Economics of innovation, Slide di Economia Dell'innovazione

Scarica Economics of innovation e più Slide in PDF di Economia Dell'innovazione solo su Docsity! 1 ECONOMICS OF INNOVATION Course Summary Innovation and firms • Innovation: basic concepts and definitions • Innovation in the history of economic thought. • Innovation, markets and firm performances. • Innovation appropriability and IPRs. • Open innovation Innovation, industries and governments • Financing of innovation and market failures • Innovation systems The geography of technological innovation • Innovation diffusion • Globalization of Innovation Course objectives Provide students with the basic tools of the economic analysis to study innovation under different points of view: • Determinants: sources(ànew ideas, R&D, learning), incentives. • Characteristics of Innovation and effects on society To answer some fundamental questions: • What is Innovation? • How the innovation process can be characterized? • Which factors affect/promote innovation? – Intellectualpropertyrights(IPRs) – Marketstructure – Resourcesavailability(competences,finance,etc.) • How important is Innovation? – Fortheenterprise – Forthesystem/society What is the Economics of Innovation? Economics of Innovation as a multi-disciplinary field. Microeconomics: Understanding the processes and the incentives of a firm to innovate. Macroeconomics: Innovation as a key driver of economic growth and as a determinant of the economic cycles. Economic Policy: Are there market failures in the innovation process and what, if anything, should the government do? Business Strategy: This is not a course for advising firms on how to innovate, but does include some insights on which innovation strategies are more suited in different technological and competitive “regimes”. Why study Economics of Innovation? • Innovation is one of the most important economic and business phenomena of our time: it has widespread implications for our economy and society. • Innovation phenomena were ignored for a long time in mainstream economics -> today, innovation has a more important role in modern economies. • Today, innovation economics is one of the most popular areas of research for young economists. • Economic and quantitative competences for analyzing innovation are increasingly requested by: Ø National governments as well as international organizations today are involved in innovation research (EU, OECD, UN, EPO). Ø Also many (large) enterprises have research units/teams focusing on doing innovation research. Ø Consultants and advisors specialized in innovation topics (e.g. IPRs strategies, analysis of market dynamics). How to study Economics of Innovation? • A simple application of the general principles of microeconomics is not sufficient to understand the complex nature of innovation processes. 2 • Besides economic reasoning, a good understanding of the technological determinants is necessary. There are huge differences between different industries and technologies. • Besides mainstream neoclassical economics, some appreciation of evolutionary economics is necessary to capture the dynamic and history dependent innovation phenomena. • Adjacent disciplines like economic sociology, psychology, engineering, management as well as modern complexity sciences play an important role: Innovation economics is interdisciplinary! The importance of innovation - 1 “... the capability to innovate and to bring innovation successfully to market will be a crucial determinant of the global competitiveness of nations over the coming decade...” “(...) intellectual assets taken as a whole — a concept seeking to aggregate measures of human capital, R&D and capacity to conduct it, patent valuations as well as intangible assets such as brand value or firm- specific knowledge — are rapidly becoming the key to value creation through a number of channels.” (Innovation and Growth, OECD* 2007) * The Organisation for Economic Co-operation and Development (OECD) is an intergovernmental economic organisation with 35 member countries, founded in 1960 to stimulate economic progress and world trade. The importance of innovation - 2 "Business has only two functions, innovation and marketing" (Peter F. Drucker, founder of “modern management") “Innovation distinguishes between a leader and a follower” (Steve Jobs, co-founder and CEO of Apple) “....coming up with an idea is the least important part of creating something great. It has to be the right idea and have good taste, but the execution and delivery are what’s key.” (Sergey Brin, co-founder of Google) “Creativity is thinking up new things. Innovation is doing new things.” (Theodore Levitt, economist and editor of the Harvard Business Review) Innovazione e crescita economica The Global Innovation Index (GII) is an annual ranking of countries by their capacity for, and success in, innovation. It is published by Cornell University, INSEAD, and the World Intellectual Property Organization, in partnership with other organisations and institutions. 5 → While invention can often come from a single ingenious individual, innovation is usually the outcome of a collective effort (e.g. innovation systems). • Diffusion: the spread of a new invention/innovation throughout society or at least throughout the relevant part of society. →Without this final step, innovator and society cannot gain full benefits. Some of this represents ‘spillovers’ or ‘positive externalities’. →Diffusion can occur through adoption and/or imitation (not always there is a clear distinction). 25 Definition of Invention: Usher Abbott Payson Usher (American economic historian, 1883 – 1965) defines inventions as the emergence of “new things” which require an “act of insight” going beyond the normal exercise of technical or professional skills. “Inventive acts of insight are “unlearned activities” that result in new organizations of prior knowledge and experience...”. Acts of skill include all “learned activities” that can be undertaken by technicians and engineers who possess the “state of the art” technology in their routine activity. “Such acts of insight frequently emerge in the course of performing acts of skill, though characteristically the act of insight is induced by the conscious perception of an unsatisfactory gap in knowledge or mode of action.” Science and Technology • Science: All human activities devoted to the discovery and dissemination of new knowledge through research (generic, undirected). The new knowledge is public good (non-rival in use), hence creates externalities. It may represent the basis for technological advances. • Technology: Application of knowledge to fulfill specific needs (usually to ‘production’ by firms driven by profit incentives). Can be a private good: investment (R&D) projects, use of intellectual property and secrecy to exclude others from the use -> it can be expressed by an isoquant or a family of isoquants. • Technique: point on the isoquant of the production function. • Technological Change (TC): change of the isoquant (of the production function) over time. Technological and Technical Change K New isoquant= New technology Technique used to produce output Y = combination of a pair of factors on a given isoquant Old isoquant= old technology Production function: Y=f(K,L); Y=output; K=capital; L=labour L Technological Change (TC) Given a production function Y = f(K,L), TC can be: • Embodied: refers to improvements in the design or quality of new capital goods or intermediate inputs (e.g. investments in better equipment)->Tangibility 6 • Disembodied: shift in the production function (frontier) over time without investing in new capital goods or intermediate inputs. (e.g. learning by doing)->Intangibility->Embodied in humans (Human Capital) NB: In general Innovation ≠ Technological Change (TC) TC is always an Innovation but not vice-versa: an Innovation can also be non technological (e.g. organizational or marketing innovation). Technological Change (TC) Technological Change can also be: i. Hicks neutral if the ratio K/L remains the same (e.g. K/2, L/2) ii. labour saving if L diminishes more than K (K/L increases) iii.capital saving if L diminishes less than K (K/L decreases) Radical vs Incremental innovations How “disruptive” is the new technology with respect to the market? • Drastic or Radical innovations: Drastic or radical innovation introduces a completely new type of product or production process with a wide range of applications and gives rise to a whole new genre of innovative products. (steam engines-> internal combustion engines, thermionic valves -> transistors, mainframe computer->personal computer). • Incremental innovations: (Minor) Improvements or refinements of existing products/processes with respect to existing dominant designs. Incremental innovations are usually introduced through imitation Christensen (1997) defines incremental innovation in terms of: ‘a change that builds on a firm’s expertise in component technology within an established architecture.’ NB: The latter type (incremental innovation) is more frequent, the former one (radical) is less frequent, but usually this distinction can be assessed only ex post. often associated with the introduction of a new technology. In some cases this will be a transforming technology, perhaps even one associated with the transforming effect Radical vs Incremental innovations of a Kondratiev long wave. Radical Innovations are usually followed by groups of several Incremental Innovations. They have a huge impact on society and they are at the basis of the theory of so-called “long waves” of economic development. (Kondratiev Table 2.4: Radical Innovations waves, see next Lectures). Examples of radical Innovations: Drastic vs non drastic innovation 7 Formal definition (based on Arrow, 1962, will be analyzed in Lecture n. 4) of radical (drastic) vs. incremental (non-drastic) process innovation in caso of monopoly (right) e and contestable markets (limit price theory – left). Demand and marginal revenue function: a refresh Revenues function: R=P(Q)*Q Marginal revenue: MR=R’(Q)=!"/!$=P(Q) + P’(Q)Q with P’(Q)<0 The marginal revenue MR (the increase in total revenue R) is the price the firm gets on the additional unit sold, less the revenue lost by reducing the price on all other units that were sold prior to the decreased in price. Hence: in this special case the MR function curve has the same intercept (a) of the inverse demand curve, with twice its negative slope (-2b). Modular vs Architectural innovation Classification proposed by Henderson and Clark (1990) Modular innovations: may result in the complete redesign of (all or some) core components, while leaving linkages between the components unchanged (e.g. Trevor Baylis’s mechanical radio – only power unit changed) 10 New product demand curve: temporary monopoly in the new created market A product innovation represented by a shift in existing demand curve: increased willingness to pay in the existing market. Technological vs.non technological innovations Non-Technological innovations: 11 Organizational Innovations: significant changes in the management and internal organization of the firm or in the external relationships with others (e.g. new alliances, M&A, etc.) Marketing Innovation: adoption of significantly improved marketing strategies OECD (2005) Oslo Manual Definitions Organizational Innovation: A new organisational method in business practices, workplace organisation or external relations. Examples: New business organization practices (Supply Chain Management, Business Re- engineering, Knowledge Management, etc...) Marketing Innovation: A new marketing method involving significant changes in product design or packaging, product placement, product promotion or pricing. Examples: social media marketing. Example: The electric mixer is a product or process innovation? • For the producer of cooking tools? • For the chef/restaurant? Leontief’s input–output flow matrix How to measure interdependencies across sectors or nations? How much integrated is an economy? Example: Two sectors economy: Machineries (M) and Steel (S). AMM and ASS = Shares of output used as intermediate inputs by firms in the same sector. AMS and ASM = Shares of output used as intermediate inputs by firms in the other sector. FM and FS = Shares of output consumed by the final demand F. XM and XS = Gross output produced in each sector. The Leontief’s input–output matrix can be used to measure also other flows than output (example: Investment flows) 12 Example OECD input/output tables http://www.oecd.org/trade/input-outputtables.htm Innovation in the History of Economic Though Innovation in the History of Economic Though: how it evolved. A. Smith (1723-1790) 15 Robert Solow: Was the first mainstream economist to demonstrate the huge importance of technological change: “Gross output per man hour doubled over the interval 1909 – 1949, with 87,5% of the increase attributable to technical change and the remaining 12,5% to the increased use of capital”. (Solow 1957) The evolutionary school J. A. Schumpeter (1883-1950) He is the most important writer on the economics of innovation. He was the first one to analyze in a comprehensive and systematic way the role of innovation in modern industrial economies. 1912 The Theory of Economic Development: An inquiry into profits, capital, credit, interest and the business cycle -> he coined the term: “creative destruction” 1942 Capitalism, Socialism and Democracy He emphasized the role of innovation (and the entrepreneur) as key competitive factor. “It is not [price] competition which counts but competition from the new commodity, the new technology, the new source of supply, the new type of organization ...” Christopher Freeman (1921 – 2010) He has founded the Sussex Policy Research Unit (SPRU) the first institute in Europe focusing on innovation. He was also a strong supporter of the idea that innovation economics has to be an interdisciplinary discipline. I.e. to understand innovation processes one has to use contributions form other disciplines than economics such as engineering, management, sociology, psychology, philosophy, and many others. He is also the inventor (with B.A. Lundvall e R. Nelson) of the important concept of National Innovation Systems which covers the synergies in the innovation processes undertaken by private actors together with all kinds of institutions (e.g. governments, universities etc.) in an economy. Other influential economists of the Evolutionary School Nathan Rosenberg: The innovation incentive can be either demand pull (driven by sophisticated consumers) and technology push (driven by research, discoveries and new technologies). Richard Nelson and Sidney Winter: authors of the very influential book “An Evolutionary Theory of Economic Change” (1982) in which they present several evolutionary innovation models with firms (characterized by heterogeneity, bounded rationality, routines) and market mechanism for mutation and selection. Eric von Hippel: has important contributions in the fields of informal know- exchange (innovation networks) and in the demand side phenomena of the innovation processes where he coined the notion of democratizing innovation. The evolutionary school: Paul David Paul David is one of the originators of the idea of path dependence in the economics of innovation: the equilibrium to which an economic process converges depends on historical path (accidents) that may lead to lock-in effects. This strongly contrasts with the idea of neoclassical economics where the “optimal” equilibrium is always attained independently of the initial conditions or “path followed” (ex. QWERTY keyboard). 16 The evolutionary school: Giovanni Dosi Giovanni Dosi: He introduced the concept of technological paradigm (TP)(taken from the concept of scientific paradigm by Khun). “TP is the general outlook on the problems (faced by firms) and the specific knowledge related to their solutions”. Paradigm shifts are rare events which become likely when the technological opportunities are depleted and increasingly bottlenecks of further development appear. Only in these periods major or radical innovation might become possible which then shape a new technological paradigm NB: Interesting reading (optional) on new technological paradigms on cloud computing. http://www.digitaltonto.com/2016/cloud-computing-just-entered- totally-new-territory/ The evolutionary school The term Evolutionary is generally used to refer to: - The neo-Schumpeterian economics: for its view of technical change and industrial dynamics as evolutionary processes. “[We shall designate by the term Economic Evolution] The changes in the economic process brought about by innovation, togheter with all their effects, and the response to them by the economic system” (Schumpeter, 1939, BC, Vol.I, p.86) - The Evolutionary school originated by Nelson and Winter (1982) for the analogies of their models with Darwin’s Evolutionary Theory (heterogeneity: firm’s technology and practices = genotypes, strategies = phenotypes, market competition = selection mechanism) - Other heterodox schools Aim: do develop a theory of Evolutionary Economics alternative to the (neoclassical/mainstream) one based on general equilibrium (and other assumptions). Evolutionary and neoclassical schools: main differences. 17 Next Lecture: The innovation process Lecture 2 (14-15/10/2020) The sources of innovation: Knowledge, R&D and Learning Characteristics of a public good: 20 - Recipes: To achieve a final outcome or service → necessity of a sequence of cognitive and physical acts. A recipe aims to specify the sequence of procedures that are legal (in the sense they are admissible and necessary to achieve the result) - Routines: Different people and groups are assigned to different part of the (persistent) process. • Distribution of various pieces of knowledge across many individuals. • Each individual receive a signal and answer with a different one • The repetition of this operations creates organizational routines - Artifacts-centered representation of technology Routines Routines, as stressed in Nelson and Winter (1982), • (i) embody a good part of the memory of the problem-solving repertoires of any one organization; • (ii) entail complementary mechanisms of governance for potentially conflicting interests. • and(iii)might well involve also some“meta- routines” (routines regulating other sub-routines) apt to assess and possibly modify “lower-level” organizational practices. Artifacts The procedure-centered representation of technology is highly complementary to what we could call an artifact- centered account of what technologies are and their dynamic over time • Recipes often involve designs of what it is there to be achieved as a final output. • Even when the procedure involves a notion of design, the latter is in general only one of the many possible configurations which can be achieved on the grounds of any one knowledge base. • In fact, when outputs are physical artifacts, it is useful to study their dynamics in the design space, defined by the properties of the components which make up the final output and their combinations. • Hence the history of technologies can be usefully tracked, from one angle, through the dynamics of outputs in their appropriate characteristics space. Example: retrieve knowledge through reverse engineering Market failures: remedies • Public provision of basic knowledge (E.g. through Universities and Public Laboratories -> general purposes and benefits for all society) • Subsidies, R&D Tax credits and Public Incentives (e.g. Prizes). Problems: – How to identify research projects that worth to be financed? – How to distinguish R&D from other expenditures of the enterprise? • Collaborations (Research Joint Ventures) – Internalization of Externalities • Intellectual Properties Rights (IPRs). Problems – Create temporary monopolies for the use of innovation. – Need to set a balanced system of IPRs (ex. Patent length and breadth) in order to balance Dynamic efficiency (innovation, but no competition), Static efficiency (competition, but no innovation) and social welfare (diffusion and de-duplication of R&D). Sources of Innovation: R&D 21 Research and Development (R&D): Research activity organized and formalized (e.g. routines) by businesses and other organizations finalized to introduce innovations. → it has become the main source of innovations during the XX century. The Frascati Manual (OECD, 2015) is the internationally recognised manual for collecting and using R&D statistics (NB: the OSLO Manual focuses instead on Innovation outputs). Official definition. “...Research and experimental development (R&D) comprise creative work undertaken on a systematic basis in order to increase the stock of knowledge, including knowledge of man, culture and society, and the use of this stock of knowledge to devise new applications...”. Sources of Innovation: R&D R&S can be divided in three phases: 1) Basic Research: finalized to increase/widen the scientific knowledge, without any specific direction or application in mind (undirected research); 2) Applied Research: uses scientific knowledge or combines existing technological knowledge to create new knowledge for a specific purpose (e.g: new product or processes). 3) Development: specific work systematically undertaken to finalize the production of new product or processes. NB: These three phases usually follow sequentially, but not always (differences across industries) R&D – (OECD Frascati Manual) Science vs. Technology • Basic Research vs. Applied Research↔ Science(S) vs Technology(T) – Content (type of knowledge produced): S = codified, general, undirected, a-finalized T = tacit, specific, directed, finalized Output: S = new models/formulas/principles and (testable) general laws. 22 T = projects, machines, products, specific knowledge OECD (2012) https://data.oecd.org/rd/gross-domestic-spending-on-r-d.htm 25 LbD is also an important informal strategy to increase the level of appropriability (on innovations) by giving the innovator a “first mover” advantage (over imitators) in terms of lower average costs or higher product quality. Important in industry in which experience matters and IPRs are not effective. Learning by doing: famous quotes Arrow (1962) “The Economic Implications of Learning by Doing”: “...technical change in general can be ascribed to experience, that it is the very activity of production which gives rise to problems for which favorable responses are selected over time”. “...Lundberg [9, pp. 129-133] has given the name “Horndal effect" to a very similar phenomenon. The Horndal iron works in Sweden had no new investment (and therefore presumably no significant change in its methods of production) for a period of 15 years, yet productivity (output per manhour) rose on the average close to 2% per annum. We find again steadily increasing performance which can only be imputed to learning from experience.” Learning by using (LbU) • Incremental improvements in of products/machineries coming from user experience. • It can be devided into two types: pure/non embodied and embodied LbU. Pure/Non embodied LbU can be defined either as the understanding of: • “the performance the components and materials within a complex system” • “the interaction of the complex system with the environment” that cannot be fully understood ex-ante by the producer (e.g. through scientific models or simulation/testing in labs) but only ex-post through knowledge coming from repeated use by costumers. Examples: maintenance and improvements of Turbo-JET engines, machine tools, software (automatic error reports), and other durable goods with system complexity or high fixes cost and long time to market. Learning by using (LbU) Embodied learning by using: – Improvements not only from the producer who use the machinery, but also from sophisticated consumer (lead users) with expertise that use the product – ‘...lead users of a novel or enhanced product, process or service are those displaying two characteristics with respect to it: lead users face needs that will be general in the marketplace – but face them months or years before the 26 bulk of that marketplace encounters them, and lead users are positioned to benefit significantly by obtaining a solution to those needs.’ (Von Hippel, 2005)“Democratising innovation” Lead users: costumers that are aware of their needs and have a strong incentive to fulfill them by exploiting their creativity and technical knowledge of the product itself (technical competence + incentive to innovate). Examples: Open Source software development, snowboard, mountain bike. Improvements can be developed and incorporated directly by lead users, or they can push producers to do so. Learning by interacting Very important in high-tech industries because of the complexity and variety of knowledge required to develop innovations • Scientific parks (eg: Area Science Park (Trieste), Kilometro Rosso (Bergamo), Como Next)). • Technological clusters (Eg. Silicon Valley) • Collaborative networks – Research joint ventures (RJV) – License agreements – Informal networks of knowledge exchange Learning from spillovers • Technological spillovers are unintended externalities coming from R&D performed by other firms or institutions usually located in the same region/cluster. • They represent positive externalities for the receiving firm. • Factors that affect the level of spillovers: – Efficacy of appropriability mechanisms (such as IPRs, secrecy, etc) – Tacitness of the knowledge involved (codified knowledge tends to “spill” out the firm more easily than tacit one) – Mobility of human capital – Absorptive capacity Horndal at Heathrow (ICC 2003) • Case study of organizational innovations introduced (also) through learnig • “Horndal at Heathrow” (Tether and Metcalfe, Industrial and Corporate Change, 2003) Increase in production capacity (Runway capacity- Air Traffic Management ((ATM)) in Heathrow, Gatwick e Frankfurt airports through: • Minimization of runway occupancy times (Rapid Exit Taxiways (RETs), changes in practices and better co- ordination between pilots and air traffic controllers. • Reduction of time-airbone separations between aircrafts during take-off/landing. 27 F o r m a l i z a t i o n s o f t h e I n n o v a t i o n P r o c e s s Main conceptualizations of the Innovation Process If innovation can be formalized as a process, how does it take place? • Linear model vs. chain linked model • Technology push vs. Demand pull Innovation as a process: the linear model • There are 4 main sequential phases that lead to the introduction of innovations. • Unique directions -> (no jumps or multiple arrows) • No backward feedbacks <- “Science The Endless Frontier” A Report to the President by Vannevar Bush, Director of the Office of Scientific Research and Development July 1945 The linear model: interim knowledge outputs 30 LIMITS OF THE LINEAR MODEL 1. Sometimes not all the phases are necessary [“Serendipity innovations” (no phase A)] 2. Sometimes not all the final stages (C,D) are reached [strategic patents] 3. Knowledge flows may be bi-directionalàe.g.. “feedbacks” from consumers to innovators [Open source communities]. Innovations by “accident” (serendipity) Post-itnote(3M): In1968,Dr.Spencer Silver, a scientist at 3M in the United States, was attempting to develop a super-strong adhesive. Instead he accidentally created a "low-tack," reusable, pressure-sensitive adhesive. For five years, Silver promoted his "solution without a problem" within 3M both informally and through seminars but failed to gain acceptance. In 1974 a colleague who had attended one of his seminars, Art Fry, came up with the idea of using the adhesive to anchor his bookmark in his hymn book. Instead he accidentally created a "low-tack," reusable, pressure-sensitive adhesive. Innovations by “accident” (serendipity) Aspartame(APM): Aspartame was discovered in 1965 by James M. Schlatter, a chemist working for G.D. Searle & Company. Schlatter had synthesized aspartame as an intermediate step in generating a tetrapeptide of the hormone gastrin, for use in assessing an anti-ulcer drug candidate. He discovered its sweet taste when he licked his finger, which had become contaminated with aspartame, to lift up a piece of paper of his notebook. He had the presence of mind to trace the sweetness back to a simple molecule of two amino acids. Innovations by “accident” (serendipity) Microwaveoven:Microwave ovens use microwaves, which are radio waves at a frequency of approximately 2,500 MHz, to heat food. 31 Percy Spencer was an engineer working for a company named Raytheon, developing microwave radar transmitters during World War II. One day, while Spencer was working on building magnetrons for radar sets, he was standing in front of an active radar set when he noticed the candy bar he had in his pocket melted. Spencer wasn’t the first to notice something like this with radars, but he was the first to investigate it. Critiques to the linear model • Serendipity innovations and recent studies have shown that innovation is far from being a linear and sequential process, but it can be generated from multiple sources following different development paths. • ENDOGENEITY OF SCIENCE with respect to technology and business. There are several reasons to believe that basic research can be affected by applied research and the economic context in general (reversal path direction <-) 1. Applied research may generate huge databases when studying current technologies -> these data can be used by science to validate its theories. 2. Relevant technical problems that cannot be solved incrementally by applied research can be solved by new fundamental research on the “core” of the problem (ex: CPU) 3. Technological change may alter the economic incentives in doing basic research. 4. Technology arrives first and sometimes contradict and falsifies “old scientific paradigms” (Ex: steam engine was invented before the discovery of thermodynamics laws, the first airplane came before most of the aerodynamics laws, radio and theories of electromagnetic vawes) 5. Improvements of technological instruments may open up opportunities for new scientific research and discoveries (ex. Microscopy -> medicine, telescope->astronomy) • IMPORTANCE OF FEEDBACKS – Commercialization and diffusion→ Innovation and development (ex: LbD and LbU) – Invention/Innovation→ Basic research The Chain-linked model (Kline and Rosenberg -1986) Innovation: high degree of uncertainty it is NOT the result of a linear model, but comes from a set of relations and interactions amongst the different phases and from continuous input-output feedbacks. It is an enriched version of the linear model: 1. once a target market is identified, firms first look for a solution internally. If no internal solution is found, they try to acquire external knowledge and/or do R&D. 2. innovation process centered around the concept of analytic design = analysis of original combinations of new knowledge (innovation) or existing knowledge (new design) -> innovation is not considered as a novelty in absolute terms, but also as an intelligent and original re-combination of existing knowledge. 3.The new analytic design is then developed→ tested→produced→commercialized 32 4.Importance of the “feedbacks” along the chain (arrows F) 5. Final innovations can affect the advancements of the scientific research: -> endogeneity of science. The Chain-linked model (Kline and Rosenberg -1986) The Chain-linked model (Kline and Rosenberg -1986) In addition to the “central chain”, there are 4 main “links”: 1) feedbacks from the bottom phases to the top; 2) new designs can follow directly from new scientific and technological advancements (link between Research and Phase 2) 3) Scientific and Technological knowledge are important not only for new design generation, but also throughout all the innovation chain (C links) 4) Innovations can drive/facilitate new scientific research/discoveries (double direction arrows instead of one- direction arrow of the linear model) Inside the black box: Rosenberg (1982): in many occasions technology affected science (ex. Galileo’s telescope opened new opportunities for the development of modern astronomy) Technology push vs. Demand pull Where the first stimulus to innovate comes from? • technology push ... scientific discovery invention production marketing • ...vs demand pull problems/suggestions from users invention production “It seems that all models of innovation exists in a wide range of forms and contexts. In general the pharmaceutical industries seem to lean towards technology push and science based models of innovation, whereas consumer products seem to be more demand oriented, while the innovation models in the assemble industries and sectors are more based on integrated and parallel oriented innovation models” (Rothwell 1994) 35 MEASUREMENT OF INNOVATION Introduction: Innovation and its implications can be measured and analysed – There are many problems, since the innovation process is lengthy and complex, which is subject to a large amount of uncertainty in its output – Without measurement & analysis, understanding and policy will be based on rhetoric, anecdote and lobbying • In a nutshell, economists try to act as natural scientists • However, in economics, experiments (field or lab) are difficult and hardly reproducible, hence the best viable alternative is to rely on observational data (like astronomy) to validate theories. • Therefore, economists (should) try to act a good statisticians: they (try to) provide robust empirical evidence characterizing observed phenomena and construct theories explaining and surviving to this empirical evidence Main indicators of Innovation used in economics (but there are many others): • R&D expenditures and R&D personnel (inputs) • Patent statistics (outputs) • Innovation Surveys (ex. Community Innovation Survey) • Technology balance of payments (TBP) • Productivity measures • Composite indicators R&D expenditures and personnel Measures of the innovative effort/input. • Advantages: good indicator for formalized innovation activity (input) and to proxy the stock of knowledge (absorptive capacity). Extensive and permanent data collection (almost) worldwide on yearly basis. • Disadvantages - The link with innovation outputs is uncertain and delayed (ex: time lag in pharmaceuticals) - Relevant for Science-based sectors, not for other sectors such as services (or supplier-dominated and scale- intensive Pavitt sectors). - R&D mostly performed by large firms. - R&D expenditures are not easily defined(see Frascati Manual),not always reported in the balance sheets. Tend to be affected by the economic cycle(pro-cyclicality) Evidences on R&D 36 Evidences on R&D: firms Patents as indicators A patent is a property right (granted by governments) to exclusively use a knowledge asset for economic purposes => useful measures of innovative output (more on this in next lectures) • # of patent applications at the firm, industry, country level over time • # of patents weighted by the number of subsequent citations that the patents receive • # of citations from one patent to another -> an imperfect but useful map of the links between these “bits” of output or knowledge (more on this in next lectures) Using patents as indicators requires to know and understand what “they are” how and why they are granted, 37 administered, enforced and how all this changes over time and across countries. Trademarks are an indicator of new product innovation and marketing activity, which tend to react to economic conditions, and not of inventive effort, which is more continuous -> represent minor innovations than patents, such as new varieties of existing products, as there is no novelty test. • Advantages - They represent inventions (output),some times before their commercialization (innovations). - Their novelty is assessed by examiners. - High frequency data( day by day) - Very detailed information on the technological content - Cross country coverage (national and international patent offices) • Disadvantages – Not all inventions later become innovations (no commercialization) – Their value is highly skewed – Different patent propensity across sectors and countries – Strategic use of patents (patent trolls) – Not all innovations can be patented – Cross country differences in legislation. EPO (European Patent Office) patent filings in 2016 (country of origin of the applicant) Analysis based on the Eu European filings under the EPC and international filings under the PCT). 40 https://www.oecd-ilibrary.org/science-and-technology/main-science-and-technology-indicators- volume-2015-issue- 2/technology-balance-of-payments-payments_msti-v2015-2-table67-en TBP Example: Japan Productivity and growth To measure real output value added is used • Value added is defined as sales minus raw materials used • Indicates what the firm has truly produced when transforming the raw materials into the final product • Both sales and raw materials have to be deflated for any price inflation when measuring over time Definitions of partial factor productivity: • labour productivity (value added per unit of labour) • capital productivity (value added per unit of capital) • High labour productivity is often largely explained by high levels of capital per worker (e.g. in mining and the steel industry) • High capital productivity will be present when labour is used intensively (e.g. in developing countries with scarce capital) Measuring total factor productivity (TFP) • This measure improves on partial factor productivity by correcting for growth in inputs • Derivation of total factor productivity: Suppose value added (Y) is produced by two input factors capital (K) and labour (L) and by total factor productivity (A) according to: 41 • A represents “technology.” The term “technology” in this context is a catchall that includes any impact on Y that is not accounted for by K or L. This could include process or product innovations that raise value added. • In productivity studies A is most often called total factor productivity (TFP) Measuring total factor productivity • Growth of TFP is calculated by residual: g A =g Y –αg K –βg L proof: Taking logarithms and assuming constant return to scale (α+β=1): l o g (Y ) = l o g ( A ) +  l o g ( K ) +  l o g ( L ) Differentiating both sides with respect to time: • g Y -> growth rate of value added or GDP • g A -> growth rate of technological change or TFP • g K -> growth rate of capital • g Y -> growth rate of labour Annual average growth in GDP per hour worked (1970-2006) Average growth of GDP per capita in emerging markets 42 Other economic growth resources • There is a vast amount of productivity and economic growth data on web that could be used to look at specific countries, periods or industries e.g. – National statistical agencies – World Bank, OECD (includes regular country studies), IMF – The Groningen Growth and Development Centre – Penn World Table Composite indicators Composite Indicators are computed as weighted average of single indicators Ex: Summary Innovation Index (European Commission) SII encompasses 27 different variables. Used in the «Innovation Union Scoreboard». 3 pillars”: Enablers (drivers of innovations), Firm activities, Outputs 8 “dimensions” covered Advantages: • Summarize different dimension of the innovation activities; • International comparability. Disadvantages: • Loss of information on the single dimension/indicator; • Availability only at the country/regional level. 45 EU15 1.92 0.76 39.6 54.6 Germany 2.49 0.76 30.5 67.1 France 2.16 0.94 43.5 50.8 UK 1.79 0.73 40.8 43.9 Japan 3.20 0.71 22.2 74.5 USA 2.66 1.06 39.8 61.4 “GOV” is the government sector and “BES” is the business enterprise sector. The conduct of R&D by business, government and universities in 2003 Country R&D/GDP conducted by BES R&D/GDP conducted by GOV R&D/GDP conducted by HES Sum of columns 1 to 3 Total R&D/GDP EU25 1.22 0.25 0.41 1.88 1.90 EU15 1.26 0.25 0.42 1.93 1.95 Germany 1.76 0.34 0.43 2.53 2.52 France 1.37 0.36 0.42 2.15 2.18 UK 1.24 0.18 0.40 1.82 1.88 Japan 2.40 0.30 0.44 3.14 3.20 USA 1.86 0.33 0.37 2.56 2.67 “BES” is the business enterprise sector; “GOV” is the government sector; “HES” is the higher education sector. Percentage allocation of government R&D support by objective in 2005 Country Land Health Energy Industry University Defence All Other EU25 9.6 7.3 2.8 10.9 32.0 13.6 24.0 EU15 9.3 7.3 2.7 10.9 32.4 13.8 23.8 Germany 8.9 4.4 2.9 12.4 40.3 5.8 26.0 France 6.5 6.1 4.5 6.2 24.8 22.3 29.5 UK 8.5 14.7 0.4 1.7 21.7 31.0 22.0 Japan 10.3 3.9 17.1 7.1 33.5 5.1 23.0 USA 4.5 22.8 1.1 0.4 .. 56.6 14.6 The Government-University Axis • Basic Knowledge tends to be a public good (non-rival), hence market mechanism alone cannot generate optimal amount (free-riding and underprovision) – Government funding of university research, and government research labs, are main solutions in modern economies. – The view that science, and its development into new technology, was fundamentally important to society slowly became clear after the experiences of World War II (e.g. Bell laboratories and the linear model). – The public provision of a public good, in this case scientific knowledge, with the public bearing the costs of knowledge production, which is then made available to all at a zero marginal cost, reflects its nonrival nature in use. Changing provision of basic science for knowledge economy Historical system: • Provision of basic science as a public good 46 • Discoveries were placed in the publi cdomain without any private ownership • Motivation of scientists to disseminate quickly their findings to the scientific community or ‘peer review’ • Use of science base open to al ltypes of business Recent changes: • Government finance for research is conditional on the research having more immediate application in industrial and commercial products. • Rise of academic patenting... The University-Business Axis • University-business links - many dimensions: – IPRs held by university: By universities patenting their scientific inventions and then licensing their use to industry. – Research joint ventures (RJV): By universities and firms engaging in jointly funded research and agreeing how to share the findings via contracts about future patents and licenses. – Spin-outs/start-ups: By the scientists from academic institutions actively forming new companies to develop and manufacture products that apply their scientific findings. – Personnel pooling: By exchanging and sharing science and engineering personnel between commercial firms and academic departments. • Growth of university IPRs • – US Bayh-Dole Act 1980 stimulated change. • – Before - government owned any patents coming from federally funded research and then issued non- exclusive licences. – After – university/scientists own IPRs and can licence exclusively to key firms. (Public knowledge becomes a private good!) – Often achieved via technology transfer offices (TTOs): staffed by skilled administrators and university technology managers, who assist with patent filing and licensing – Many EU countries have followed these changes. – Professor privilege: in most European countries, individual inventors, although they are employed by universities or public research institutes, are entitled to privately own the patents that emerge from their research in the service of the university (or public research institute) (OECD, 2003). University-Business Linkages Collaboration in Research • Joint, contract, and commissioned research, • Consultancy by academics Spin-outs, Start-ups, Science Parks • Formation of spin-outs and joint ventures • Formation of university incubators • Growth of science parks near to university 47 Personnel Linkages • Formal and informal social and professional networks • Continuing professional development and education, including public university lectures and workshops • Academic-scientist exchanges with firms • Recruitment of students from universities by firms The Government-Business Axis Key areas of innovation policy: • IPRs - the enforcement of IPRs can be influenced by national policy, as is legislation to some extent • Tax policy - corporate tax policy can affect innovation in various ways; key areas include R&D tax concessions, rules surrounding IP, and venture capital • Competition policy - the stance of competition policy matters, especially when decisions involve innovation (e.g. a firm has a dominant market position but also leads the industry in terms of innovation) Further key areas of innovation policy: • Government-business targeted funding – can be of specific research areas, technology development and small business • Standard setting - government is involved in setting various standards for measurement, performance, safety, testing and interoperability • Procurement policies - as a large purchaser of goods and services, the government can influence business activity (e.g. its decisions about purchasing computers) NaEonal InnovaEon Systems in Emerging Markets • SouthKoreaandTaiwan: - 50 years ago both were poor countries - Their governments promoted research and technology setting up important university research institutes - Firms were encouraged to do R&D - Initial approach was reverse engineering and technology transfer from the rich world (weak IPRs system) - Later graduated to developing world class innovations • ChinaandIndia: - Began with large populations but small % highly educated - China has encouraged FDI and technology transfer - India less open, but in 1990s expanded higher education 50 Regularities of the ILC (taken from various studies) On entry and exit 1. At the beginning of the industry, the number of entrants may rise over time or it may attain a peak at the start of the industry and then decline over time, but in both cases the number of entrants eventually becomes small. 2. The number of producers grows initially and then reaches a peak, after which it declines steadily despite continued growth in industry output. Stabilization 3. Eventually the rate of change of the market shares of the largest firms declines and the leadership of the industry stabilizes. Technology related 4. The diversity of competing versions of the product and the number of major product innovations tend to reach a peak during the growth in the number of producers and then fall over time. 5. Over time, producers devote increasing effort to process relative to product innovation. 6. During the period of growth in the number of producers, the most recent entrants account for a disproportionate share of product innovations. Utterback, James M., Abernathy, William J., A Dynamic Model of Process and Product Innovation, Omega 3, 1975, 639-656 Consider an new product (or technology or combination of technologies) introduced into the market. The product develops over time where in the beginning design and variety are most relevant and change whereas later product standardization and cost orientation are prevalent Three phases of product development: (1) performance maximization • high product variety and frequent changes herein; most (incremental) innovations are market stimulated (induced); high uncertainty with respect to market potential; relevant technology is multifold and derives from several sources (2) sales maximization • competition is based on product differentiation; some variants tend to dominate; firms consider more and more competitors when they design strategies and decide (3) cost minimization • product variety declines; standardization; product characteristics are clearly defined Utterback, James M., Abernathy, William J., A Dynamic Model of Process and Product Innovation, Omega 3, 1975, 639-656 Characteristics of the innovation process and the innovation strategies of firms vary with respect to technological and economic environment, product policy, growth and status quo of process technology Three phases of process development: 1) uncoordinated phase 51 • no prevalent standard; changes are easy to implement and are often accomplished manually; sub processes are loosely coupled; the production / technology system is ‘organic’ (2) modular phase • production systems are more mechanic and rigid; specific tasks are more specialized; formal control mechanisms are installed; automation and process control are introduced (3) systemic phase • the related production process is highly developed and sophisticated, capital intensive and shows low flexibility Dominant Design (DD): definitions Dominant design (DD): Dominant design is a technology management concept introduced by Ucerback and Abernathy in 1975, idendfying key technological features that become a de facto standard. Ucerback, James M., Suàrez, Fernando F., Dominant Designs and the Survival of Firms, Strategic Management Journal 16, 1995, 415-430 Definidon of dominant design • special path along the design hierarchy which dominates all other possible paths; this path and the consdtudng technological concept define a so-called standard or a dominant design (DD) (comp. to paradigm-trajectory approach, Dosi 1982) Genesis Dominant Design (DD) • determined by market as well as technology; considered as the result of a combination of technological, economic and organizational factors 52 • considered as the result of a competition among different technological alternatives (where not always the technologically superior / best technology will succeed) Influence of competition • early on experimental approaches with a lot of product / technology variants -> establishment of a DD with accepted, fixed product / technology features Industrial development • firms entering before the establishment of a DD show higher rates of survival • but: before the establishment of the DD firms run risk to select the ex-post wrong DD • later on DD serves as a barrier of entry since • • the DD is not easy to imitate by entering firms • • to enter with a product / technology not DD is highly unlikely to be successful Shake-out: definition Shake-out: the moment in the industry life cycle when abruptly the number of entry collapse and number of exit rises, drasMcally reducing the total number of companies. • Jovanovic, Boyan, MacDonald, Glenn M., The Life Cycle of a CompeMMve Industry, Journal of PoliMcal Economy 102, 1994, 322-347 • Genesis of industry: by a basic or fundamental innovaMon • Shakeout • shakeout as a result of a fundamental improvement • the respecMve invenMon may arise outside the industry under consideraMon • as the invenMon appears firms in the industry a^empt to adopt and uMlize it • those who successfully manage that change, the winners of this adopMon race, stay in the market and increase sales and market share whereas the losers exit the market/industry Industry Life Cycles 4 main Phases: (Dosi et al 1997: Klepper 1997) Phase 1: Introduction: • A new industry is started by the introduction of a (radical) innovation followed by a cluster of incremental innovations introduced by both the first innovator and the imitators. • Few firms: low competition on market shares but high competition on standard settings (multiple standard/design may co-exist). “Turbulent” and volatile market demand because of low product knowledge and utility awareness by consumes. • Low barriers to entry: many imitators or competitors with new standard/design may enter the new market -> Low production scale volumes (often inefficient). Phase 2: Growth: • Product standardization: a dominant design emerges. Product variety decreases. • Investments in plants and equipment to increase the volumes of production and reach an efficient production scale -> lowering of average costs. • Competition now more focused on price/quantity than on product innovations. • Barriers to entry become relevant (high efficiency scale, marketing and customers loyalty -> industry concentration increases. • Producers that didn’t adopt the emerged dominant design exit from the market (shake out) or get acquired. 55 Yearly Densities of Automobile Manufacturers: France and Germany, 1886 to 1981. Source: Hannan, Carroll et al. 1995. Cantner, Dreßler and Krüger (2006), German automobile industry: historical background • Start in 1886 with firms founded by Goclieb Daimler and by Karl Benz • Slow development because of railways and social structure • Increased growth at the end of 19th century; early 20th century large firms showed interest • Interrupdon during World War I; most firms produce military cars • Second big founder period early 1920s, followed by a shakeout in 1924 • Early 1930s emergence of the „Big Four“: Opel, Ford, Daimler- Benz and Auto-Union dominate the industry German automobiles 2040 #0 80 100 120 10 i É 3 H 2 \xu 2 teso iS00 isto iso ino sso teso sa00 tmio iso mo 10 rene ear e. al ca sl o Ésl Hea 2 mt2_ al i800 1900 1910 is0 i850 1940 year Phase 1 2 3 {Exploration phase) (development phase) (maturity phase) periode 1886-1924 1925-1930 1931-1945 # of years 39 6 15  57 Economic cycles and the theory of “long waves” The “long waves” (Kondratiev waves), in the Schumpeterian view are a direct consequence of tendency of innovations to concentrate in specific industries and time periods (clusters). According to this (still debated theory) the whole economic trend can be explained by the overlapping of three different economic cycles: • Kitchin (3-5 years) –> short/medium term cycle (linked to demand cycles, replacement of inventories and materials). • Juglar (7-11 years) -> fixed investment cycle (replacement of machineries, equipment and other fixed capital). • Kondratiev (45-60 years) -> introduction of radical innovations and technological changes in the economy. Long waves (Economist 1999) 60 Mark II: technological opportunities more linked to basic/fundamental research; Mark I: more linked to applied research) life cycle may be close to Mark I regimes, later phases to Mark II regimes. Schumpeterian hypothesis: from a dynamic perspective, (some) market power and large firms are fundamental for innovation and the technological progress, because of the nature of R&D investments (uncertain, costly, cumulative and long term returns). Over time, an industry may change technological regime. Early phases of the Malerba, F., Orsenigo, L., Technological regimes and sectoral pacerns of innovadve acdvides, Industrial and Corporate Change 6, 1997, 83- 118 Relationship between stages in ILC and the structure of innovative activities in the respective sector: • Two extreme cases: (1) entrepreneur, entering / creating the market with a new product in the early stages of the ILC development (entrepreneurial regime or Schumpeter I) (2) large firm, which in the phase of maturity produces a standard product on a large scale (routinized regime or Schumpeter II) Malerba, F., Orsenigo, L., Technological regimes and sectoral patterns of innovative activities, Industrial and Corporate Change 6, 1997, 83- 118 Sectoral structures of innovative activities Neo-Schumpeter hypotheses: size and market concentration purely economic Malerba/Orsenigo: • stability in innovator hierarchy • technology (market) entry technological learning Schumpeter I • small firms • low innovation concentration • low stability (dynamics) in rank orders • creative destruction 61 Entrepreneurial Regime Schumpeter II • large firms • high innovation concentration • high stability in rank orders • creative accumulation Routinized Regime Concentration of innovators Size and rank orders INNOVATIVE ENTRY Technological entry Share of patent applications by firms applying for the first time ina given technological class in the period 1986-1991 over the total number of patents in the same period (ENTRY)" High ENTRY Low ENTRY Clothing Organic chemicals Fumiture Macromelecular compounds Mining Electronic components Chemical processes Consumer electronics Machine tools Telecommunications Civil engineering Lighting systems Sports * this indicator measures innovative entry and not entrepreneurial birth: a new innovator may in fact have been around for quite a long time. (Malerba / Orsenigo 1997) Schumpeter I and II idealized Schumpeter | Schumpeter Il low concentration high concentration low asymmetry high asymmetry low firm size high firm size {) Vow stability in rank orders high stability in rank orders high entry rate low entry rate (-) negative correlation 62  65 Features of the evolutionary approach Individuals and organizaWons with procedural raWonality (H. Simon) adopt “simple” behavioral rules in order to: - Take decisions in a context with incomplete informaWon - Exploit only the (few) informaWon available (observe the history of own past success/failures or the acWons and outcomes by others and argue/infer on their level of “saWsfacWon”). Example of “sa#sfying” strategy of technology adopWon by firms with procedural raWonality: The evolutionary school: Nelson e Winter (N&W) 1982 1982 Nelson, R.R e Winter, S.J. An evolutionary Theory of Economic Change Differently from the original Schumpeterian school: 1. More rigorous formalization of the cognitive bounds of the agents (firms) through the concept of “organizational routines” 2. Less emphasis on radical technological discontinuities: incremental innovations and imitations are also important. 3. More emphasis on firm’s heterogeneity and market selection mechanism for explaining inter and intra sectoral differences. 4. More extensive use of the analogies with evolutionary biology: “...in analogy with evolutionary biology, one is able to identify four principal building blocks of an evolutionary theory: (i) a fundamental unit of selection (the genes); (ii) a mechanism linking the genotypic level with the entities (the phenotypes) which actually undergo environmental selection; (iii) some processes of interaction, yielding the selection dynamics; (iv) some mechanisms generating variations in the population of genotypes and, through that, among phenotypes” [Dosi e Nelson, 1994, JEE, p.155] The main distinctive feature of the model of Nelson e Winter are firm’s routines (organizational memory or “inertia”). 66 A Routine is “an executable capability for repeated performance in some context that has been learned by an organization” (Cohen et al., 1996). Routines are made of decision rules and “automatic” procedural behaviors (i.e. what the firm can do) that affect the life of an organization. • They are linked to the context/environment → variety • They embody a good part of the memory of the problem-solving repertoires of any one organization • A routine is not the raw sum of individual team member’s capabilities, but it’s the result of complementarities and interactions amongst them Technology Evolution as an “Evolutionary Process” i) Unit of selection: Firms/agents (phenotypes) are characterized by organizational routines (genotypes) that drive their decisions (e.g. investments, production, innovation, etc.). ii) Process of mutation: the satisfying procedural process for searching of new routines (Herbert Simon), “firms are driven to consider alternatives ... under the pressure of adversity” (N&W, 1982, p. 211) iii) Process of selection: market forces (and environment in general) drive the selection processes for firm’s survival and growth. These 3 features form the basis of the main principles of evolutionary biology: i. Variety: heterogenous firms as the result of the joint processes of selection, mutation and transmission (heritage) ii. Heritage: best routines are preserved by the firms and adopted (imitated) by others (possibly with small modification) iii. Selection: the competitive “struggle” for firm’s survival (innovate or die, survival of the fittest) Main features and stylized facts of the Nelson e Winter (1982) model Synthesis of the main dynamics of the model: • An initial population of N heterogeneous firms, searching for more efficient routines Rit, that are selected if they are more efficient in terms of average cost of production c: c(Rit)<c(Rit-1). • The search S for new routines, either through imitation (Im) or innovation (In) is uncertain (0<p<1) and its costs depend on the difficulty of the searching process: S(In)>S(Im) • Probability of a successful search ps depends on the amount spent in R&D->p(S). • The investment in research S depends on the strategy (innovation or imitation) and the amount of financial resources available rΠ (r=share of profits invested in research) • Large firms can spend more in R&D, have higher probability to find better routines and can apply cost reduction to larger volumes of production-> higher survival probability. • By setting “plausible” initial values of the parameters and by running several simulations for long time periods, the final market structure tends to be concentrated with few large innovating firms. • The Schumpeterian hypothesis is confirmed: in a dynamic perspective, (some) market power positively boosts innovation-> Dynamic Efficiency (innovation, but no competition) vs Static Efficiency (perfect competition, but no innovation). Main difference with evolutionary biology: whilst in biology the generation of new varieties (speciation) occurs through a random combination of genes (alleles) and the natural selection determines the “survival” of new species, in evolutionary economics the way in which new varieties are generated from old ones doesn’t follow a pure random process, but is driven by human intervention in the determination of the direction of the speciation. 67 The human willingness to solve some technical problem through R&D (it is the human deciding in what project to invest) is constrained by: • operating experience (e.g. learning by doing), rather than through formal training in the sciences; • efforts at inventing and solving technological problem are constrained by the range of options that are perfectly understood; • firm heterogeneity in knowledge and practices (even in presence of identical relative prices or cost of inputs). Such differences hardly come from either science or engineering principles, but rather form idiosyncratic experience. Having said that: Are there invariances in the knowledge structure and in the ways technological knowledge accumulates? Technological paradigms Each technology needs to be understood as comprising: • (a) a specific body of practice—in the form of processes for achieving particular ends—together of course with an ensemble of required artifacts and components on the “input side”; • (b) quite often some distinct notion of a design of a desired “output” artifact “in mind”; • (c) a specific body of understanding, some relatively private, but much of it shared among professionals in a field. • These elements, together, can be usefully considered as constituent parts of a technological paradigm Technological paradigms A paradigm entails: - an outlook: a definition of the relevant problems to be addressed and the patterns of enquiry in order to address them; - a view of the purported needs of the users and the attributes of the products or services they value; – patterns of solution to selected techno-economic problems — that is, specific families of recipes and routines— based on highly selected principles derived from natural sciences, – jointly with specific rules aimed at acquiring related new knowledge (heuristic of search) – design concepts which characterize the configuration of the particular artifacts or processes that are operative at any time Sometimes the establishment of a dominant paradigm is not associated with a dominant design. Example: pharmaceutical technologies and biotech which do involve specific knowledge basis, specific search heuristics, without any dominant design or artifact characteristics. Technological Trajectories Technological paradigms “channel” the evolution of technologies along distinct technological trajectories. As paradigms embody the identification of the needs and technical requirements of the users, trajectories may be understood in terms of the progressive refinement and improvement in the supply responses to such potential demand requirements. 70 How demand side affects technological change Changes in relative prices can easily induce changes in the directions of the technological changes brought to practice by users/adopters of new technologies, even holding search behavior constant. New production techniques or new machines to be sold to a users will be selected only if they will yield total costs lower than those associated with the incumbent techniques/machines. But the outcome of the comparison obviously depends on relative (input) prices. Change in paradigms of mobile phones: from miniaturization to enhancement of battery storage performance S ALecture 4 Innovative Firms and Market Structures (Neo-Schumpeterian Hypothesis) Innovation and market structure • What kind of market structure provides more incentives to innovate? – Incentives are represented by the increase in profits that the innovative firm can enjoy after the innovation. – but profits depend (also) on the degree of market competition and market structure (number of firms, relative dimension, degree of differentiation…). – Should we expect a higher innovation intensity in more or less concentrated markets? Is there any association between firm size and propensity to innovate? • Is also innovation affecting the market structure? – How the lead-time advantages (in term of lower cost, or higher product quality) of the innovator can affect its relative size (with respect to competitors) and thus modify the market structure? 71 The Schumpeterian hypothesis (1942) • Innovation is the core of dynamic competition “It is not [price] competition which counts but competition from the new commodity, the new technology, the new source of supply, the new type of organization … competition which commands a decisive cost or quality advantage.” Schumpeterian Hypothesis (2 parts): R&D mainly performed by: – (1st part) Concentrated industries ( some market power is needed for appropriability) – (2nd part) Large firms with large R&D labs → Perfect competition is not only impossible to “attain”, but it is also sub-optimal (in a dynamic perspective) since it doesn’t provide enough incentives to innovate. • Creative accumulation (Mark II) vs Creative destruction (Mark I) Innovation and Market Structure • Neo-Schumpeter-Hypothesis I: firms with larger market shares should innovate more • Large market share gives more certainty about recouping returns to R&D once innovation occurs • It also implies more current profits to finance the expenditure on R&D • This hypothesis has led to substantial theoretical and empirical work on the relationship between market structure, competition and innovation • Possible there is an inverted U-shaped relationship (see next slide), but economists cannot yet identify the optimal degree of competition C* Inverted U-shape between innovation and competition Innovation and Market Structure • Neo-Schumpeter-Hypothesis II:Large firms produce more technological innovations than small firms • Innovation propensity tends to increase with firm’s size because: 1. R&D projects usually involve high sunk costs that require large scale economies and volumes of production in order to be recovered. 2. Large and diversified firms can more easily from unintended/unexpected results of the innovation process (economies of scope) 72 3. Large firms can diversify the (technological and strategic) risks of R&D failures by performing multiple R&D projects. Empirical evidence on this second hypothesis is mixed: • Large firms are more likely to do R&D or be IP active • But smaller firms that are R&D or IP active have higher intensities of such activity Evidence on returns to innovation Evidence of private rates of return to R&D: • Investigated using either market value or productivity approaches • Both approaches suggest private rates of return to R&D are higher than for standard, tangible investment projects • Excess returns may be reward for higher risk • High rates of return also suggest that there is not free entry into R&D • Could be due to barriers, e.g. raising finance, lack of skilled labour, or IPRs • Also possible R&D requires complementary assets e.g. tacit knowledge and skilled labour. Evidence on interaction between competition and innovation • Absolute firm size is not necessarily beneficial to innovation • Larger market share has been found to increase the returns to R&D • But those with very high degree of market dominance may become complacent • Recent evidence relating rates of patenting to degree of product market competition supports the inverted U-shape R&D ranking of the top EU companies (2011 EU Industrial R&D Investment Scoreboard) Arrow (1962): Drastic innovation Drastic innovation G"-Gl=qLe, G=0 Arrow (1962): Non Drastic innovation (p1">co): Incentive to innovate in perfect competition: * Equilibrium before innovation: (po°=co, g°). Profits before innovation: Gp°=0. Price=marginal cost-> no extra-profits. * Equilibrium after innovation: (p1°=c1+L=co, 91°=0°). Profits after innovation: G,°L=co- c->license fee per unit that maximize the innovator/s profits. * Additional per period profit (incentive): AG°=(c-c1)g°=(p1°-c1)q°. Incentive to innovate in monopoly: * Equilibrium before innovation: (po, q0”). Profits before innovation: Go"=(po"-Co) qo">(p1"-cp) q1" since Gy" is the maximum profit that the monopolist can get with the old technology Co * Equilibrium after innovation: (p1", q1"). Profits after innovation: G,"=(p,"-c,)q1". * Additional per period profit (incentive): AG"=G"-Go"=(p1"-c1)q1"-(Po"-co) gdo” which is lower that the following quantity (Pm Ap, MA * Since py">co-> qiM<q"-> AGM=G1"-Go"<(c9-c1)q1"<(co-c1)a°=AG". Even with non-drastic innovation the incentive to innovate is lower in monopoly than in perfect competition. 75  76 Arrow (1962): Main results Result According to Arrow (1962) the incentive to innovate in competition is always higher than in monopoly. The reason for this result is as follows: • The monopolist already gains a profit before innovation which is cannibalized by the new technology (replacement effect) • The profit incentive of the inventor on the competitive market leads to a monopoly situation (with positive extra-profits) which hasn’t existed before (no previous extra profits are “cannibalized”). Arrow (1962): conclusion Conclusions: Is the model by Arrow (1962) contradicting the Schumpeterian hypothesis: yes, but... The the model by Arrow (1962) has several limitations: • Static character of the model: the Schumpeterian hypothesis assumes dynamic competition (e.g.: new products/processes vs. existing ones), whereas the model by Arrow (1962) is a static one. • Asymmetry when comparing competition and monopoly: • Different output levels in t0 before innovation (qc >q m). • Consequently the cost reduction in monopoly concerns a lower volume of output; hence, the profits tend to be lower already for that reason and therefore the incentive to innovate. • When considering equal initial conditions in t0, either in term of initial volumes of production or quantity elasticity with respect to price (εQ,p) (e.g. by assuming different demand functions), Demsetz (1969) and Kamien e Schwartz (1970) found an higher incentive to innovate in case of monopoly. Arrow (1962): other critiques Criticisms: 77 The model by Arrow (1962) has been criticized also because: • It assumes perfect appropriability ex-post of the innovation (i.e. no spillovers). • Innovation is “exogenous” and do not depend on the innovator’s effort -> only fixed costs of development are assumed • Only analysis of extreme cases (monopoly versus perfect competition) and no analysis of intermediate cases (e.g.: oligopolies) -> Interdependencies of innovation and market structure are not entirely taken into account. • Assumes no uncertainty on the economic returns of innovation. In reality it is a very uncertain process due to both technological uncertainty (R&D projects may fail, see e.g. Nelson (1959)) and strategic uncertainty: even if R&D projects succeed, competitors may still innovate first (patent races and “winner takes all”) or imitate quickly. Innovation and size of the firms: Nelson (1959) • Scientific research may be defined as the human activity directed toward the advancement of knowledge, where knowledge can be either represented by new facts or data observed in reproducible experiments and theories or relationships between facts and data (usually, but not always, equations). • These advancements are not fully appropriable by private firms. Scientific knowledge is a public good and very often has practical value in many fields. • But… Few firms operate in so wide a field of economic activity that they are able themselves to benefit directly from all the new technological possibilities opened by the results of a successful basic research effort. • A firm with a narrow technological base is likely to find research profitable only at the applied end of the spectrum, where research can be directed toward solution of problems facing the firm, and where the research results can be quickly and easily translated into patentable products and processes. Such a firm is likely to be able to capture only a small share of the social benefits created by a basic research program it sponsors. • On the other hand, a firm producing a wide range of products resting on a broad technological base may well find it profitable to support research toward the basic science end of the spectrum. Nelson (1959) • Nelson (JPE, 1959) (Basic) scientific research (characteristics) • Technological uncertainty • The marginal benefits (MB) of scientific research depend on how many innovations the firm will be able to commercialize from its discoveries → how many? For how long? In which sectors? Simplifying assumptions: • Marginal cost (MC=dC/dR) of basic research is constant and equal to 1 • Marginal benefits (MB=dΠ/dR) are decreasing (=dΠ2/dR2<0) (where Π are the firm’s profits) Who has the higher incentives to innovate? • Small specialized firm? • Large diversified firm? • Government? LARGE VS SMALL ENTERPRISE • Advantages of large diversified firms: - higher chances to appropriate the returns of research from a larger product/innovation portfolio and for longer time. Model: Firms: n, i=1,...,n Homogeneous output of firmi: g,, Î=1,...,1 c dp d Aggregate output: Q= Va, Demand: p(0), P <0, L >0 il dQ d0 de, dc, . Technical progress (unit cost of production): c;(), —<0,+->0, Vi dr, dh Factor price of R&D: 1 R&Dexpenditures: C,(7;)=7;, i=1,....n î Profit Function: 77,(g;.1;)= 11,(1,9;)-C;0;) = P(4,):q; =): 7% Gross return from innovation: 11,(;,9;) = p;(4;)*d;76;0;)q; Where SH >() and dH <0 dr dr Dasgupta & Stiglitz (1980): isolated decision (monopoly) In the optimum the ci margina! retum of an = innovation corresponds to VI(#) 2 4 tic margini sis I RED. 7(r) | Î) | “n AS x(n) Y I 0 s Figure 1: The decision of an isolated innovator 80 Dasgupta & Stiglitz (1980): decision in competition Non tournament model - The de. ion in competition In competition, firms are considering the demand function, which is not only determined by their production g; but by the overall production of all firms in the market Q, where Q= Va * Profit maximization: MaxT,(g;,1,)= p(0):q;-c;(G;):g;-1, ViEn di Atomistic competition: production decision by firmi doesn't affect the other's. * First order conditions (1): da, pd dp | dq; da IO, (O) | SI] (0-0 dg, 900 dq; d0| dq; rr: positive . re (absolute negative value)! - =c;(r) (1 * Innovator will choose a quantity where the price is larger than the unit costs (here equals marginal costs of production) in order to sustain the cost of R&D (if price elasticity of quantity £ap iS elastic (£0,p>1) or at least £ap>q/Q.) Dasgupta & Stiglitz (1980): decision in competition Non tournament model - The decision in competition + First order conditions (2): de , , der da, __de, 09; where£ —=|L ia -Zij-1=0 + ge CESSI (@ en ar al n a) ore| | dr r + Innovator will choose an R&D level where the unit costs reduction induced by the last R&D unit multiplied by the quantity is equal to the marginal costs of the last R&D unit used (equals factor price) *. Byplugging (1) in (2) we get: r , ing, | P(Q)q; (CoA + R&D intensity of the innovator: willingness of the innovator to Invest in R&D out of its sales this willingness is the higher - the higher technological opportunities (£;..)->elasticity of c with respect to r - the higher demand elasticity (en) - the lower the relative firm size (q/0) s di with <p 81 Dasgupta & Stiglitz (1980): decision in competition What about the R&D intensity of the whole industry? Market structure: Symmetry assumption. * q;andr,identical forall î: q=£, r=l, C=C; n We can re-write the first order conditions (1) and (2) as follows: U ; 1 from (1: —g; + p(0)-c;(1,)= p(Q)|1--2— |-c;()= sonore (1-4 0 E0» 11 -nofi-1)-c-0 G) NE 9001 . cmq 1 E, =1 4 dr n r From (2): Dasgupta & Stiglitz (1980): decision in competition Market structure: Free market entry condition * with free entry equilibrium, profits will be 0 Va XP -cM)q-r]= 1 p0)-c@)g=r]=0+ iel il =>(p(0)-c(r)gn=nr> sc nr =c(r) > p(0)-—=c(r) (5) Q R&D intensity of the sector by plugging (5) into (3) we get: (0) i 1 1-2 |a po) si 2 Q P(0) nr 11 nr Ù nr (0)- po) P22 La lie, POTPO ron O Mio, DOO n° N00 (6) 82 Technological spillovers, decision in competition Levine and Reiss (1984): - First order conditions: diodi q +0) (1, 0R)=0= (0) It - (0) (1 =(£. 24 Be peg LOI, _ r * Where £,,= Technological spillovers, decision in competition Levine and Reiss (1984): Ui * Plugging(1)into (2) and re-arrangingtermsi —__= (E. pt de, 7 P(O)q; Similar result as in Dasgupta and Stiglitz (1980), but in this case, the relationship between R&D intensity, demand elasticity and relative firm's size is mediated by: * Ex-(elasticity of firm’s unit cost with respect to its own R&D): firm’s R&D intensity increases with the degree of technological opportunities; * Ecalelasticity of firm’s unit cost with respect to market's R&D): the impact of firm’s R&D intensity with respect to the technological dimension of appropriability (8) depends on the sign and magnitude of E, in particular: * if Eg,>1, complementary know-how: competitors increase their joint efforts when firmi increases its own R&D effort. * iFO<,,<1, substit ive know-how: competitors increase their R&D less intensively than firm î. * IfEg,<0: “free-riding”: competitors decrease their joint efforts when firm i increases its own R&D effort. 85 R&D intensity of the industry: symmetry and zero profit condition: * Applying the symmetry condition and multiplying both sides of the profit function by n: 1,(q,,7,,R)=0> |rl0) -c(1,0R)|a, =,> I) - c(r,0R)|na = nr * Dividing both sides by p(Q) and Q=ng: P(Q)-c;(1,0R) iS e (3) PO) P(O)Q The left hand side is the Lerner index (of monopoly power) and the left hand side is the industry R&D intensity with respect to sales. * Byplugging (1) into (3) we have (as in Dasgupta and Stiglitz (1980): 1-|1-£ 1 4 1 MM 1 =%, nr QE0n) P(WDQ Qeo, POQ n p(0)Q Industry concentration (1/n) is proportionate to industry R&D intensity, where elasticity of demand is the factor of proportionality. Exercise 1: Optimal R&D investment (I) The gross profit from innovation ÎI which depends on the innovator's investment in R&D r, is given by: Il=ar, a>0 R&D costs C are given by: C(r)=br?, b>0 a. Calculate the maximum net profit from innovation n given the R&D investment r. (Provide a graphical illustration of your solution) b. Assume that, due to imitation the innovator can only appropriate the fraction y (0 <p < 1) of the innovation success. How does this change the innovator’s decision to invest in R&D? c. Forwhich y is the incentive to innovate completely vanished? 86 Exercise 2: Optimal R&D investment (II) The gross profit from innovation TT which depends on the innovator's investment in R&D r, is given by: Î1=4r°5 R&D costs C are given by: C(r)=0.5r a. Calculate the maximum net profit from innovation rr given the R&D investment r. Provide a graphical illustration of your solution! Due to the presence of imitators, the innovator can only appropriate a certain fraction y of the initial innovation success, given by u = 0.8. b. Howdoesthis change the optimality condition of the model? (Provide a graphical illustration of this change) A simple patent race model with entry * Profits for M: Before the race: IIm After the race: — (IIw- cost of the license) if M innovates; — II if E gets the license and innovates - pergamena amount that M is willing to pay for the license: IIy- D * Profits for E: Before the race: 0 After the race: — 0 if M gets the license; — IIp - cost of the license if E gets the license and enters. - DENTE amount that E is willing to pay for the license: IIp-0 7 D 87  90 innovation is radical, then potential entrants may have an incentive to invest in innovation larger than the one of the incumbents (replacement effect>efficiency effect) • The previous model can be extended to take into account also of the technological uncertainty of R&D, e.g.: – Low risk strategy→ Incremental innovation with high probability of success – High risk strategy→ Radical innovation with low probability of success – The previous results suggest that incumbents have higher incentives to introduce “small” incremental innovations (less “risky”, less “disruptive” and competence enhancing), whereas new entrants have higher incentives to introduce “higher” radical innovations (more “risky”, “disruptive” and competence destroying). Famous examples: Xerox One of the great inventions of the 1960s was Rank Xerox’s technology of electrostatic copying (“xerography”). This technology allowed for copying onto plain paper at a substantially lower cost than photography-based methods. With a view toward protecting its near-monopoly, Xerox patented not only the process of xerography but also every imaginable feature of its copier technology. As claimed in the suit later filed against it by the SCM Corporation, Xerox maintained a “patent thicket” wherein some innovations were neither used nor licensed to others. It would appear that the only purpose of these “sleeping patents” was to prevent competitors from inventing a technology similar to Xerox’s. The result was that, when IBM and Litton entered the market in 1972, Xerox sued them under literally hundreds of patents. More than 25% of IBM’s budget at the time was devoted to patent counsel, not R&D. As a result of mounting complaints against Xerox’s exclusionary strategy, the Federal Trade Commission eventually ordered Xerox to license its patents to all entrants at nominal cost. Within a few years, prices of plain-paper copiers were cut in half. Xerox’s market share dropped from 100% in 1972 to less than 50% in 1977 Source: Cabral 2015 Famous examples: Nintendo vs Sega Video games are a big business. Nintendo’s main product was an 8-bit machine and a series of games featuring the popular Mario. Since the late 1980s, Nintendo had been developing a faster, 16-bit machine. Nintendo, however, was not in a hurry to launch the new product: In fact, by the late 1980s, Nintendo’s 8-bit machine was at the peak of its sales. Launching the 16-bit machine might significantly cannibalize the market for the slower system. Sega did not have to worry about such trade-offs. In October 1988, it introduced its 16-bit Mega Drive home video-game system. Eventually in September 1991, Nintendo introduced its own 16-bit machine. A fierce price war ensued, with Nintendo and Sega sharing the market in approximately equal shares. In the transition from the 8-bit system to the 16-bit system, Nintendo lost its position of near-monopoly, having to share the market with Sega. Source: Cabral 2015 Innovation and market structure: empirical evidence 91 Model (very simplified version) by Aghion et al (QJE, 2005) – Hypotheses. Hypotheses: Suppose we have 2 types of industries, in each of them we have only 2 firms. • Industry type NN (Neck-and-Neck, or leveled): the two firms share the same technologies and thus compete “neck to neck”. • Industry type LL (Leader and Laggard, or unleveled): one firm has a superior technology (the leader) and the other has an inferior one (the laggard-follower). Suppose that in the industry type LL, the follower has higher incentives to innovate (and thus is the only one to innovate) because of the presence of technological spillovers form the leader to the follower, (low appropriability regimes, the follower can easily imitate and the leader has no incentives to innovate). The net incentives to innovate are measured by the difference between ex-ante and ex-post innovation profits: ∏POST – ∏PRE Analysis: • In industries NN if the degree of competition is LOW (e.g. near collusion), since existing profits are high (∏PRE high) the incentives to innovate (∏POST – ∏PRE) are low (keeping ∏POST fixed). • An increase in the degree of competition would decrease ∏PRE and thus increase the incentives to innovate (∏POST – ∏PRE): if competition ↑-> innovation↑ (this effect is called “escape competition effect”). • In industries LL if the degree of competition is LOW, since the ex-post potential profits of the follower are high if it innovates (∏POST high) the incentives to innovate (∏POST – ∏PRE) are high. • An increase in the degree of competition in this case would decrease the value of ∏POST and thus decrease the follower’s incentive to innovate (∏POST – ∏PRE): if competition ↑-> innovation↓ (this effect is called “Schumpeterian effect”). Implications (I): If the degree of competition is LOW: • In industries NN we have low incentives to innovate (high replacement effect because high ∏PRE ) and thus less innovation; • In industries LL we have high incentives to innovate (because ∏POST high) and thus more innovation: these industries, since only the follower innovates, will tend to switch to a NN type. 92 IMPLICATION 1: If the degree of competition is LOW, most of the industries tend to be NN type, and in these industries there will be less innovation (because ∏PRE high) . IMPLICATION 2: in this situation, if competition INCREASES (since most of the industries are NN), the incentives to innovate (and thus innovation) INCREASE (∏PRE decreases) because of the “escape competition effect”. Implications (II): If the degree of competition is HIGH: • In industries NN we have high incentives to innovate (because of low ∏PRE ) and thus more innovation (these industries will tend to switch to a LL type); • In industries LL we have low incentives to innovate (because ∏POST low) and thus less innovation: these industries will tend to remain in a LL status. IMPLICATION 3: If the degree of competition is HIGH, , most of the industries tend to be LL type in which, because of the high competition, there will be less innovation (because of ∏POST low) . IMPLICATION 4: in this situation, if competition INCREASES, (since most of the industries are LL), the incentives to innovate (and thus innovation) will DECREASE (because of the “Schumpeterian effect”). Aghion et al (QJE, 2005) – Graphical representation of the implications 95 M will always have the incentive to block the entry of E. Thus M will produce as much as varieties until E makes no profit if it enters: – Profits of M with 1 variety + entry: (1/2)p – F – Profits of M with 2 varieties + entry: (3/4)p – 2F – Profits of M with 2 varieties + no entry: p – 2F p – 2F > (3/4)p – 2F > (1/2)p – F Brand proliferation through product varieties is driven by the threat of entry. With no potential entrants, the best strategy that maximizes the profits of M would be to produce only one variety (profits= p - F) Innovation and market structure: suggested readings 1. Neo-Schumpeter Hypotheses •Hagedoorn J. (1996), Innovation and Entreprenuership: Schumpeter Revisted, Industrial and Corporate Change 5(3), 1996, 883-896 •Mueller D.C.,J.E. Tilton (1969), Research and Development Costs as a Barrier to Entry, The Canadian Journal of Economics 2(4), 570-579 2. Structure-Conduct-Performance Approach •Arrow K. J. (1962), Economic Welfare and the Allocation of Resources to Invention’, in R.R. Nelson (ed.), The Rate and Direction of Economic Activity, Princeton University Press, N.Y., 609- 626 3.New Industrial Economics: 96 •Dasgupta P., J.E. Stiglitz (1980) Industrial Structure and the Nature of Innovative Activity, Economic Journal 90, 266-293 4.Empirical evidence: •Cohen W. (1995), Empirical Studies of Innovative Activity, in: Stoneman P., editor, Handbook of the Economics of Innovation and Technological Change, Oxford: Blackwell •Cohen W.M., R.C. Levin (1989), Empirical Studies of Innovation and Market Structure, in: Schmalensee R., R.D. Willig, Handbook of industrial Organization, Amsterdam: North Holland, 1989, 1059-5107 Lecture 5 (4-5/11/2020) The financing of R&D and innovation: problems and market failures. The “funding gap” problem for investment in innovation Market failure causes of under-investment in innovation: - Imperfect appropriability: new knowledge underlying the innovation is a (quasi) public good; - Positive externalities associated with innovation: the innovator cannot price and let society pay for all the innovation-related benefits. - Indivisibilities (i.e. fixed costs that cannot be broken down in smaller “parts”) related to the optimal scale of production (barriers to entry) and (technological) uncertainty. Both uncertainty and indivisibilities could be solved if capital markets worked perfectly. -> E.g. investors would correctly evaluate the expected value of any investment project (including R&D projects) and allocate funds to the projects with the highest returns. Uncertainty can be dealt with by investors diversifying their portfolios. However, there are reasons to expect frictions in capital markets and problems in financing innovation. Banks, venture capitalists, and other investors may have difficulties in understanding and evaluating the project if it is related to innovation. Hence, asymmetric information in financial markets is another type of market failure that can cause an underinvestment in innovation (if the innovator doesn’t have enough own funds to finance its projects). Asymmetric information in financial markets Asymmetric information problems may result in the failure of many markets in which the buyer and the seller don’t have the same information on the product traded (E.g. Akerlof (1970) in the “The Market for ‘Lemons’” analyzed this problem in the market for used cars). In financial/credit markets, where there is a time lag between the provision and the repayment of the money exchanged (and the contract between lender and borrower is mainly based on trust) the sources of asymmetric information can be: - adverse selection: the investor/lender cannot observe and assess the quality of the project/borrower when deciding to finance it or not. - moral hazard: the investor/lender cannot monitor the advancements/due diligence of the project/borrower after it has been financed. Hence the investor bears a higher level of uncertainty and risk of not being paid-back because of the limited liability of the borrower (e.g.: if the debtor doesn’t pay back, the lender has only limited legal/procedural tools to seize money from the borrower.) Consequences of asymmetric information: credit rationing L=loans, r=interest rate, S=supply of loans, D=demand of loans, r*=equilibrium market clearing interest rate (symmetric and complete information) r1=credit rationing interest rate (asymmetric information) Investors may try to compensate the higher riskbcaused by asymmetric information by simply asking higher interest rates (so-called risk premium). However, it is not always possible or convenient for the lender to increase the interest rate to its market clearing level, thus generating a situation of 97 equilibrium credit rationing: lenders don’t supply additional credit to borrowers, even if the latters are willing to pay higher interest rates. Hence there will be an excess demand for credit at the equilibrium rate of interest. The main intuition behind this result is that safe borrowers would not be willing to tolerate a high interest rate, as, with a low probability of default, they will end up paying back a large amount of money to the lender. Credit rationing: a simple numerical example 2 different projects G and B with different levels of risk that need an initial investment of I=100. G=risk-free project with a fixed return of 15% -> expected return= 115. B= risky project associated with an expected loss of 30% an a gain of 30% with probability p=½ -> expected return=70*(1/2)+130*(1/2)= 100. Scenario I: Bank interest rate of 10%: Both G and B will accept the loan. Expected gain by the bank: from G: 10 (with probability 1) from B: 10 (with probability 1/2) and -30 (with probability 1/2) =10*1/2 - 30*1/2 = 5 - 15 = -10 If the bank finance both G and B equally (e.g. because of asymmetric information), its expected profits will be 10-10=0 (= no incentive to fund the projects). Could the bank increase its expected profits by raising the interest rate? Scenario II: Bank interest rate of 20%: Only B will accept the loan (since G would have to pay back 20 from a risk-free gain of 15). Expected gain by the bank: from G: 0 (no loans) from B: 20 (with probability 1/2) and -30 (with probability 1/2) =20*1/2 - 30*1/2 = 10 - 15 = -5 If the bank charges an interest rate of 20% it will experience an expected loss of -5, since only B will accept the terms of the loan. The participation constraint for G involves a lower interest rate thanthe one of B. A too high interest rate will push the less risky borrower out of the market (another famous example: Akerlof (1970) market for lemons of used cars). Equilibrium credit rationing: Stiglitz and Weiss (1981) An entrepreneur wants to invest an amount I>0 in a 1 year long innovative R&D project. The entrepreneur doesn’t own money (A=0) and has to ask for funds to the (perfectly competitive) credit market to finance his/her project. In the market there are 2 types of borrowers: G (good) and B (bad). • In t=1: the lender (bank) decides whether to finance the project or not and sets the interest rate r. • In t=2: if financed, the result of the R&D project is observed. It can fail and have no return (the entrepreneur bankrupts) or can succeed and have a return of X (with X>I), in this case the entrepreneur realizes positive profits and repays the debt. The probability of success for the good entrepreneurs (G) is p, whereas for the bad entrepreneurs (B) is q, with p>q. No discount rate assumption (1€ in t=1 is equal to 1€ in t=2). Stiglitz and Weiss (1981): symmetric information Scenario I: Symmetric information (no adverse selection). In this scenario investors are able to observe the quality of the entrepreneur (G or B) asking for credit and set different loan terms (i.e. interest rates) for the two “risk profiles”, under the “expected zero profits” constraints for the lender (competitive credit markets). In particular: - The interest rate paid by good entrepreneurs (G) will be: p*(1+rG)I+(1-p)*0 – I = 0 -> p*(1+rG)I=I -> rG=(1/p)-1=(1-p)/p - The interest rate paid by bad entrepreneurs (B) will be: q*(1+rB)I+(1-q)*0 – I = 0 -> q*(1+rB)I=I -> rB=(1/q)-1=(1-q)/q It follows that: rG<rB because p>q and so (1-p)/p<(1-q)/q Low-risk entrepreneurs G pay a lower interest rate with respect to high-risk ones B. Stiglitz and Weiss (1981): symmetric information Scenario I: Symmetric information (no adverse selection). In regulated credit markets (as it happens in almost all countries) there is maximum legal non-usury interest rate rMAX at which loans can be made.