Assessing recent developments in innovation measurement: the third edition of the Oslo Manual

Transcription

1 Science and Public Policy, 34(1), February 2007, pages DOI: / X190487; Assessing recent developments in innovation measurement: the third edition of the Oslo Manual Carter Bloch This paper assesses developments in innovation measurement that have been incorporated into the recent revision of the Oslo Manual. Among the most central points emphasized in recent innovation literature are a need for: data that is adequate for the analysis of innovation systems and for policymaking; and a suitable measurement framework to capture the substantial innovation activity within services. The new measurement framework, which includes an expansion of the innovation concept beyond technological product and process innovation and a more complete treatment of linkages and knowledge flows, provides greater coverage of these topics. However, while these changes are a significant step forward in innovation measurement, there are a number of areas in which survey guidelines can be improved further, among them human resource development and the role of demand in innovation. The paper discusses some options for building on these changes in the future. F ROM THEIR BEGINNINGS in the 1980s, the measurement of firm innovation activity has grown at a rapid pace. Innovation surveys are now conducted in a broad range of countries, including those of the European Union (EU), other Organization for Economic Cooperation and Development (OECD) economies and a growing number of non-oecd ones. The formation of international guidelines in the OECD/Eurostat Oslo Manual (OECD, 1992; OECD/Eurostat, 1997; 2005) has had a substantial influence on these developments, both in terms of survey type and content. The first edition of the Oslo Manual, published in 1992, was a synthesis of the experiences from a broad group of innovation surveys in the late 1980s, providing a standardized framework for collecting firm-level data on technological product and process (TPP) innovation in manufacturing industries (OECD, 1992; see also Smith, 1992). This framework Carter Bloch is at The Danish Centre for Studies in Research and Research Policy, University of Aarhus, Finlandsgade 4, 8200 Aarhus N, Denmark; carter.bloch@cfa.au.dk; Tel: The author has greatly benefited from comments from the editors William Page and Susana Borrás and two referees. was later updated in the second edition in 1997 to include innovation in service sectors. These guidelines and the surveys based on them have made important contributions to our understanding of innovation and innovation processes, which, previously, was primarily based on R&D and patent data and ad hoc surveys. However, a number of papers 1 have pointed to their limitations and the neglect of a number of areas of importance to understanding innovation. Among the most central of these critiques are: an inadequate coverage of innovation in services; a lack of focus on non-technological innovation; and a lack of information on the dynamics of innovation from a systems viewpoint. 2 The recently completed third edition of the Oslo Manual (OECD/Eurostat, 2005) has undergone a number of substantial changes, with the aim of keeping innovation measurement abreast of policy needs and developments in the innovation literature. Among the most important changes are: a broadened definition of innovation to include marketing and organizational innovations; a much expanded coverage of knowledge flows and the role of linkages in the innovation process; and an adaptation to reflect increased awareness of innovation in less R&Dintensive industries, such as services and low-tech manufacturing. Science and Public Policy February /07/ US$08.00 Beech Tree Publishing

2 Carter Bloch is an assistant professor at the Danish Centre for Studies in Research and Research Policy, University of Aarhus in Denmark. His research concerns innovation measurement, service innovation, and the relationship among R&D, innovation and productivity. He is main draft writer (with Peter S Mortensen) of the third edition of the OECD/Eurostat Oslo Manual. Current projects include innovation indicators (NIND), service innovation (ServINNo) and innovation and productivity (IGNORed), all of which are supported by the Nordic Innovation Centre. The purpose of this paper is to discuss and assess the main changes that have been undertaken in the new Oslo Manual, and to examine how well the new measurement framework addresses recent developments in our understanding of innovation processes. The next section discusses recent developments in innovation theory and their main implications for innovation measurement. We then review recent survey experience and discuss the main changes to the Oslo Manual. Finally, we assess how well these changes are able to meet research and policy needs, and discuss possible areas where innovation measurement could be improved or expanded upon. Conceptual issues It has long been recognized that firms do not innovate in isolation. Influential works such as Nelson and Winter (1982) and Kline and Rosenberg (1986) highlighted the role of interaction among actors in the development of knowledge and technology. Kline and Rosenberg s chain link model of innovation provides a conceptual framework for the first and second editions of the Oslo Manual. It models innovation in terms of the interaction between the firm and other actors, where the completed innovation is the result of a (nonlinear) process involving testing, feedback and subsequent redesign. Systems of innovation approaches (for instance, Lundvall, 1992; Nelson, 1993; Edquist, 1997; more recently, Lundvall et al, 2002; Edquist, 2005) focus on how these interactions operate and on the role of cultural, organizational and institutional factors in affecting innovation. The concept of national innovation systems has become very widely used as a perspective both in thinking about innovation and in analyzing science and technology policy. Viewed in terms of the share of overall S&T policy funding, R&D support is still by far the major instrument, implying that S&T policy still generally adheres to a linear model of innovation. However, a number of policy initiatives have been implemented (such as those involving industry science relationships, technology transfer and innovation incubators) that focus on interactions and improving firms and other organizations framework conditions. These policy measures are relatively small measured in expenditure, yet they provide an indication of the growing emphasis in policy circles on viewing innovation from a systems perspective. The Aalborg school (Lundvall et al, 2002) places special emphasis on the factors that influence relationships among actors, such as norms, habits and rules, though there are a number of other factors in national innovation systems that may impact innovation, including laws and regulations, market demand, and institutional organizations. Innovation systems Edquist (1997; 2005) argues that a definition of national innovation systems should include all important economic, social, political, organizational, institutional and other factors that influence the development, diffusion and use of innovations, in part because we do not know enough about how these factors affect innovation and which are the most important determinants of innovation processes. There is a need for systematic analysis of activities in innovation systems. Edquist (2005) provides a list of these activities, which include demand-side factors, competence-building activities, innovation support organizations and other institutions: research and development (R&D); competence building in the labor force; formation of new product markets; user feedback and interaction; innovation support organizations (entrepreneurship, public research, technology transfer, consulting, and so on); networks (formal and informal); institutions (laws and regulations); financial and administrative institutions. The concept of innovation systems has subsequently been applied to levels other than the nation state, in particular to regional systems of innovation (for instance, Asheim and Gertler, 2005; Asheim and Isaksen, 2002; Cooke, 2001; see also Porter, 1998) and sectoral systems of innovation (for instance, Breschi and Malerba, 1997; Malerba, 2005). The basic motivation for these alternative or additional concepts is that there may be a number of factors that are specific to regions or sectors, thus making it relevant to analyze how innovation processes evolve at these levels. An additional important argument for examining innovation systems at these levels is that much policy may be conducted or targeted at the regional or sectoral level, making regional and sectoral innovation systems useful from a policy perspective. Borrás (2004) considers to what extent the EU can be conceptualized as a system of innovation. There is wide variation in cultural factors, which, among others, the Aalborg school identifies as being central to national innovation systems. However, some social and cultural values are being formed at the EU level, in particular concerning issues of risk and the role of science in social progress. On the other hand, 24 Science and Public Policy February 2007

3 there are a number of policy organizations that operate at the EU level, implying that examining the EU as a system of innovation has great relevance for policy. Organizational practices The role of organizational practices and organizational change in innovation processes has received increasing attention. A number of studies have, for example, identified the importance of organizational change for the successful introduction of process innovations, particularly those that are related to information technology (IT) (for instance, Brynholfsson and Hitt, 2000; Gera and Gu, 2004; Murphy, 2002). Lam (2005) discusses literature on organizational innovation, focusing on three perspectives: the relationship between organizational design and innovativeness, the process of organizational learning and knowledge creation, and the capacity for organizational change and adaptation. She contrasts two forms of innovative organizations, J-form and adhocracy. The J-form is more formally structured and relies on the knowledge and learning capacity of the organization as a whole through project teams, job rotation and shared cultures. Adhocracy is a more fluid organizational structure and relies on individuals knowledge and creative capabilities. Organizational structure is more decentralized and a large degree of discretion is given to the individual employee. The benefits of different types of practices and structure may to a large degree depend on the industry and types of innovation needed. Efforts have also been made to better integrate insights from organizational literature with work on innovation and innovation systems, focusing on learning and competence building, both at the firm and innovation systems level. For example, Lam and Lundvall (2006) argue that learning should take a more central role in innovation systems, focusing more on education systems and labor markets. Globalization Globalization has had a significant influence on innovation. Firms have much greater access to information and markets, and are much more easily able to undertake joint projects with firms in other countries. Globalization has at the same time meant increased international competition, making innovation vital for firms in all industries. For many firms, competition on price and efficiency is not enough. They also need to compete on product characteristics and marketing methods. Multinational enterprises (MNEs) play a substantial role in the globalization process, and are responsible for a large share of international knowledge flows through patenting, licensing, foreign direct investment (FDI), trade and scientific collaborations (Narula and Zanfei, 2005). Their operations will With globalization, firms have greater access to information and markets, and can more easily undertake joint projects with firms in other countries: it has also meant increased international competition, making innovation vital for firms in all industries often cross national boundaries, with R&D, production and sales being conducted in different countries. As a consequence, countries may be better able to benefit from foreign R&D and other innovation, but may also be less assured of reaping the full gains from domestic innovation activities if production is conducted in other countries. Furthermore, MNE outsourcing and off-shoring activities mean that countries are under increasing competitive pressure to provide the most attractive environments for MNEs innovation activities. Innovation in services One of the most debated issues in innovation surveys is coverage of innovation in services. While there is an increasing number of studies that examine innovation in services, it is argued that almost all this work draws on data from a measurement framework designed for manufacturing firms (Gallouj and Weinstein, 1997; Tether, 2005). This assimilation approach may be partly one of convenience and partly based on the view that innovation in services is similar to manufacturing, where the majority of service firms are assumed to be non-creative adopters of technology (Tether, 2005). Approaches in the literature suggest two ways of improving coverage of services. One group of research (for instance, Gallouj and Weinstein, 1997; Sundbo and Gallouj, 2000) argues that innovation in services is fundamentally different from that in manufacturing, with the implication that service firms should be surveyed separately from manufacturing. The synthesis approach (Coombs and Miles, 2000; Drejer, 2004; Tether, 2005) argues that innovation in services and manufacturing are not fundamentally different, but that both may emphasize different aspects of the innovation process. This would then argue for a survey framework that is equally suited to both sectors. Arguments that service innovation is not well captured in a technological product and process framework stress a number of other aspects that are important to innovation activities in service firms. Innovation can be seen as a continuous process, Science and Public Policy February

4 consisting of a series of small, incremental changes (Tether, 2005), as opposed to larger, discrete changes. This arguably makes it more difficult to identify service innovations as product innovations. New knowledge and technology are more likely to be embedded in workforce skills in service firms than in more measurable sources such as R&D and advanced equipment. Services also tend to involve greater interaction with customers, both in their development and provision, and a greater degree of customization to the individual customer (Green et al, 2001). Services are often characterized as non-separable (Lovelock and Gummesson, 2004), with production and consumption occurring simultaneously. It is argued that this makes it difficult for service firms to distinguish between product and process innovations. Finally, organizational change is argued to have a central role in innovation activities for service firms, as changes in services will often involve changes in work procedures and practices. Role of demand The role of demand in innovation has also been the subject of increased policy interest (see, for instance, Aho group report, 2006; Nordic Council of Ministers, 2006; CEC, 2006). However, demand is a broad concept that may influence, or play a role in, firm innovation in a variety of ways. Examining the marketing and innovation literature, we can identify four main aspects of how demand may affect innovation: the role of the user in innovation; the impact of market demand and market structure; understanding user needs; and utilizing market knowledge in the firm s innovation activities. User-driven innovation involves the use of knowledge of customer needs in the firm s innovation activities. To some degree, demand influences all innovations, as was clearly pointed out by Mowery and Rosenberg (1977) in their critique of a number of earlier studies of demand and supply influences. User-driven innovation should thus imply that user needs are a driving force in generating new ideas for product development. Lead-user innovation involves the participation of lead users in the actual development of new products (for instance, Lilien et al, 2002). In the leaduser process, interaction with customers not only concerns identifying user needs, but also seeking solutions for the development of new products. This attributes users a greater role as a linkage source, potentially also as a source of new technological knowledge. User innovation concerns the development of innovation for own use, either as final products or as intermediate inputs. Von Hippel (2005) argues that users are a very important source of innovation and the creation of new knowledge. Among the implications of this view are that policy should not focus on a few individuals, as it is hard or impossible to predict where new ideas will originate. In addition, the benefits of open innovation increase when we take into account that many new ideas may come from users. Another aspect concerns how firms collect and utilize information on demand in their internal innovation processes, that is, methods used by firms to analyze market demand and the flows of this information within the firm. Market analysis can range from traditional techniques that examine identified needs to newer research methods that involve identifying customers latent needs. An example of newer techniques is ethnographic design research (Gilmore, 2002), which involves deeper research into potential customer needs through observation of their habits, routines, views and preferences. As opposed to examining average user preferences or needs, its aim is to uncover idiosyncrasies and new perspectives. The focus of this type of research is on customers experiences as opposed to simply product characteristics and functions (Suri, 2003). This type of design method can have important implications for innovation processes. It places a clear focus on the user (and research conducted on the user) as the source of new ideas, charting the direction for product development and potentially also that of general research on new technologies. Griffin and Hauser (1996) analyze and review work on R&D-marketing interaction, examining both barriers to communication and cooperation across functions, and organizational approaches used to enhance integration of marketing and R&D. Among these approaches are: relocation and physical facilities (that is, keeping marketing and R&D activities physically close to one another); personnel movement (such as job rotation); informal social systems; organizational structures (such as coordinating groups and project teams); incentives and rewards; and integrative management processes. More general characteristics of market demand and market structure are important aspects of innovation systems, both through their influence on incentives to innovate and for knowledge flows. Hauser et al (2006) refer to the concept consumer innovativeness as the propensity of consumers to adopt new products. Consumer innovativeness is an important determinant of the rate of diffusion, and of firms propensity to innovate, since the more responsive customers are to new products, the greater the incentives and return to innovation. Consumer innovativeness in home markets may also be important for development and testing. Even in cases in which domestic markets are relatively unimportant in terms of share of sales revenue, they may be essential for the development and testing (market and otherwise) of new products. Network externalities (Hauser et al, 2006) refer to demand that is affected by the number of users with similar products or the availability of related products. Hence, for new product areas, innovation 26 Science and Public Policy February 2007

5 may be very dependent on the existence of a market for the product. Standards, platforms and regulations may also play an important role in establishing demand for certain product groups. Implications for innovation measurement What are the implications of recent innovation literature for the collection of innovation data? Here we draw out some main points from the discussion above that can be used to assess the revision of the Oslo Manual and recent innovation surveys. Measuring frameworks need to be able to capture adequately the substantial innovation activity in services. Concerning the question of whether to collect data for services and manufacturing jointly or separately, an issue of great significance is that it is in practice extremely difficult, if not impossible, to isolate service activities. Boundaries between service and manufacturing have become increasingly blurred. Many manufacturing firms are also active in services, both in support of their goods, and as independent products. Likewise, many service firms sell products that also include tangible goods. This duality in goods and services innovation is an important topic for innovation studies and, at the same time, necessitates a single, synthesized measurement framework. Surveys need to collect data that is both adequate for analysis of innovation systems and for policymaking. This has also been a central critique by Salazar and Holbrook (2004) of the 1997 Oslo Manual (and surveys based on it), that it devotes too little attention to the diffusion of knowledge, to linkages and to learning, training and absorption capacities. Related to this, data on regional factors are important for both regional or clusters analysis and the formation of regional innovation policy. It is increasingly important to obtain data on globalization and MNEs. At a minimum, surveys should be able to identify MNEs and their subsidiaries among respondents, along with international linkages and knowledge flows. A more complete understanding of how firms innovate needs to go beyond technological innovation and examine the introduction of other (nontechnological) innovations. The purpose here is to examine both the importance of non-technological innovation activities and how they are linked with (technological) product and process development. More data is needed on the role of users and demand in innovation, both for policy-making and to increase understanding of the impact of market factors. This area is relatively unexplored, though the discussion above suggests some possible directions for future measurement. Recent innovation surveys As noted above, there has been a considerable increase in recent years in the number of countries that have conducted innovation surveys. 3 These surveys represent the most recent experience in innovation measurement and thus have had a significant influence on the Oslo Manual revision. This section discusses recent innovation surveys, including experience with organizational and marketing innovation. The Third Community Information Survey, CIS3, for EU countries covered the period and included manufacturing and services firms using the same questionnaire. Following the 1997 Oslo Manual, it covered TPP innovation, although it also included a group of questions on organizational and marketing changes. Recent surveys in Australia, Canada and Japan have a number of similarities to CIS3, but also included greater coverage of organizational aspects such as human resources and knowledge-management practices. All these surveys cover linkages to other actors and sources through questions on information sources for innovation, active innovation cooperation, and the acquisition of external knowledge. For example, CIS3 included a question on the importance of various information sources and on the type and geographic location of innovation cooperation partners. The Australian Innovation Survey 2003 included organizational innovations as part of the overall innovation concept. 4 The questionnaire also included additional questions on linkages (in comparison with CIS3), asking about types of collaboration activity (for instance, marketing, production, R&D) and about methods used to acquire knowledge from other firms and public research institutions (for instance, through new employees, consultants, new equipment, purchases/use of new research). The Canada Innovation Survey 2003 (for instance, Lonmo, 2005) included (in addition to linkages questions as in CIS3) questions on reasons for cooperating on innovation activities. The survey also included a question on different types of government There has been a considerable increase recently in the number of countries that have conducted innovation surveys: these represent the most recent experience in innovation measurement and thus have had a significant influence on the Oslo Manual revision Science and Public Policy February

6 support organization, thus allowing the examination of linkages to institutional organizations. Finally, the Canadian questionnaire had a question on the importance of a series of objectives and practices, including human resource and knowledge-management practices. The Japanese National Innovation Survey 2003 (Ijichi et al, 2004) models questions on nontechnological changes along the lines of CIS3, though with a much greater coverage in terms of the number of questions. This allowed individual questions to address more specific aspects, yielding greater detail on firms organizational changes. All these surveys included a question on the novelty of product innovations, and some also on process innovations. There are two approaches, one in terms of novelty with respect to geographic regions and another in terms of the firm s competitive market. There are advantages and disadvantages to both approaches. Geographic regions are equally understood by all, but may not fit well with the geographic market in which the firm sells its goods (Holbrook and Hughes, 2001; Salazar and Holbrook, 2004). The firm s market fits much better with its own competitive environment, but novelty will vary greatly from firm to firm depending on the size of its market. One way to strengthen novelty measures based on firms competitive markets is to cross the question with information on what markets the firm operates within (for instance, regional, national, international). There have been a handful of surveys focusing specifically on service innovation, among them the Italian survey of innovation in service firms 1995 (for instance, Sirilli and Evangelista, 1998) and the German survey of innovation in service companies 1995 (for instance, Hipp et al, 2000). The degree of standardization is often identified as an important characteristic of service firms and one that has significant influence on service innovation (for instance, Hauknes, 1998; de Jong et al, 2003). The German survey includes a question on standardization by asking firms to divide sales into standard services, partially customized services and oneoff custom-made services. Subsequent analysis (Tether et al, 2001; Hipp et al, 2000; Tether and Hipp, 2002) has found this to be an important factor of service firm innovation that can be used in capturing the wide variation among firms in the service sector. While the experience with organizational innovation is somewhat limited in general innovation surveys, it is now fairly extensive in specialized surveys (see, for instance, Wengel et al, 2000). These vary widely in their coverage and objectives, but some general areas that have been examined are: organization of the workplace; human resources development and management; external relations; management systems and strategies; and the use of information technologies. These surveys also generally examine both existing practices and recent changes. There is, in contrast, much less survey experience with marketing innovation. Beyond individual questions on marketing changes in CIS3 and other surveys, there are almost no examples of innovationrelated surveys devoted to marketing practices. 5 The recent CIS4 survey, while officially based on the 1997 Oslo Manual, incorporates a number of the changes from the more recent one. For example, the new wording for the definitions of product and process innovation are used, and questions on marketing and organizational innovation are included at the end of the survey (though these definitions vary somewhat from the final definitions arrived at in the new Oslo Manual). In addition, each type of innovation is divided into subtypes, for example, product innovations are divided into goods and services and process innovations into production processes, delivery or logistics processes and other support functions. This may allow for a variety of interesting analyses and interpretations. One example is to examine manufacturing firms that are active in service innovation. Coverage of other areas in CIS4, such as linkages (information sources and cooperation) is very similar to that in CIS3. Oslo Manual revision By far the most significant change to the Oslo Manual is the broadening of the overall innovation concept to include organizational and marketing innovations (OECD/Eurostat, 2005). An innovation is the implementation of a new or significantly improved product (good or service), or process, a new marketing method, or a new organizational method in business practices, workplace organization or external relations. [original emphasis] The overall definition of an innovation is modular; for example, we can still focus on the narrower concept of product and process innovation. However, the definition clearly reflects that firms innovation The overall definition of an innovation is modular, for example, we can still focus on the narrower concept of product and process innovation: however, the definition clearly reflects that firms innovation activities are more than product and process innovation 28 Science and Public Policy February 2007

7 activities are more than product and process innovation and that a broader view is needed to obtain a full picture of innovation. Types of innovation Operationalizing this definition for use in surveys presents a number of challenges, among them creating individual and distinguishable definitions for each type of innovation and adapting other innovation survey topics to the broadened innovation concept. The 1997 Oslo Manual defines technological product and process innovation as implemented technologically new products and processes and significant technological improvements in products and processes (OECD/Eurostat, 1997). The term technological is not defined in the 1997 Oslo Manual, which recognizes possible difficulties in interpretations (OECD/Eurostat, 1997: 22). The general interpretation of technological in the 1997 Oslo Manual is that it has to do with objective performance characteristics (OECD/Eurostat, 1997: 25 26). Subsequent innovation surveys have moved away from the use of the word technological, while, at the same time, retaining the focus on objective performance characteristics. 6 The purpose of these modifications is likely not to broaden the scope of the definitions (although it cannot be ruled out that they had this effect), but to make them less manufacturing oriented. This shift of emphasis from technological to characteristics or intended uses is meant to make the definitions more applicable to innovation in services. The Manual now also acknowledges that innovations are not restricted to single discrete changes, but may also consist of a series of incremental changes, provided these changes together amount to a substantial improvement (OECD/ Eurostat, 2005: 124; 151). 7 A product innovation is the introduction of a good or service that is new or significantly improved with respect to its characteristics or intended uses. This includes significant improvements in technical specifications, components and materials, incorporated software, user friendliness or other functional characteristics. [original emphasis] A process innovation is the implementation of a new or significantly improved production or delivery method. This includes significant changes in techniques, equipment and/or software. [original emphasis] In forming a definition of organizational innovation, a first issue is when organizational changes should be considered innovations. One criterion is the objectives of the change. Organizational innovations share the objectives of process innovations in that they may aim to enhance production or delivery capabilities through improvements in productivity, efficiency or quality, or by facilitating the production of new products. However, following Lam (2005), a broader view of the innovative firm also includes capabilities to create new ideas, use existing knowledge in new ways and transform knowledge into marketable products. 8 This suggests that organizational innovations should not only be linked to changes that aim to improve the production and delivery of existing goods and services, but also to those that aim to improve capabilities for developing new products and processes (OECD/Eurostat, 2005). An organizational innovation is the implementation of a new organizational method in the firm s business practices, workplace organization or external relations. [original emphasis] Organizational innovations are defined as consisting of one (or more) of three elements: business practices; workplace organization; and external relations. Business practices are routines or procedures for the conduct of work. These can range from practices for sharing knowledge to the sets of procedures involved in management systems. Workplace organization involves organizational structures and the distribution of responsibilities and decision-making, while external relations involve the organization of relationships with other firms or public research institutions. The 1997 Oslo Manual excludes aesthetic or other subjective changes to products from the definition of technological product and process innovation. While it is apparent that these types of change do not fit the objective performance characteristics criteria, it is also widely acknowledged that they may be very important to firm performance (OECD/ Eurostat, 2005). A marketing innovation is the implementation of a new marketing method involving significant changes in product design or packaging, product placement, product promotion or pricing. [original emphasis] Many firms devote substantial resources to product design and packaging, where their activities may involve market research, testing and the use of new knowledge. The same is also true for marketing practices, and many firms may focus on marketing and design in their efforts to improve competitiveness and performance. Product design and marketing practices have in common that they both may be aimed at improving the presentation and promotion of products. Drawing on this, the definition of marketing innovation is based on the 4P s model for marketing strategies (see, for instance, Perrault and McCarthy, 2005), that is, product, price, placement and promotion. Science and Public Policy February

8 Product design changes refer to changes in product form and appearance that do not alter the product s functional or user characteristics. They also include changes in the packaging of products such as foods, beverages and detergents, where packaging is the main determinant of the product s appearance. Product placement involves methods used to sell goods and services to customers. Promotion includes concepts for promoting a firm s goods and services, such as new advertising methods or new brand symbols. Pricing involves the use of pricing strategies to market the firm s goods or services. An important issue is how to identify significant changes in characteristics, in order to apply similar criteria to organizational and marketing innovations as for product and process innovations. This is addressed by requiring that marketing and organizational innovations consist of methods that have not been used before in the firm and are the result of strategic decisions taken by management (OECD/ Eurostat, 2005: 171; 179). A significant challenge in defining the four types of innovation is addressing how best to distinguish among them for borderline cases. Efforts are made in the Manual to minimize borderline cases, though it is not considered feasible or desirable to make clear-cut distinctions among types. Distinctions will often depend on the nature of the firm s business and on the specific characteristics of a firm s innovations. Often, even if it is hard to make a clear-cut distinction among the firm s activities (that is, product development, production and delivery processes, organizational practices, marketing), it may be less difficult to distinguish among types of innovation. An example is product and process innovations in services. As noted above, it is often argued that production and consumption are frequently blurred in services, making it difficult to distinguish between product and process. However, when viewing the characteristics of the actual changes made, service firms may still be able to identify whether the changes are product innovations (the characteristics of the service that the customer experiences are new or significantly improved) or process innovations (the firm makes significant improvements to how the service is provided, possibly without changing the service s characteristics) or both. Linkages Both the first and second editions of the Oslo Manual (OECD, 1992; OECD/Eurostat, 1997) recognize the importance of linkages for innovation. The first edition lists a number of different sources of external knowledge, including internal sources, commonly available sources of codified knowledge, acquisition of embodied and disembodied technology, and cooperation with a number of external actors. These sources could be examined as sources of innovative ideas that contribute to the initiation of innovation projects, or as factors contributing to the success of innovative projects. The first edition also discusses the outbound diffusion of technology through the sale of products to user sectors, and R&D cooperation. These same topics are included in the second edition, though with some modifications. The two lists of factors are consolidated to a single set of sources of information for innovation, where no specification is given as to how the information is obtained. For example, firms and public institutions can be information sources either through passive (one-way) transfer of knowledge or through active cooperation. In the years following the publication of the second edition of the Oslo Manual, innovation surveys 9 have built on this framework and included separate questions on information sources and on innovation cooperation, 10 a topic that is not formally discussed in the second edition. These developments, combined with an increasing awareness of the importance of linkages, motivate a deeper and more extensive discussion of linkages in the third edition. A separate chapter is devoted to linkages: it presents a coherent framework in which linkages are characterized by their source, cost and level of interaction. Three types of external linkage are identified: open information sources; acquisition of knowledge and technology; and innovation cooperation. Open information sources provide access to knowledge without the need to pay for the knowledge itself, although there may be marginal fees for access (membership in trade associations, attendance at conferences, subscriptions to journals) (OECD/ Eurostat, 2005). This type of linkage involves the transfer of codified knowledge, though some open sources, such as attendance at fairs or exhibitions, can give access to some tacit knowledge through personal interaction with other participants. Acquisition of technology and knowledge involves the purchase of external knowledge and technology without active co-operation with the source (OECD/Eurostat, 2005). External knowledge can be embodied in machinery or equipment, in new employees, or in the use of contract research and consulting services. Disembodied technology or knowledge also includes other know-how, patents, licenses, trademarks and software. Innovation co-operation involves active participation in joint innovation projects with other organizations (OECD/Eurostat, 2005). Innovation co-operation allows enterprises to access knowledge and technology that they would be unable to utilize on their own. It can take place along supply chains, involving customers and suppliers in the joint development of new products, processes or other innovations, or it can involve horizontal collaboration, with enterprises working jointly with other enterprises or public research institutions. 30 Science and Public Policy February 2007

9 This framework, which forms the core of the chapter on linkages, mainly deals with external linkages and primarily with inbound diffusion, with the exception that cooperation involves the two-way flow of knowledge. The third edition also discusses outbound diffusion, though this is more difficult to measure as firms will generally be much less able to assess the impact of knowledge flows outside the firm. One possible option in this area is to collect information on the users of a firm s innovations. The Manual suggests the following classification: consumer markets; - domestic - foreign inputs to other firms; - domestic (inside/outside enterprise group) - foreign (inside/outside enterprise group). Knowledge management An additional area covered in the Manual is knowledge-management practices. Knowledge management involves activities relating to the capture, use and sharing of knowledge by organizations. It thus encompasses the management of both external and internal linkages, including methods and procedures for seeking external knowledge and for managing relationships with suppliers, customers and research institutions. A number of surveys on knowledge management practices have been conducted in recent years, most notably in Canada (for instance, Foray and Gault, 2003). In addition, questions on knowledge management have been included in innovation surveys, for example, the Japanese NIS 2003 Survey, the French CIS3 and the Australian Innovation Survey In general, the Oslo Manual devotes less attention to internal linkages. Knowledge-management practices are one exception, as they deal with both the internal flow of information and coordinating within firm activities to improve the capture and use of external knowledge. Discussion and conclusion The revision of the Oslo Manual, in varying degrees, touches on all the main criticisms and needs expressed by researchers and policy. By broadening the innovation concept, it opens up for examination a wider range of innovation activities. This section assesses these changes and examines where more progress can be made. In considering this, we look both at potential changes or additions to survey questions and also at what more can be done to get the most out of existing data or within this new framework. A number of areas of innovation literature, such as innovation systems and service innovation, argue for the need to broaden the innovation concept. A number of areas of innovation literature, such as innovation systems and service innovation, argue for the need to broaden the innovation concept: the introduction of marketing and organizational innovation in the new Oslo Manual comprises an important step Thus, the introduction of marketing and organizational innovation in the new Oslo Manual comprises an important step in this direction. It allows both the examination of firms primarily active in soft types of innovation and the links among different types of innovation activity. Pilot modules to the upcoming CIS2006 include, for example, questions on whether organizational and marketing innovations are linked to product and process innovations. However, as we will discuss in greater detail below, questions on organizational and marketing innovation are likely not to address fully the need for broader data. For example, interest in organizational innovation is just as much directed at characteristics of existing organizational practices as the implementation of new organizational methods (that is, organizational innovations). An example of this is the analysis of the relationship between work organization and innovativeness by Arundel et al (2006). In the same way, interest in commercialization is not limited to innovation in marketing methods, but also includes the variety of marketing and organizational practices firms may use to conduct market research and to utilize market knowledge in their product development activities. One potential problem that would need to be overcome in covering both existing practices and innovations is avoiding confusion between the two. There is some experience with including questions on both the existence of organizational practices and whether changes have been made to them. 11 However, no explicit criteria are made for these changes to be considered as innovations, in contrast to the definitions of organizational and marketing innovations. While this is an issue that would require careful consideration in survey design, it would not seem to be insurmountable. While the new Oslo Manual provides more expanded and coherent coverage of linkages, in terms of actual survey implementation, it is mainly keeping up to date with recent surveys. For example, CIS3 and CIS4 essentially already follow new Manual guidelines. However, it should be recognized that recent surveys already provide a wealth of information on linkages, their importance, whether Science and Public Policy February

10 they involve one-way flows or active cooperation, and their geographic location, that has yet to be fully explored. For example, what types of linkage are most important and how are they related to the development and implementation of innovations? Further evaluation of this data is needed to assess the usefulness of these indicators and whether they could be improved upon. Survey coverage of diffusion is generally limited to inbound diffusion. 12 Questions on outbound flows, while probably difficult to measure in a specific way, could very feasibly be included in a general sense as suggested by the new Oslo Manual (see above). This type of information would allow the identification of the structure of demand for firms innovations. Innovation processes may differ significantly between firms that primarily sell products to private consumers and those that have other firms as their main customers. It may, for example, be much more straightforward for firms to assess the user needs of other firms than those of private consumers, and firm clients are more likely to be closely involved in product development. An additional aspect of linkages that has not been covered in recent surveys is at what stage of product development they are utilized. Firms may rely on a variety of internal and external sources for ideas on new products and on contributing towards the actual development and implementation of product innovations. An example is industry science relations (see Polt et al, 2001). Public research institutions can potentially contribute to firm innovation as suppliers of new research results, or they may be actively involved in the development and testing of new products. Information on which stages of the innovation process public research contributes to, can aid in targeting research policy. An additional example concerns user-driven innovation and the role of market interaction. Of interest here is the role of demand in the innovation process, for example, the extent to which product innovations are initiated by user needs as opposed to technological developments and also how relatively important demand sources are for the development process and subsequent market introduction. The upcoming Danish CIS 2006 questionnaire includes a question on the usage of linkages at different stages of product development. The question draws on the Carnegie Mellon R&D survey (Cohen et al, 2002) and is also related to the earlier classification of sources for innovation in the first edition of the Oslo Manual (OECD, 1992). Regional innovation systems literature points in particular to three areas of relevance for innovation surveys: linkages to regional actors; linkages to regional institutions (and their importance); and the importance of regional markets for innovation activity. In fact, most surveys already examine this geographic dimension to some extent, though not necessarily for small enough areas that would be considered as regions or clusters or in sufficient detail to analyze regional innovation. While there is a great deal of coverage of linkages with other actors in recent surveys, less attention is given to linkages with institutional organizations. The question on public innovation support organizations in the Canadian Innovation Survey 2003 offers a feasible method that could be used in other surveys, and where a number of country-specific organizations or programs could also be included. Recent literature on innovation systems emphasizes the role of learning and competence building. While one important determinant of firms absorptive capacity, R&D, is included in innovation surveys, the Oslo Manual offers little guidance on firms human resources 13 and activities to build human resource capabilities (for instance, training, skills development, job rotation). An example of possible approaches to cover this is the Canadian Innovation Survey 2003, which includes a question on the importance of a series of human-resource management practices. Related to organizational and human-resource management practices are knowledge-management practices (Foray and Gault, 2003) for which, as noted above, there is some survey experience. However, all three of these areas are very interrelated, implying that any consideration of these issues should be done jointly. A number of modifications have been made to the innovation measurement framework in the new Oslo Manual to make it applicable to both services and manufacturing. It would be useful to evaluate how well these changes perform in capturing innovation in services. The service sector includes an extremely diverse group of firms. An important issue is whether innovation surveys are able to capture this diversity. Two important dimensions in characterizing types of service firm are product standardization and the degree of client interaction (for instance, de Jong et al, 2003). While innovation surveys contain some information on client interaction, there is generally no information on product standardization, with the exception of the German service innovation survey. The question on product standardization included in the German survey may be of value for future surveys. A final question concerns whether the new measurement framework is adequate in characterizing how firms innovate. An examination of this issue should first look into ways of improving use of existing data. Tether (2001) and Arundel and Hollanders (2005) construct classifications of types of innovative firm using composite indicators based on data from several innovation survey questions. This line of work can be brought even further, in particular with the use of the broadened definition of innovation. 14 In addition, innovation surveys can ask about subtypes of innovations. This approach was followed in CIS4 and allows greater detail on types of 32 Science and Public Policy February 2007