophy. In the USPTO system, protection is focused primarily on the rights of the inventor. The European system from its part primarily aims for a timely dissemination of new technological information, to stimulate the pace of technological progress. Such a distinction is also reflected in various legislative principles. For USPTO, until recently, the ‘first to invent’ principle applied (intellectual property rights belong to the entity / actor who can demonstrate that he / she came up first with the idea leading to the invention). This contrasts to the EPO the ‘first-to-file’ 'principle (intellectual rights go to whoever was first in filing the patent application). Recently however – since 2011 – the USPTO has begun the transition from the ‘first-to-invent’ to the ‘first-to-file’ principle.

 

Patent documents

Patent documents worldwide are structured in a specific way. They consist of three basic parts:

  1. The title page with bibliographic information;
  2. The text with a description of the invention, detailed examples including drawings, diagrams and graphs;
  3. The claims.

Patent analysts usually only use the information that is available on the title page of a patent document. This title page contains bibliographic data that are of great interest to the analyst such as title of the invention, inventor and applicant details, application and grant dates, technological classification, and a list of patent citations or references of interest in the research process.

‘References cited’ concerns a list of references that the examiner considers to be relevant prior art (part of which may be provided by the patent applicant) and that may be a prelude to a narrowing the claims in the original application. These references can cite other patent documents, as well as non-patent (or ‘other’) documents. A significant part of the latter are references to scientific articles.

 

2. Patent statistics

For studying sources of economic growth, pace of technological change or competitive position of companies and countries, direct measures are lacking. As a consequence, researchers and policy makers are committed to using related indicators or ‘proxies’. Patent statistics are an appropriate ‘proxy’ for technological and innovative activity. Patent-related indicators have a number of methodological and technical benefits. Like any indicator, they also have limitations that must be taken into account. They are listed in the following table:

Benefits of patent-related indicators

Limitations of patent-related indicators

The close relation of patents to the outcome of industrial R&D and other inventive and innovative activities enhance their value as a proxy indicator.

Patents are present in almost every technology domain that is useful for analyzing the diffusion of core technologies (except software, which is usually protected by copyright).

Patents allow for geographical analyses.
 

The detailed technological classification of patent documents (by IPC codes) allows for a range of options in levels of aggregation; from broad technological areas to specialized technological niches.
 

Patent documents contain detailed information on the level of: relevant dates (request, grant, ...), technological classification, applicants and inventors (including addresses), citations,...
 

The broad accessibility and availability of electronic patent databases have made the use of patent statistics much easier.

Companies vary in their propensity to patent (number of patents per unit of R&D spending).

Technological domains vary in their propensity to patent.

Countries vary in their propensity to patent: differences in size and geographical position lead to different expectations w.r.t. the outcome of patent protection.
 

Differences between national systems in terms of  legal, geographic, economic and cultural factors limit the use of patent indicators.
 

Patents vary greatly in value.

 

Regardless of such limitations, patents remain a unique resource for studying the processes that lead to technological change. The availability of other indicators for innovation-related processes remains scarce. When patent statistics are used to answer questions about economic growth, about the pace of technological change or about the competitive position of companies and countries, a number of additional comments need to be kept in mind for an accurate interpretation of findings.

The term ‘patent rationale’ refers to the different motives that can lead to the decision to a patent application, or to refrain from it. The economic literature has devoted much attention to strategic patenting behavior of companies. Several empirical studies have shown that active involvement in innovation does not necessarily mean that companies apply for patents. In addition, it is noted that companies do not file patents on all of their inventions or technological developments. Other mechanisms such as secrecy, quick market launch or product complexity may complement or even substitute for patenting behavior.

This has important implications for the interpretation of patent statistics as a measure of the degree of innovation of companies and countries. The following points should be taken into account:

  • Not all inventions are commercialized and used, hence they do not all necessarily lead to innovation.
  • Not all inventions are patentable; and those that are, are in practice not necessarily patented.
  • Companies may patent for objectives other than commercial exploitation of the invention.

 

3. The use of patent statistics for mapping technological activity

As mentioned earlier, patents can be aggregated and analyzed in various ways, including:

  • Type of inventors, companies or groups of companies
  • Applicability in one or more technological domains
  • Patenting activity of a country or region
  • Patterns of patenting behavior over time

Patent indicators can be developed on three levels, whereby the required level of aggregation depends on the objective of the patent study:

  • Macro: country / region
  • Meso: industry / sector
  • Micro: single firm / institute

The simplest type of patent indicator is obtained by merely counting the number of patents, based on one or more criteria (technological domain, application year, inventor, applicant,…). Comparing patent volumes between countries, industries or companies in a particular technology domain can provide insight into technological performance differences. Taking into account population size, scientific texture and technology infrastructure, analysts relate patent volumes with demographic, economic and research variables (GDP, R&D expenditures,…). Such normalizations yield patent indicators that are independent of country size, allowing for more accurate comparisons.

An indicator that is often used - but that also receives some criticism - is the ‘propensity to patent’. This indicator represents the number of patents per dollar (or other currency), invested in R&D or in R&D staff. This indicator shows the extent to which R&D input is translated into patents, and can hence be considered as a measure of R&D output. Since the definition, delineation, measurement and quantification of the input data are not always straightforward, the ‘propensity to patent’ indicators need to be interpreted with caution.

Specialization indices can be used to answer questions about the position of a specific country or region in various technological domains, compared to other countries or regions. The ‘Revealed Technological Advantage (RTA)’ is the most frequently used specialization index.

The information in patents can also be used to construct ‘maps of technology’. For this purpose – besides information about the innovating company and on features of the invention itself – information can be gathered about the references contained in each patent application, both to previous relevant patents and to scientific articles on which the invention is based. ‘Maps’ of various technological (sub-)domains can then be constructed by examining the interrelation between frequently cited patents.  In addition, co-citation, co-classification or co-word analyses are possible. Within each (sub)domain, one can assess the relative position of different players, including companies, research institutions, countries / regions. Finally, and conditional on a detailed and careful identification cited scientific sources, these citations can be used to map relations between science and technology. This includes mapping knowledge flows between science and technology (at the level of countries, regions, institutions and even individuals) or the development of concordance schemes between scientific disciplines and technology domains.

ECOOM’s technometric research group has in-depth expertise and experience on the different technometric indicators and their uses as listed above.