Final Report Summary - KNOWINNO (Making the most of knowledge)
Executive Summary:
« Making the most of knowledge » (acronym KNOWINNO) brings forward the measurement and policy agenda of the EC and the OECD on three distinct themes of mutual interest: (1) services as a source of knowledge creation and growth; (2) knowledge networks as a result of new collaborative practices, where knowledge sharing and exchange between firms, individuals and institutions have grown to facilitate knowledge creation and circulation; (3) the career path and mobility of doctorate holders, a population specifically trained for producing scientific knowledge. Key findings from the three strands of work are summarised in turn.
INNOVATION IN SERVICES: THE ROLE OF R&D AND R&D POLICIES
This work package examines the nature, prevalence and impacts of R&D and innovation activities in services and the policy frameworks used to foster new forms of service-based innovation. The quantitative part of the project aims at developing new empirical evidence on how service sector firms innovate and improving our understanding of service innovation throughout the economy. This analysis used a variety of data sources at the firm and sector-level, as well as different quantitative approaches to provide a comprehensive view of recent trends across countries. One contribution of this work was to update previous analyses on innovation (in services) by exploiting more recent microdata from a broader set of countries and new data linking possibilities. The research confirms that some service sectors are as innovative as high-tech manufacturing; that service innovation plays a pivotal role in firms’ strategies, is more reliant on external inputs and collaborative, with a strong focus on market-based relations, is associated with different appropriation mechanisms, and has a positive impact on performance. The policy and case-study module reviews the rationale for public support to innovation and discusses to what extent it applies to service-based innovation. The non-R&D based innovation elements of new service-based activities are forcing governments to rethink the focus on formal R&D, as service firms still benefit less frequently than others from public support for innovation. Service innovation can be supported by a wide range of policy instruments responding to different rationales, and each can be implemented using a variety of approaches. Proposals for improving measurement in this area are also made.
IPR STRATEGIES AND POLICIES IN THE ERA OF KNOWLEDGE NETWORKS AND MARKETS:
This work package examines the role of markets and networks for knowledge-based assets as the mechanisms and institutions facilitating the creation, exchange, dissemination and utilisation of knowledge and associated rights in the innovation system. The research provides new evidence on the knowledge sourcing strategies used by firms, their role in shaping their innovation strategies, and their link to business characteristics and performance. It introduces proposes a conceptual framework for understanding the purposes served by KNMs to support knowledge flows and the transfer of IP rights, illustrating a number of novel examples. It considers more specifically some developments in the market for IP rights, looking in first instance at the evidence on the size of the market and the role of intermediaries. The role of public policies in the IP marketplace is also considered, with particular emphasis on the new forms of policy interventions such as government-sponsored patent funds. It also reviews some features of the markets and networks for knowledge originating in public research organisations, as well as the role of intermediaries such as technology transfer offices whose role has been rapidly changing in recent years. The analysis of knowledge markets is finally extended to the market of knowledge embodied in highly skilled employees. The ambiguous impact of mobility on innovation is noted, considering in particular the use of agreements to restrict the movement of human capital and the potential implications of their enforcement. Some proposals for the future measurement agenda are outlined.
THE CAREER PATH AND MOBILITY OF DOCTORATE HOLDERS:
This work package has revised the CDH survey methodology building on the experience of the first two large-scale data collections; expanding the coverage of the project to additional countries worldwide (France, United Kingdom and Japan) using graduate first destination surveys and promoting the first time adoption of representative graduate/doctorate population surveys in a number of countries, including Germany. The research has supported the construction of a quality indicator database that users can easily access for the monitoring of labour market outcomes, career path and mobility of doctorate holders. The research results have demonstrated new uses for the available data, including additional indicators but also has illustrated the potential of CDH data for more structural types of analysis which allow for more like-for-like comparisons, for example, across gender and experience. The research has revealed how different countries place a different labour market premium on the mobility of researchers, and how different structural conditions impact on the careers of researchers and their choices. The project has enabled the formation of a research network, extending the number of users of CDH data beyond the statistical community to researchers in universities and analysts in ministries. The project has made a strong case for making the data more easily available for research uses, under appropriate arrangements that protect confidentiality, and facilitating linking to other databases. The CDH project has also engaged with the policy maker community in OECD member countries, in order to increase awareness about the potential of the CDH data, its synergies with complementary data sources, as well as the policy relevance of this body of evidence.
Project Context and Objectives:
Overall context and objectives of the proposal
The objective of this project is to bring forward the measurement and policy agenda of the EC and the OECD in three areas of common interest: (1) services as a source of knowledge creation and innovation; (2) knowledge networks and markets; and (3) the career path and mobility of doctorate holders, a population specifically trained for scientific research.
These three distinct themes address empirical and policy issues around knowledge as a source of innovation and productivity gains: i.e. knowledge as a production activity (in services); as a circulating asset which may or may not involve commercial exchanges (knowledge networks and markets); and as embedded in scientists who are mobile across sectors and countries.
Context and objectives by theme or work package:
Work package 1: “Innovation in Services: the role of R&D and R&D policies (SERVINNO)”
There is increasing recognition that knowledge creation and innovation are a feature of services activities and that having a dynamic and innovative service sector is key to boosting productivity. Moreover, service functions are being outsourced to non service sectors and thereby act as a source of knowledge creation and diffusion throughout the economy. However, the extent and nature of knowledge creation and innovation in services has been difficult to measure. For example, services are now well covered in most national R&D surveys but there are concerns that: a) some R&D undertaken by service firms might be under-reported; and b) that a significant share of service innovation is not R&D-based, therefore falling outside the scope of traditional policies for fostering R&D investment.
The objective of this work package is to examine, based on new measures, the extent and forms of R&D and innovation activities in services and to assess the role and impacts of innovation policies in the services sector. In particular the work package aims at:
• Improving current definitions and measures of R&D/innovation in services;
• Studying the relationship between R&D and innovation in the service sector;
• Assessing the impact of R&D and innovation in services on productivity, employment and growth in both the service and manufacturing sectors;
• Identifying the barriers to innovation in services and the role and impacts of policies stimulating service innovation performance.
Work package 2: “IPR Strategies and policies in the era of Knowledge Networks and Markets (KNM)”
The growth in transactions involving knowledge and the related emergence of new types of mechanisms for the governance of these transactions, labelled “knowledge networks and markets” (KNM), are changing the nature of innovation and raising new challenges and opportunities for policies in a variety of areas. KNM include IP marketplaces, IP funds, open access mechanisms and a variety of high leverage public-private partnerships. Understanding the conditions of emergence of KNM, notably in relation with the use of intellectual property rights (IPR) is a pre-condition for integrating KNM in policy frameworks, which will ensure the proper functioning of KNM and which will leverage KNM for serving broader policy goals (e.g. environment, development etc.). This work package aims at:
• Measuring and analysing stocks and flows of knowledge, including patents;
• Analysing the patenting behaviour of various categories of firms in OECD countries and investigate its determinants
• Developing a typology of knowledge networks and markets, all types of mechanisms which aim to create, access, share and use knowledge;
• Identifying approaches that can quantify prevalence and impact of such arrangements;
• Analysing ways in which policies that influence the circulation of knowledge may need to change in order to facilitate and/or govern emerging knowledge networks and markets
Work package 3: “The career path and mobility of doctorate holders (CDH)”
Doctoral graduates are seen as having a key function: firstly, they are specifically trained for scientific research; secondly, they hold a diploma at the highest level of the higher education system and as such are considered as the best qualified for the creation, implementation and diffusion of knowledge and innovative products and processes. However, not much is known about their labour market outcomes, career path and mobility. It is with this in mind that the OECD initiated a collaborative project with the UNESCO Institute for Statistics and Eurostat which has led to a first large-scale data collection in 2007 currently followed by a second one in 2010.. It is a new source of individual-level data that can be a rich source for new comparative empirical and policy analysis. This work package aims to:
• Improve the CDH survey methodology building on the experience of the fisrt two large-scale data collections. In particular, refine the measurement of certain key aspects of doctorate holders’ careers such as their post-doctoral activity or intra-sectoral and international mobility;
• Expand the project to cover an additional and growing number of countries worldwide;
• Continue to develop and maintain a quality database that will secure a regular monitoring of the labour market outcome, career path and mobility of doctorate holders;
• Exploit the individual-level data to produce a new set of policy relevant comparable indicators and analysis.
Project Results:
Improving the conditions for innovation and economic growth is high on the agenda of both the OECD and the European Union.
Central to this is the concept that knowledge is the key prerequisite for progress and value creation.
This project develops empirical and policy analysis on the creation, circulation and returns from investing in knowledge.
« Making the most of knowledge » (acronym KNOWINNO) brings forward the measurement and policy agenda of the EC and the OECD on three distinct themes of mutual interest:
(1) services as a source of knowledge creation and growth
(2) knowledge networks as a result of new collaborative practices, where knowledge sharing and exchange between firms, individuals and institutions have grown to facilitate knowledge creation and circulation
(3) the career path and mobility of doctorate holders, a population specifically trained for producing scientific knowledge. Key findings from the three strands of work are summarised in turn.
INNOVATION IN SERVICES:
THE ROLE OF R&D AND R&D POLICIES:
Services are an important part of the economy, accounting for around 70 percent of GDP in OECD countries. In the context of lagging productivity and slow job growth, many OECD governments are looking for new sources of growth. Boosting productivity in services through innovation is one way to re-start growth. The INNOSERV project contained two main modules, focusing respectively on measurement and policy issues.
The measurement module aimed at developing new empirical evidence on how service sector firms innovate and improving our understanding of the nature, prevalence and impacts of service innovation throughout the economy.
In addition, the work included a review of existing data sources and indicators for R&D and innovation in services in order to assess their strengths and weaknesses.
Finally, the work examined sectoral taxonomies.
The analysis in this part of the project used a variety of data sources (e.g. business R&D, innovation surveys, IPR data) at the firm and sector-level, as well as different quantitative approaches (e.g. indicators, econometric methods, exploratory data analysis) to provide a comprehensive view of recent trends across OECD countries. One contribution of this work was to update previous analyses on innovation (in services) by exploiting more recent innovation survey data from a broader set of countries (up to CIS-2010, or surveys covering a similar period outside of Europe), and new data linking possibilities (e.g. with patent and trademark data).
The second module focused on policy understanding of the nature of service R&D and innovation. It examined the rationale for public support to service R&D and innovation. It also reviewed the key features of policies and measures used by governments to support innovation in service-based activities. Business case studies were used to make observations on the nature of service innovation in firms. In addition to identifying good practices across a range of programme types, the study compared service R&D and innovation policies across 14 countries, covering United States, China, Australia, Europe and Asia.
The comparisons drew on publicly available material and a newly conducted survey of companies member of the Business and Industry Advisory Committee to the OECD (BIAC).
CHARACTERISING SERVICE INNOVATION:
Some service sectors are as innovative as high-tech manufacturing sectors; service innovation plays a pivotal role in firms’ strategies, is more reliant on external inputs and collaborative, with a strong focus on market-based relations, is associated with different appropriation mechanisms, has a positive impact on performance.
• The service sector accounts for a relatively modest share of business R&D in most countries, but not all innovation is based on R&D, and this is particularly important for services: almost half of all product innovating firms in services did not carry out any internal R&D activities
• Data from innovation surveys show that the service sector is heterogeneous and that some service industries are actually amongst the most innovative. Considering all types of innovation together (product, process, organisational or marketing innovation), the rates of innovation intensity in KIS are comparable – sometimes even higher – to those in high-tech manufacturing sectors.
• Innovation rates across sectors tend to vary depending on the type of innovation considered. High-tech manufacturing sectors show the highest rates of companies introducing new goods, whereas Knowledge Intensive Sectors (KIS) are the leading sectors regarding the introduction of new services
• However, service innovation also takes place outside the service sector. A significant share of manufacturing firms innovating in products – almost one out of three in the EU-15 countries – introduce both new goods and new services.
• This also reflects a trend towards a “servitisation” of products sold by manufacturers, a change in their business models. Likewise, a number of service firms also introduce new goods, which suggests that goods and service innovation are both hybrid in terms of sectors.
• Some manufacturing sectors are among the most likely to introduce new services. The highest rates of firms introducing new services are found in KIS and in high-tech manufacturing sectors (with respectively 28% and 16% of service innovating firms), followed by other services and other manufacturing (both 9% of service innovating firms). This ranking does not change when considering the propensity to introduce service innovation across sector, controlling for firm size, country, and belonging to a domestic or a multinational group.
• Service innovation plays a central role in the innovation strategies of firms. It tends to uniformly co-occur with multiple types of innovation, with a slight tilt towards marketing innovations and some process innovations such as the adoption of new logistics methods. Innovation in goods is found to be closer to technological activities such as internal R&D and the acquisition of machinery and equipment. Service innovation is also found to be more closely related to the acquisition of external knowledge, which points towards a more open mode of innovation than innovation in goods.
• The question of openness of innovation is a key issue for policy-making, as it may underpin potential market failures relating to the coordination across different actors. Those market failures, if they are found to be associated with specific types of innovations, may constitute a justification for policies targeting those types. Service innovation and innovation in service firms are more frequently reported to be developed outside the firm or jointly with other firms or institutions than innovation in goods and innovation in manufacturing firms respectively. Potential explanations are that service innovations tend to involve the participation of other actors, such as customers, and that they rely less on proprietary knowledge than other types of innovation.
• In terms of sectors, the firms which are the most likely to rely on external inputs for their innovations are those in services other than KIS. Firms in KIS are the most likely to actively collaborate
• Differences in collaboration rates appear to be mainly driven by market-oriented types of collaboration. Service firms tend to collaborate significantly more than manufacturing firms with suppliers, clients or competitors, whereas the level of collaboration with universities, consultants or public research institutes is similar in the two broad sectors.
• Service innovation is positively related to firms’ performance: overall, service innovation introduced by firms is found to be positively and significantly related to employment growth in those firms, the relationship being particularly strong in KIS.
• Beyond firm-level impacts, a key-concern for policy makers is the question of spillover effects. The results of our analyses are consistent with the existence of positive spillover effects associated with service innovation and innovation in services. In KIS and in other services, firm performance is found to be significantly and positively related to the share of competitors introducing service innovations. Moreover, firm performance in KIS is found to be significantly and positively related to the share of competitors introducing innovations whichever the innovation type considered, except innovation in goods. In high-tech manufacturing sectors, by contrast, no evidence is found of the existence of positive innovation-related externalities.
• Trademarks are the predominant type of IP right used by firms in service sectors, with 15% of firms using trademarks in KIS and 10% in other services. Patents are only rarely used in services (only 4% of firms are found to use patents in KIS and 1% in other services, versus 7% in high-tech manufacturing) as innovations in these sectors are often not patentable because of their intangible or non-technological nature. Trademarks may represent an alternative way to protect such innovations:
• In the majority of countries, most patenting activity takes place outside the service sector. However in a few countries services account for a significant share of patent applications, such as in Ireland, Poland or UK, where more than half of patent applications by firms at USPTO and EPO stem from business services.
• Service trademarks are widely used, often in combination with goods Significant differences are observed across countries, with certain countries such as Spain, Luxembourg or several Eastern European countries more specialised in service trademarks, and others such as Japan, Korea or BRIICS countries strongly specialised in goods trademarks. Moreover, trademarks are frequently registered both in goods and services classes, which tends to confirm that the delimitation between innovation in goods and in services is not clear-cut.
FOSTERING SERVICE INNOVATION
The non-R&D based innovation elements of new service-based activities force governments to rethink the focus on formal R&D. The various instruments to foster service R&D and innovation do not all respond to the same policy rationale.
Service innovation activities vary greatly in nature from firm to firm. Service firms benefit less frequently from public support for innovation. Service innovation is supported by a wide range of policy instruments, and each can be implemented using different approaches.
• As in the case of manufacturing firms, the main rationale for governments to promote innovation in service firms is based on market failure arguments that prevent firms from investing in R&D and innovation at socially optimal levels. Some of the market failures faced by manufacturing and service firms are alike, such as information asymmetries, capital market failures, or knowledge spillovers which limit the appropriability by firms of their innovations and hence lead to under-investment in innovation activities. Nevertheless, the literature on the economics of innovation shows that market failures in innovation may particularly affect service firms because of the very nature of service activities (i.e. intangibility, low tradability, customer proximity for delivery).
• Policy should recognise service innovation that is not based on formal R&D. Some firms engage in well-defined innovation projects, such as the development and introduction of new service concepts and products, whereas others primarily make continuous improvements to their services, business models and operations. Both types of firms are innovative. Service innovation can consist of the implementation of a single significant change or of a series of smaller incremental changes that together constitute a significant change.
• With a few exceptions, firms in services are less likely to receive public financial support for innovation. Public financial support for innovation, as reported by companies in innovation surveys, is generally more frequent in manufacturing than in services. In a number of countries (Australia, Austria, Estonia, Finland, Italy, Japan, the Netherlands and the Slovak Republic) public support is always more frequent in manufacturing than in service sectors, with low-tech manufacturing firms receiving more support than firms in knowledge-intensive services, which may suggest a potential bias against services.
• In most countries the gap between manufacturing and services in terms of the relative likelihood of receiving support has however narrowed slightly within the last decade, which may reflect a shift in countries’ policy strategies to support innovation.
• As far as indirect support is concerned, it does not appear to be biased towards certain sectors. The sectoral distribution of R&D tax incentives closely resembles that of business R&D spending, which confirms that they are a horizontal policy instrument and do not target specific sectors.
• Public support is found to be significantly related not only to innovation in goods, but also to service innovation. Firms which are involved in activities related to product or process innovation have a 15% higher probability of introducing service innovations when they receive public support than when they do not (versus a 22% higher probability for innovation in goods). However, the fact that the effect is lower for service innovation suggests that the current design of direct public support instruments may be better adapted for innovation in goods.
• Relative to high-tech manufacturing, firms in services tend to report fewer or similar obstacles to innovation. These results tend to counterbalance the policy rationale for supporting service innovation, but may also reflect a lack of awareness by service firms and a bias in the mindset regarding innovation and innovation policies
• More and more countries have some type of service innovation policy strategy but only a few countries have actually developed an action plan, and the number of fully-fledged policy strategies is even lower. Some policy strategies are part of a broader vision on innovation, or relate to major social challenges and innovation in the public sector Although the absence of a national orientation on service innovation can indicate a lack of awareness, there are also cases where service innovation policy is integrated in broader strategies. Still, most seem to be struggling to develop a balanced policy mix that sufficiently addresses the specificities of services.
• Service innovation is supported by a wide range of policy instruments, and each can be implemented using different approaches. There are many instruments and initiatives that contribute to service innovativeness, including measures outside the domain of R&D and innovation. Regarding policy instruments explicitly devoted to service innovation, they mostly focus on fostering an innovation culture including development and diffusion of management principles, and aligning research to innovation. Often, these measures coincide with SME-support.
• Policy instruments directly targeting service innovation are often found on the supply-side, while novel instruments are mostly indirect and demand-oriented
• Evidence of the design and implementation of policy instruments for service innovation is still limited, and as many instruments are new, impact assessments are rare or not forthcoming soon enough. Moreover, most of the already existing service innovation instruments are part of a more generic scheme, such as tax credits and voucher schemes. Such instruments will probably be evaluated, however, without a specific focus on the service innovation aspect
• Of great importance is the still insufficient availability of metrics that capture the peculiarities of service innovation. Without indicators that can truly deal with the tacit and intangible nature of services, policy makers will continue to struggle with demonstrating the impact of innovative efforts and policy support.
KEY MESSAGES FOR IMPROVING MEASUREMENT
• There is a need to better clarify the distinction between extramural and intramural R&D: the increasing complexity in how R&D activities are organised within and across firms, especially in services, raises a number of measurement challenges (e.g. around purchased R&D, activities of on/off-site consultants, etc.).
• There is also a need to clarify the definitions and boundaries of R&D in surveys and how they relate to service firms/activities: e.g. the distinction between service sector R&D versus service sector innovation.
• Work must continue on improving the international comparability of data on R&D by industry. Not all countries use the same approach for reporting R&D by sector of activity: some classify R&D according to the main activity of the firm in terms of value-added, others according to the nature of the R&D activity itself. Better guidance is needed regarding the way in which to collect and report R&D statistics by industry.
• Current measurement frameworks still tend to favour technological and R&D-based innovation which is less frequent in services: e.g. within innovation surveys only a few questions are included on service innovations which are only considered as a sub-type within product innovation. Likewise, there is much less detail collected on marketing and organisational innovation, and limited information on knowledge-based assets, all of which are usually associated with service-based innovation.
• There are still challenges regarding the understanding, applicability and relevance of certain innovation-related concepts to firms, particularly in services. For example, the notion of ‘significantly improved’ seems problematic for some firms, particularly as it relates to service innovation which is often gradual/incremental.
• Further work should focus on developing and testing new indicators of non-R&D based and service innovations, particularly regarding impacts.
• Efforts should be continued to improve the coverage of innovation surveys: some service firms still fall outside the scope of innovation surveys due to their sectoral coverage and firm size thresholds. This should be continuously reassessed.
• A number of areas related to service innovation should be further explored including: o Measurement of design and other knowledge-based capital; o Public procurement as a tool for fostering service innovation; o Public sector innovation; o Measurement of policies and framework conditions.
• There is a need to further experiment with alternative analytical approaches, for example by looking beyond industrial classifications or focusing on the role of service innovation in addressing social challenges.
• In addition to surveys, efforts should focus on exploiting alternative data sources such as administrative data.
• Given the heterogeneity of firm strategies within and across sectors, this work confirms the importance of having access to firm-level data at the national and international level to properly measure and study service innovation at the level of firms and sectors.
• Efforts should be made to facilitate data linking in order to better understand the specificities of service innovation in both services and manufacturing sectors; and to ensure open access to public data.
IPR STRATEGIES AND POLICIES IN THE ERA OF KNOWLEDGE NETWORKS AND MARKETS
Over the last two decades, there has been a major surge in the amount of information, knowledge and associated intellectual property (IP) rights that have been generated on a global basis across all domains of innovation and economic activity. Investment in various forms of knowledge-based capital has experienced a sustained increase, while increased competition from emerging economies which are moving towards more knowledge-intensive activities are challenging traditional assumptions about the way firms and countries specialise along innovation and value chains.
New digital technologies have brought about large reductions in the cost of copying, storage and distribution of data and information. In this environment in which information has become more abundant, easy to access and copy, IP rights have become more important as the tool that allows knowledge creators to exclude third parties from using an idea or the expression of an idea to extract value from knowledge-based assets. This unprecedented availability of information and volume of overlapping rights has led to a continued search for mechanisms that help individuals, business and organisations navigate this complex ecosystem, identifying relevant sources or providers of knowledge beyond their immediate network of contacts, enabling them to clear IP rights and transact upon them with minimal transaction costs.
Efficient knowledge markets make it possible to separate various components in the production and use of knowledge across firms, thus encouraging specialisation through the division of innovative labour. Within knowledge markets, enforceable IP rights help address the threat of misappropriation in downstream markets which could otherwise lead firms to adopt less efficient, internal development strategies. Knowledge markets can thus contribute to encouraging the exploration of knowledge development avenues which might otherwise be left unexploited.
The concept of knowledge networks and markets describes the diverse set of systems, institutions, procedures, social relations/networks and infrastructures that aim to allow companies, organisations and individuals to safely engage in the meaningful exchange of knowledge and associated rights. The common defining feature of these mechanisms is the expectation that they can provide a number of critical services to actors in the innovation system along the various phases throughout the process of exchanging knowledge and associated rights. These services range from facilitating the search for and matching to relevant counterparties and their knowledge, through to evaluating, executing and enforcing agreements. The concept of KNMs is probably too broad to be considered as a single, all-encompassing element for analytical purposes, particularly those of a quantitative nature.
A wide range of approaches, from data sources to multi-disciplinary strategies, are needed to fully grasp the implications of policies in this area. Here are some selected findings and messages from the project: • Broad categories of KNMs can be defined on the basis of whether their focus is on facilitating the transfer of disembodied knowledge, as in the case of searchable registers and repositories of already-existing data and information; and platforms for sourcing solutions to ad-hoc problems and challenges, which includes platforms for implementing inducement prize incentives or identifying consultants to assist with new R&D projects. KNMs can instead directly focus on the facilitation of transfer of intellectual rights on disembodied knowledge. IP brokers, pools and funds primarily deal with the allocation of IP rights, which may frequently but not necessarily be driven by a knowledge transfer motivation, and the creation and management of financial assets and liabilities attached to them. As a further category, it is important to emphasize the role of institutions and actors that specialize in ensuring the transfer of knowledge embodied in goods or people.
• Analysis of detailed business innovation survey data for a majority of OECD EU countries reveals a small set of attributes which can describe a firm’s innovation strategy. Most of these (four out of five) reveal specific approaches for sourcing knowledge and collaborating - of which only two apply to R&D active firms. A fifth one, linked to non-technological forms of innovation introduced in the latest version of the Oslo manual, are seemingly more inward looking. However, this may only reflect the unequal treatment of some forms of non-technological innovation in the survey questionnaires, as no linkage questions are asked in relation to these other types of innovation that can be also very open in nature. This new “taxonomy” appears to provide a fair description of what are the prevalent modes of innovation and openness across countries. However, a more complete and balanced description of business innovation strategies still requires collecting evidence on inside-out activities, which are currently absent from most official surveys.
• R&D data collected under OECD Frascati guidelines can be used to examine key patterns and trends on the balance within internal and outsourced R&D activities by the organisation. A number of countries have also been experimenting with new questions on collaborative research and IP-related transactions relating to R&D. These are topics which may potentially feature in the forthcoming revision of the Frascati Manual to be undertaken over the next two years.
• Standard economic statistics are only beginning to account for the full implications of the market for ideas. In some countries, corporation tax data on licensing incomes provide an additional source of evidence on the growth in the knowledge market which complements the picture emerging from a wide range of ad hoc studies and already available data on international transactions on IP. The incorporation of R&D as a capital formation in the System of National Accounts will likely result in improved sources on intellectual asset creation and trade across and within national boundaries.
• New statistical data on IP specialist firms and intermediaries are now becoming available across many countries, resulting from the implementation of a new international statistical classification of economic activities. Data show that the value of their services is relatively small in comparison with the investment made in knowledge-based capital but appears to be increasing over time. The current comparison between the United States and European countries indicates that the European market appears to be significantly less developed.
• Markets for intellectual property rights are particularly complex for policy analysis purposes and, on the evidence available, it is not possible to conclude that achieving larger markets is necessarily a sensible policy objective to pursue. The absence of a healthy market for IP rights may be a symptom rather than the cause of weakness in the innovation system. Policy makers should concentrate on identifying these ultimate causes and evaluating the appropriate mechanisms for dealing with them. Among these, measures that improve the ability of markets to address the fundamental asymmetries that limit their effectiveness could feature quite prominently, but due care should be given to potential unintended impacts.
• Patent assertion entities are a relatively new species in the IP marketplace. They have been described by experts in the field as a path-breaking, (so far) legal disruptive technology for monetising patents that eliminates traditional obstacles to enforcement. Like many other operators in this space, they exploit some weaknesses in the system to their advantage and that of their customers. Their controversial business model is not universally successful, indicating that patent system features do have an impact on their viability.
• Government-sponsored IP funds, typically involving patents, are another new addition both to the range of intermediaries operating in the IP market place and to the portfolio of policy instruments considered by governments and public authorities. Their rationale, which is not unique across countries but has a common element in the desire to improve the valorisation of IP, address patent thickets and provide with a defence against disruptive litigation. The case for this type of instrument is by no means uncontested. The use of public funds to invest in IP titles should and the alignment of this practice with international treaties should be thoroughly scrutinised if implemented at all.
• Matched IP and business data provide a first set of insights into the relationship between IP strategies and business performance. The experimental matching of patent and business data on a global basis carried out under this project provides some evidence on the propensity to patent across firms based in different countries and in different size and industry groups. The data matching exercise also reveals the limitations of commercially available business registers – which are often limited in scope – and suggests an avenue for further data infrastructure development, bringing together IP-based administrative data and statistical business registers and other linked information, such as R&D and innovation data.
• This project has highlighted the importance of knowledge flows through people for the mobility of knowledge and the incentives by companies. Our understanding of the impact of institutions and regulations on the extent of job mobility, knowledge transfer and business innovation is still very incomplete. Less conventional data sources will likely need to be combined with traditional measures and tools to gain real insights on policy relevant aspects of knowledge transfer through people. This is particularly important for graduates and researchers who can take their knowledge from one workplace to another and interact in a number of channels.
• This evidence limitation on the impact of regulations applies for example to the legal enforcement of practices restricting a former employee’s ability to work for a competitor or set up her own business. Evidence collected throughout this project suggests that enforcement practices vary significantly across OECD countries and partner economies.
A number of countries share some of California’s statutory restrictions on the enforcement of non-compete agreements and among those, some share some entrepreneurship dynamics in specific sectors although the correlation is by no means perfect. The impact of these agreements is likely to vary across economies with distinct types of labour market institutions and the type of innovation system.
THE CAREER PATH AND MOBILITY OF DOCTORATE HOLDERS
Over the 2011-2012 biennium, the OECD activity on CDH has been partly sponsored by the European Union’s Seventh Framework Programme as part of the broad OECD KNOWINNO project. This supported the development of the CDH database by the OECD and has helped produce a set of internationally-comparable indicators based on the results from the 2010 CDH data collection.
In order to address a number of policy and analytical questions, four key areas of work were identified for detailed investigation:
1) early career of doctorate holders;
2) job-to-job mobility;
3) international mobility and
4) competences and skills of doctorate holders.
Each topic was led by a participant organisation: NISTEP/Japan for early careers, DGEEC/Portugal for job-to-job mobility, CSIC/Spain for international mobility and ECOOM/University of Ghent/Belgium for competences and skills.
In order to extend the number of countries for which comparisons could be carried out, a special effort was made to define common populations of doctorate holders among surveys of university graduates available for France, Japan and the United Kingdom, and subsamples within CDH surveys carried out in other countries. Access to and use of micro data has been instrumental in facilitating these comparisons that were carried out under the “early career module” of the project. Using a data coding guide provided by the OECD, ten volunteer countries harmonised their micro data sets in order to implement tabulations and econometric analyses using a common programming code developed by CSIC and NISTEP.
The OECD secretariat also carried out for demonstration purposes a comparative analysis of micro data on doctorate holders and individuals holding other postgraduate qualifications for the United States and the United Kingdom.
• The project highlights the potential of CDH data to inform policy questions that bear on the labour market and careers of doctorate holders and researchers. Beyond descriptive analysis of tabulations, the project makes use of econometric analysis based on micro-data to identify systematic patterns in the data which underpin overall differences.
• The past decade has witnessed a steady increase in the number of doctoral degrees being awarded across the OECD, rising by 38% from 154 000 new doctoral graduates in 2000 to 213 000 in 2009.
• There is only a weak association between the proportion of doctor graduates in the labour force and a country’s R&D intensity. Switzerland has the highest workforce share of doctorates, close to 2.8 per cent.
• Despite the growing supply of doctorates, the available evidence points to a sustained -possibly increasing- labour market premium on individuals holding doctoral qualifications, relative to other highly qualified individuals.
• Although female and younger doctorate holders fare relatively worse in terms of employment rates and earnings than their older and male counterparts, these biases are less marked for doctorate holders than for individuals with lower levels of educational attainment. The gender bias persists but is lower in size after controlling for other differences.
• Employment rates for recent doctorates are also high, but figures for 2009 still do not account for the wave of fiscal adjustment in many countries and its potential impact on the public funding of R&D. Temporary positions are increasingly common in the academic sector, but less so in business.
• Although the higher education sector is the main sector of employment for doctorates, demand for doctorates is apparent across other knowledge intensive sectors. The take up of jobs outside higher education is not necessarily, but often related to take-up of non research occupations. Work as a researcher becomes less likely as careers progress and other competencies are acquired.
• Natural scientists and engineers are more likely to be engaged in research while social scientists find more opportunities in non-research occupations that are nevertheless related to their degree.
• Job mobility patterns differ markedly across countries but mobility is more frequent among doctorates not working in research. Very few countries have more mobility from the higher education to the business sector than vice versa.
• Earnings in agricultural sciences and humanities are below the overall median in most countries, whereas doctorate holders in medical and health sciences tend to be paid above. Doctorates in the business sector are typically better paid than in other sectors, but surprisingly not in all countries.
• International mobility is a widespread and increasingly important phenomenon, although less common than it might be assumed for researchers. It has positive impacts on knowledge exchange and diffusion, but surprisingly it is not systematically associated with higher earnings.
• Individuals who have already experienced an episode of international mobility are more likely to report an intention to move abroad.
• Even when not in research, jobs are in most cases related to the subject of doctoral degrees and doctoral graduates are satisfied with their employment situation. A wide range of monetary and non pecuniary factors contribute to explaining the reported attractiveness of research careers. Satisfaction levels on aspects other than pay are particularly high for individuals working in the higher education sector.
The results of past CDH data collections, related documents and new reports arising from this project can be found in a dedicated website:
www.oecd.org/sti/cdh.
The project has also resulted in the production of updated methodological guidelines, comprising a revised model questionnaire –which includes new questions on competencies mainly developed by Belgium and the Russian Federation– for countries to use when collecting statistics on doctorate holders and thus ensure the internationally comparability of the results. The outlook for future CDH data collection differs across countries as some will be collecting data for the first time in 2012/2013; some will re-conduct a survey; while others have indicated they lack the resources to carry out new surveys. The OECD intends to continue to play an active role in collecting and disseminating statistics on doctorate holders as a key dimension of human resources for science and technology.
The OECD is also keen to raise the profile of CDH as an additional evidence resource to decision makers and to encourage statistical agencies to make the CDH micro-data available for wider research and analytical purposes, which are to date relatively underexploited and remain difficult to access by potential users. The results of the CDH-KNOWIINO work package aim to showcase the potential of CDH data to inform policy questions that bear on the labour market and careers of doctorate holders and researchers.
This potential goes beyond the mere analysis of tabular results, but also involves the use of micro-data based econometric methods to identify systematic patterns in the data which underpin overall differences. This type of analysis has shown, for example, that only part of the gender bias found for earnings can be explained by observable characteristics, or that the impact of international mobility on earnings greatly varies across countries and may reveal some fundamental, structural differences.
By demonstrating the feasibility of carrying out a wider range of possible analyses among a reduced set of countries, this project has also attempted to make a strong case for improving researchers’ access to data on doctorate holders with a view to promoting the coordinated use of micro-data across countries and conduct more in-depth analyses of the factors shaping career choices, labour market mobility and individual performance, linking whenever possible to other available databases. Carrying out the analyses has required extensive efforts by participating experts to overcome a wide range of barriers to render the data accessible and suitable for analysis. This would not have been possible without the goodwill and relentless efforts made by partnering organisations. Some best practices have also been identified, for example cases in which the data has been anonymised and made publicly available.
Future survey collections should foresee the future analytical and research intended uses of the data, secure the necessary consents and provide the necessary infrastructure. In their absence, data collected at a high expense to taxpayers could be left unused, a luxury that countries can ill-afford. CDH data should be considered as a critical but not exclusive source of evidence on the policy questions of interest regarding doctorates and research careers. The project has equally outlined the importance of utilising mainstream household data sources such as census and labour force data. This requires an explicit effort to capture information on the educational attainment of individuals at a sufficiently detailed level.
In many countries, there is room for using employment registers to trace mobility with much higher accuracy than at present. Business surveys could potentially inquire about the employment of individuals at the doctorate level to help inform the analysis of the demand for doctorates. The emergence of the internet and the wealth of information available from social networks of professionals should also provide an additional, much needed tool for improving the quality of registers upon which CDH surveys are based.
More hypothetically, they could also provide complementary and timely information on key phenomena of interest. Further work should be carried out to assess the statistical properties of such sources. Throughout the project, a number of topics have raised particular interest among the participating research teams, pointing to future areas of survey development for testing. Indeed, the changing economic environment, the increasing diversity of career patterns and the changes in the organisation of the research landscape may require the use of a different and broader set of skills. This dimension needs to be measured and analysed with the appropriate tools.
The revised methodological guidelines and model questionnaire include proposals for capturing information that is relevant to these questions. CDH data can also provide a useful tool for analysing the contribution of doctorate holders to entrepreneurship. There is increase interest in the phenomenon of academic entrepreneurship; and observers have also noted the importance of doctorate training for individuals who started, but never completed their doctoral studies as they chose to develop their inventions by starting up new businesses.
In conclusion, existing CDH data should be made more openly available for analysts and decision makers in order to judge which analysis, findings and recommendations can be supported by evidence.
It is only through continued effort, drawing from best practices in the compilation of labour, STI and skills statistics that a robust evidence base will ultimately facilitate an improved understanding of the role played by doctorates in the innovation system and the broad knowledge economy.
Potential Impact:
Various facets of the KNOWINNO project have been discussed and disseminated at meetings of the OECD Committee for Scientific and Technological Policy and the Working Parties of Innovation and Technological Policy (TIP) and Research Institutions and Human Resources (RIHR) for policies and National Experts on Science and Technological Indicators (NESTI), in addition to the CIIE’s Working Party of Industry Analysis (WPIA) in the case of the work on patent indicators and analysis.
In doing so the three strands of the project have helped advancing both the policy discussion and the measurement agenda at OECD.
The results of the INNOSERV project (WP1) can be found in the dedicated website: www.oecd.org/sti/innoserv. This project has provided a series of recommendations for improving measurement in this area which will help advance the innovation measurement agenda at OECD. Some of the recommendations point to the need to better clarify the distinction between extramural and intramural R&D. There is also a need to clarify the definitions and boundaries of R&D in surveys and how they relate to service firms/activities: e.g. the distinction between service sector R&D versus service sector innovation. Better guidance is needed regarding the way in which to collect and report R&D statistics by industry. Implications for improving innovation surveys were also highlighted. As well as priority areas for future work, namely: Measurement of design and other knowledge-based capital; Public procurement as a tool for fostering service innovation; Public sector innovation; Measurement of policies and framework conditions.
The results of the KNM (WP2) project can be found in the dedicated website: www.oecd.org/sti/knowledge. This project has helped build up awareness among the international community of policy makers on science, technology and innovation of recent developments in the marketplace for technology and associated IP rights. This has so far been reflected in the 2012 Science, Technology and Innovation Outlook, a flagship publication of the OECD. Discussions on knowledge markets and technology transfer are becoming a more pervasive feature in country innovation reviews and policy discussions. The added emphasis on the dimension of knowledge flows through highly qualified personnel has also been welcome by OECD committees. From the measurement perspective, the project has identified a number of relevant indicators that will be reflected in the 2013 STI Scoreboard. It has also highlighted a number of promising avenues for developing OECD guidelines on measuring R&D and innovation. The results of the KNM project have fed into the OECD horizontal project on Knowledge Based Capital and will provide a sound basis for future OECD work in this area following the Ministerial meeting to be held in May 2013.
The results of the CDH project (WP3) can be found in a dedicated website: www.oecd.org/sti/cdh. The project has resulted in the production of updated methodological guidelines, comprising a revised model questionnaire for countries to use when collecting statistics on doctorate holders and thus ensure the internationally comparability of the results. Throughout the CDH project, a number of topics have raised particular interest among the participating research teams, pointing to future areas of survey development for testing. By demonstrating the feasibility of carrying out a wider range of possible analyses among a reduced set of countries, this project has also attempted to make a strong case for improving researchers’ access to data on doctorate holders with a view to promoting the coordinated use of micro-data across countries and conduct more in-depth analyses of the factors shaping career choices, labour market mobility and individual performance, linking whenever possible to other available databases. Overall the project has demonstrated new uses for the available data, including additional indicators but also for more structural types of analysis; shown the importance of complementing CDH surveys with existing surveys, such as studies of university graduates (as in Japan, United Kingdom and France) and data from registers, etc. It has also strengthened the case for granting researchers access to data, under appropriate arrangements that protect confidentiality, and for linking CDH data to other databases. The project has also engaged actively with policy decision makers to demonstrate the potential of the data and secure feedback on its relevance.
The expert meetings (three per work package) helped to attract top level participants and build a community of interest around these subject areas which will last well beyond the project itself. Each project benefitted from a dedicated final event, which were all well attended and helped discussing the conclusions of the studies. In particular the KNM final conference held in November 2012 in Paris was a high level event drawing together over 130 participants, among researchers, policy makers and practitioners from business and public sector organizations. The conference was also webcasted and the video can be found on www.oecd.org/sti/knowledge.
List of Websites:
www.oecd.org/sti/innoserv - www.oecd.org/sti/knowledge - www.oecd.org/sti/cdh