Summary Innovation & Technology Dynamics 2012 Nelson & Winter (1977) - - - - Present theory of innovation it is too scattered and comes mainly from economists, who only focus on the production function. Economics try to explain that differences in productivity are caused by differences in R&D investment. Very linear view: more investments in R&D cause more productivity (see also Boekholt). Main problems with this view: o Specification. Productivity growth is also related to a host of other factors, for example the growth of the output of the industry. It is therefore wrong to assume that R&D investment is the only independent variable, there are more. o Even if the R&D – productivity growth relation is true, what explains then differences in R&D investment? Possible explanations: Institutional structure varies considerably between industries. Economists have quite rudimentary assumptions of institutional structure which do not take the differences between industries into account. R&D is more effective for technologies in certain industries than for others. Economic theory not a good starting point for innovation theory: o Innovation involves uncertainty, economists can’t cope with that o Institutional structure varies considerable between industries. Therefore, general R&D policy (such as tax credits) will not foster innovation. Theory of Nelson & Winter builds on two premises: every change is an innovation and it involves considerable uncertainty. There are considerable differences between industries. Therefore, any theory should incorporate the stochastic evolutionary (=uncertain) nature of innovation and leave room for organizational complexity and diversity. Innovation is path depended and occurs within technological regimes, which guides the search activities of engineers. o Generation of innovation: Factors on the cost side and on the demand side differ across industries, which influences innovation. Generation is seldom led by profit maximization goals, but by governmental subsidies, non-profit institutions, etc. It must be guided by an R&D strategy, with attention to both supply (what can be made) and demand (what does the market want). Natural trajectories or regimes: development is path depended, advances follow other advances. o Selection environment: how does the use of technologies change over time? Do other firms want to use it? Do consumers want to buy it? Market as selection environment. Can lead to expansion of the innovator and/or imitation by other companies. Also non-market selection: e.g. by universities. Firms/consumers are not so sharply distinguished here. A theory cannot predict future actions. For theory to be useful, it must organize knowledge and guide research. It can give some guidance in the search for a useful theory Van den Ende & Kemp (1999) - How do regimes change? New regimes originally develop in old ones, are geared to problems of the old regime, before it grows out of the old regime through a process of transformation. Novel technologies are produced on the basis of knowledge available in existing regimes. Initially, limited domain of application for new technology. Technological regime: rule-set embedded in the coherent complex of a technology which structures the search activities of engineers and the policies and actions of other actors. Technological niches: place where radical innovations emerge. They provide some protection to these innovations (incubation rooms), since their selection criteria are different. Socio-technical landscape: the whole of cultural values, macro economical developments, politics. Four themes in the discussion of the computer regime: o Variety in types of regime shifts: shift from an old regime to a new one, emergence of a nested regime, convergence of two technologies. o Contribution of the old regime to the rise of new ones: development of a new regime was promoted by developments in the old regime (in case of computers: growing tendency to schematize problems, development of more sophisticated punch card machinery). o Regime shifts are accompanied by social processes that both facilitate and constrain the transition process. Facilitated by e.g. emergence of SME’s, constrained by the discussion around the impact of computers, etc. o Path dependency created by old regime. Restricts the directions in which new one can go and defines to conditions the new regime has to fulfill if it wants to be successful. Geels & Schot (2007) - - MLP encompasses three levels: niche-innovations, ST regimes and ST landscapes. Critic on MLP Response by Geels & Schot Unclear how the analytical levels should be Level of organizational fields (totality of applied empirically relevant actors) Too much emphasis on niche as the place Niches and landscapes interact with the where regimes change regime, which leads to four transition pathways (see below) Neglect of agency (individual action) Actors are active, not passive, interpret rules, make sense and come to decision. Authors define pathways to overcome niche driven bias in the understanding of transitions. Transitions are changes from one socio-technical regime to another. Based on 2 criteria: o Timing of interaction: does landscape pressure occur when niches are fully developed or not. o Nature of interaction: Landscape developments: reinforcing (stabilizing effect on regime), disruptive (pressure on regime, creating change) Niche innovations: competitive (replacing regime), symbiotic (competence enhancing to the regime). - Four transition pathways: Transition pathway Timing of interaction Niche innovations have not yet been fully developed Nature of interaction Disruptive: moderate landscape pressure Reconfiguration Niche innovations are fully developed Symbiotic: competence enhancing. Technological substitution Niche innovations are fully developed Disruptive: landscape change is large and sudden. Through a specific shock. De-alignment and re-alignment Niche innovations aren’t fully developed Disruptive: landscape change is large and sudden Transformation - Result Regime actors will modify the direction of development and innovation activities. New regimes grow out of old regimes through cumulative adjustments and reorientations. Trigger further adjustments in the regime to solve problems. Can add up to major reconfigurations over time. Niche innovation will break through and replace the existing regime. Differs from de/realignment, since the regime is stable and there is a substitute available. Often leads to the downfall of incumbents. Regime problems, erosion of the regime. Everyone loses fait (dealignment). But: No clear substitute for the regime, multiple niche innovations compete. Eventually one becomes dominant (realignment). If landscape pressure is disruptive, then a sequence of transitions is likely: transformation reconfiguration substitution de/re-alignment. Schoonhoven & Jelinek (1997) - - Survival depends on the capability to deliver a continuous stream of innovation continuous change. This requires an attention to both flexibility and efficiency. The distinction between organic and mechanic structures by Burns & Stalker is too simple. Prevailing perspective: innovative organization ought to be unstructured, loosely couples adhocracies. However, the firms are not unstructured, key aspects of structure in successful companies. They have a clear organization and jobs: explicit formal structures, hierarchies and reporting relationships. Also clear job responsibilities. Product development is bounded by time horizons and milestones. How do these firms adapt and continuously innovate? Through: o - Frequent reorganizations (changing the formal structure of the organization to adapt to environmental change) and a dynamic tension between flexibility and efficiency (ability to be flexible through reorganizations and systematic enough to produce efficiently). Employees expect change. The change process is participative, TMT sketches major changes, but middle/lower management determine the structural change in their divisions. Such organizations are claimed to be self-designing organizations, which is possible because people are highly self-critical (leads to a willingness to reorganize) and because of awareness of the environment (to which context do you need to adapt). o Quasi formal structures: extensive use of committees, task forces, teams and dotted-line relationships (for information). Change frequently and are used for problematic circumstances that require additional managerial time. However, costs as well as benefits, because it takes time and energy. o Informal structures: strong norms and practices which guide interaction, regardless of position in the hierarchy. Collegial environment. Issue of company size: break down the company into several smaller business units. Amabile e.a. (1996) - - Central assumption: social environment influences creativity. Creativity = production of novel ideas. Focus on individual perception and the influence of those perceptions on creativity. KEYS model: to assess the perceptions of employees on the organizational work environment and which influence that has on creativity. Several aspects: o Encouragement of creativity: on three levels, organizational encouragement (of risk taking and idea generations, fair evaluation of ideas, rewards for creativity, participative management and decision making), supervisory encouragement (role of managers in for example goal clarity, open interactions, supervisory support) and work group encouragement (team member diversity, mutual openness to ideas). o Freedom: creativity is fostered when people have high autonomy, a sense of ownership and control over their own work. o Resources: resources should be allocated to projects to foster creativity. o Pressures: challenge has a positive, workload pressure a negative influence on creativity. o Organizational impediments to creativity: such as conservatism, rigid, formal management structures, etc. Results: high creativity projects scored higher on work group/organizational/supervisory encouragement, challenge, and lower on organizational impediments. No statistical evidence for, workload pressure and sufficient resources. Freedom was not significant in certain tests (only in phase 1, not 2 and 3). Thus these play a less important role in stimulating creativity. Ciborra (1996) - Platform organization: very organic and flexible. Organization is very fluid and constantly changing. The platform consists of an informal network or common ground, a shared identity, etc. It is a virtual and collective cognitive scheme. It is composed of practices, that consist of past routines, transactions and other organizational arrangements. - The platform stays the same, but the overlying formal structures change all the time. The platform facilitates the change in formal structures all the time. As such, the platform organization is ‘programmed’ to generate new structures and arrangements. Veryzer (1998) - - - Focus on the development of radical innovations. They do follow a logical process, but it is inherently different from incremental innovations. Phases of this innovation process: a. Dynamic drifting phase: exploration of various technologies. b. Convergence phase: convergence to highly discontinuous products. Driven by two forces: a product champion and a critical mass of contextual factors (e.g. company turbulence, resource availability, alliances, technology interactions that push development to the next phase). c. Formulation phase: focus on how to formulate the technology into a product (product requirements). Not driven by the customer, but focus on how to create a technical differential advantage. d. Preliminary design phase: what will the product be in terms of initial application. e. Evaluation preparation phase: design becomes much more specified. Commercial viability is researched. Project is reviewed to determine if it’s worth the funding. f. Formative prototype phase: very explorative prototype, used to determine product applications. It is during this phase that communication with key customers and target groups takes place. g. Testing and design modification phases. h. Prototype and commercialization phases: shift towards producing a commercial product. Transition of the product from R&D to an operating unit. Differences with incremental innovation process: very high degree of technological and market uncertainty, high degree of informality in which the process is managed, importance of product champions, less customer driven (focus more on creating a different product from a technological point of view, less market assessments and financial analysis) and the process is more exploratory. Why less customer driven? Market opportunities for discontinuous products are often unspecified and unclear. Customers not always understand the product. Ahuja & Lampert (2001) - How do large firms create breakthrough inventions? Organizational learning perspective: as firms improve competences to improve incremental performance, they reduce competences to produce breakthrough inventions. Three of such learning traps which constraint the ability to produce breakthrough inventions: a. Familiarity trap: focus on familiar technologies. Developing expertise with familiar knowledge bases yields more immediate and likely returns. Therefore, firms prefer this over unfamiliar technologies. Lack of exposure to unfamiliar knowledge and the reduction of experimentations reduces the chance of coming up with breakthrough solutions to problems and developing breakthrough inventions. i. Solution: exploring novel technologies (those are new to the organization). Provides heterogeneity in problem solving and you create more new-to-the- - world views. However, relationship has an inverted U-shape, since too much exploration of novel technologies is harmful (leads to information overload, confusion). b. Maturity trap: focus on mature technologies. Companies favor those, since they are usually well understood, offer great reliability, have a well developed value chain and offer legitimacy. Lack of exposure to immature technologies reduces the chance of developing breakthrough inventions. i. Solution: experimenting of emerging technologies. Technologies pose more problems in the early phases of their development than in the later phases. These represent significant opportunities, so larger change to develop breakthrough inventions in emerging technologies. However, also U-shaped relation. Exploring emerging technologies requires focus, resources, etc., too much focus creates the risk that you don’t develop anything at all. c. Propinquity trap: companies search for solutions in the neighborhood of existing solutions, because it is more reliable, people are risk averse and it saves scare resources (less time, money). However, large areas remain unexplored. However, many breakthrough inventions are invented exactly in those areas. i. Solution: experimenting with pioneering technologies. Completely new solutions or trajectory, not related to existing solutions. Jump to a completely new trajectory. Risky, but chance that at least some of them will be successful. No Ushaped relation (hypothesis 3A was not supported), because using such technologies may simply be down to involving people not directly involved in the field. This way, it does not lead to an information overload, but just worse financial figures. Critics: the technologies are pretty much the same, all are basically new technologies. This shows from the high correlation rate between emerging and pioneering technologies. O’Reilly & Tushman (2004) - - A balance between exploitation and exploration is necessary to sustain competitive. Exploitation and exploration require different structures, processes and cultures, and can therefore not be organized within the same business unit, with the same employees. Important role for top management: they emphasize the importance of valuing the business units as equally important, but top management has to make a difference in the way both are managed and be sensitive for the (conflicting) needs of both business units. The top-management makes the decision which activities are performed in the exploitative, and which in the explorative SBU. Skills of employees: more specialists, either exploration or exploitation. Therefore, also quite clearly defined roles. Alvarez & Barney (2001) - Alliances between large companies and entrepreneurial firms. a. Advantage large company: gain access to new technologies, talent and innovative capabilities b. Advantage entrepreneurial company: gain social legitimacy, access to financial resources. - - These alliances create economic value, but most captured by large firm. Reasons: difference in rate at which large firms learn about entrepreneurial technology and rate at which entrepreneurial firm is able to imitate organizational resources. Entrepreneurial firm facilitates this by explaining the technology in detail. Large firms have the resources to study product and processes in detail as well. For a entrepreneurial firm, very difficult to learn about competences large firm, because these have developed over a long period of time. They do not have the resources or the capabilities to fully understand and imitate the capabilities of the larger firm. How to reduce the risk entrepreneurial companies face: a. Do it alone. However, risky, costly and time-consuming. b. Slow the large firm’s rate of learning: by not giving full access to the technology. c. Using an alliance contract. Specify how technology will be applied. d. Building trust. Prevents the larger firm from exploiting the entrepreneurial company. e. Creating an alliance in which the entrepreneurial company creates a stream of new technologies, to foster a long term relationship. Not just rely a single technology. Hargadon & Sutton (1997) - - How do organizations develop innovative products? Through technology brokering. Exploit your network position, use gaps between industries (structural holes) to transfer resources from groups who have plenty of these, to those who need them. Act as a broker. Create new combinations of existing but previously unconnected ideas. How? Four steps: a. Access: fill in a gap in the flow information between industries, being able to see technological solutions in one area that are valuable in the other. Occupy a central position in multiple industries. Restricting the amount of ties between subgroups. This enables brokers to provide access of a subgroup to the network. Through access to multiple industries, possibility to generate new product innovations through brokering. b. Acquisition: routines to bring technological solutions into the organizational memory, where they are stored for possible future use. By talking and watching new clients, reading about industries, looking at and taking apart industry’s existing products. Ongoing process of learning from clients. c. Storage: how does an organization puts away information until it is needed. Routines for storing specific knowledge (collecting, looking at and talking about products. The knowledge is stored in products) and routines for maintaining and refreshing knowledge until it can be used (displaying products where other brokers can see them, constant conversations) d. Retrieval: how do brokers retrieve old technological solutions in forms that fit new combinations that they are creating. Analogical thinking (try to reframe the problem, generate alternative solutions), sharing problems of current projects with other people in the organization. Organization stimulates technology brokering through: a. Structure of work. Brokers should be faced with a continual flow of new problems. Don’t specialize in one industry. Form diverse team, so multiple experiences are involved. b. Norms for collaboration: stimulate knowledge sharing and that employees feel comfortable to ask for help. Not too arrogant, not too insecure. c. Reward system: Stimulate knowledge sharing and the helping of others. Based on informal reputation. d. Employee selection: screen them on cultural fit, knowledge and skills. Rely on what potential employees have shown they will do, instead of they say they can do. Von Hippel & Von Krogh (2003) - - Two dominant innovations models: a. Private investment: innovations are supported by private investments. Patents to provide some sort of monopoly. i. Problems: loss to society, since not everyone is free to use the knowledge. b. Collective action: newly created knowledge is made public. i. Problems: motivating potential contributors. You can also free-ride and wait until others create knowledge. Overcome by providing rewards, this is also why many societies provide monetary subsidies for basic research (compare with Boekholt). Authors propose a third intermediary one: private collective model, based on experiences in the open source industry. It is the middleroad, because: a. It differs from private investment: innovators freely reveal the knowledge they created. Freely revealing is not a loss of private profit, costs might still be lower than revenues, since diffusion costs are low (because of internet). Users are innovators instead of manufacturers. b. It differs from collective action: in collective action much attention towards recruiting contributors, no active recruiting in private-collective. Free-riders do not have the same rewards as contributors. Considerable more private rewards for those who contribute, such as learning, enjoyment, sense of ownership, control, etc. Sharif (2006) - - Analyzes the development of national innovation systems: system of interconnected institutions to create, store and transfer the knowledge and skills which define new technologies. Aims to understand competitiveness on the country level and explaining differences between the two. The framework in which governments implement policies. Sharif indentifies and aims to solve several ambiguities regarding NIS: a. Ambiguity surrounding the origin: both political and academia, not one of both. b. Attack on mainstream economics: NIS researchers claim that economics devote a lack of attention to knowledge, technology and change. Innovation can’t be understood with a theory that assumes equilibrium (as economics do). c. Community or informal network was used to challenge the dominance of economics. d. Flexible interpretations of the NIS concept: initially it was introduced as a macro level phenomenon, to compete with dominant economic views. Later on, also regional and sectoral innovation systems. e. Disagreement surrounding over-theorization of NIS: some scholars claim NIS needs more theory (no concrete operationalizations), others claim that fuzziness should be kept (then it can be applied to much more). f. Disagreement on the presence of NIS in developing countries: depends how you look at it. Is it a mechanism for only generating new technologies (small view) or also for diffusion of technology? (general view, than every country has one) g. Importance of wider context: got more important due to globalization and MNC’s. h. Refutation on linear model: also see Boekholt. Lundvall (2007) - - - Argues that policy makers and scholars have applied a too narrow focus in NSI. Only focus on STI, not on DUI learning (see below). Two modes of innovation: STI learning (science based: as the first step towards technology and innovation) and DUI learning (experience based: learning by doing, using and interacting) Claims that the neglect of learning as competence building in standard economics is a major weakness, which makes it less relevant for studying innovation. Claims that developing a general theory of NIS would limit its application (see also Sharif). Analysis of innovation systems is focused on the combination of innovation and learning Theory underlying innovation systems: learning processes involving skilful rational agents, who enhance their competence through searching and learning. They interact with other agents, the result is reflecting in the form of new innovations and competences. a. Therefore, it is important to understand how learning takes place within as well as between organizations, if you want to know how innovation systems work. Weak correlation between STI-learning and innovation: countries scoring high on STI, might still score low on innovation. This is due to the fact that these countries score low on DUI learning, e.g. limited organizational learning within and between firms. Research shows that firms who combine the two modes (DUI and STI) are more innovative. IS can be studied by distinguishing between the core of the system (firms and the knowledge infrastructure) and the wider setting (education, access to finance, demand, government). Some trends where NIS research might go: a. Needs to be adapted to the situation in developing countries, influence of power of certain parties on innovation, relationship between NIS and globalization. b. Role of the state and universities: knowledge becomes more important and fundamental, universities are therefore much more than just institutions who validate knowledge. Because such knowledge must be publicly available, you can’t completely privatize them. Chaminade & Edquist - Why should the government intervene to support R&D and innovation? Two theoretical approaches to innovation: a. Neoclassical: focus on research. Scientific knowledge has three characteristics: uncertainty (about the outcomes), inappropriability (you never capture all the benefits) and indivisibility (minimum investment necessary before you can create knowledge). This leads to an underinvestment in R&D by private parties. Because of this market failure public authorities have to intervene. b. Innovation systems: criticize neoclassical perspective to be too blunt. They don’t give any guidance as to how large the intervention should be, or where the intervention should - - - - be. It also focuses on individual companies. The IS approach claims that firms do not innovate in isolation, but together. Therefore, focus on collective action, not individual. Also focus on tacit knowledge, instead of only on explicit knowledge. Rationales for public intervention in IS approach: control over IS is impossible, you can’t control/build an IS. The focus is instead on systemic problems, intervene in those areas where the system is not operating well and where problems are not automatically solved by private actors. Not useful to talk about market failure in an IS approach, since it is impossible to define the ultimate or ideal type system. Therefore, comparisons are impossible and ‘failure’ loses its meaning. Examples of such systemic problems: solving infrastructure problems (both physical and scientific), transition problems (from one technological paradigm to a radical different one), lockin (in existing technologies, hampers the development of new ones), institutional problems (inappropriate laws and regulation), capability problems (insufficient competences) and unbalance between exploration and exploitation (so stimulate spin-offs). Difficult to estimate what the reaction of the system towards policy changes will be. Therefore: experiment, allow room for mistakes. Evaluation of policies very important. Policy is especially needed when uncertainty and risk are high (leads to underinvestment by private parties) and it needs to be selective (can be down based on lobbyism or by analysis of the innovation system). The authors argue for the latter: focus should be on systemic problems through a rigor analysis of the innovation system, and not on lobbyism. Rothwell & Dodgson (1992) - - During 1950’s and 1960’s, two main tracks of RTD (research and technology development) policy: science policy (support for scientific education, university research) and industrial policy (grants for R&D, equipment grants, etc.). Little coordination between both. During the 1980’s, convergence between nations in terms of RTD policy: a. More attention for SME’s, focus on stimulation/growth of NTBF’s. b. Introduction of technology policy: based on the fact that demand-side factors were important and that more factors leaded to commercialization of new product than simply R&D. It involved e.g. linkages between firms and with universities c. Introduction of regional development policy: before the 1980’s, public policy was formulated and implemented by national governments. After that, regions got more autonomy with respect to RTD policy. Creation of regional technology transfer infrastructures are the best example of this: science parks, innovation centers. Mainly focused on creating NTBF’s and on SME’s. d. Technology programs and policies on the European level. Boekholt (2010) - Describes the evolution of public policy to enhance research, technology development and innovation (RTDI). 1950’s and 1960’s: focus on basic research at universities, or creation of basic research institutes (TNO). It reflected the linear model and supply driven focus, where one expected that innovations would fall out of the pipeline when you invested enough in basic research. - - - - After 1960’s: critics on linear model. More attention for the user, market pull and the value chain. However, innovation policy was (and still is to a large extent) based on the linear model (still focus on funding R&D projects). Some new instruments which were launched from the 1980’s onwards: a. Technology transfer measures: stimulate transfer between private and public entities. Universities were encouraged to set up offices to communicate with the outside world. b. Support university spin-offs: facilitate so by creating science parks. c. Setting up regional structures to help SMEs access/absorb technologies. Through organizations which gave advice on technological developments to SME’s and brings them in contact with research centers and other companies. d. Instruments to secure finance for innovation. Make venture capital more attractive. Growing need for policy learning, evaluation and proper tools for the analysis of applied policy. Systemic instruments by Smits & Kuhlmann give some insight in this. Largest obstacle in today’s innovation policy: governance of IS. Policy is too departmentalized, lacks overall vision. One has to cross bridges and boundaries. Policy must be coherent and coordinated across boundaries. Focus on portfolio of policy instruments. Suitability of a policy instrument depends very much on the context it is in, so don’t focus on single instruments. Again, the establishment of such a policy mix is hampered due to departmentalization of the government. Lack of cross-boundary communication. Smits & Kuhlmann (2004) - - - Three trends in innovation: end of the linear model (users are more important. For policy, this means increasing need to manage producer-user interaction), rise of the systems approach (organizations don’t innovate alone, performance firm is depended on the system. For policy, this means a need to embed policy in a broader context and a need to deal with system imperfections instead of market failure), inherent uncertainty and need for learning (in innovation. For policy, this means a need to reduce uncertainty by giving actors the relevant information and giving them instruments/facilities/environment to experiment and learn) a. Leads to triple helix of innovation process, intervention and theory. As a result of these trends, innovation is not an activity which is performed in isolation, but is a result of collective action. Therefore, there is a need for instruments that focus on the system level, not only on individual organizations. Specifically, present day innovation processes require instrument that support the following five functions: management of interfaces, build/organize IS, providing a platform for learning, provision of strategic intelligence (identify sources of innovation and build links between them), demand articulation (role of users). However, traditional instruments still dominate. Analysis of systemic instruments shows that they fulfill at least three or more of the systemic functions. They are the result of co-evolution of intervention, process and theory and seem to support each other.