Science and Innovation Policies at the Digital Age Dominique Guellec Science and Technology Policy OECD Grenoble, December 2 2016
Structure of the Presentation What does digitalisation mean for science and innovation policies? 1. What does digitalisation mean for science and innovation 2. Access to data 3. Patent rights 4. Competition policies 5. Innovation support 6. Policies and Governance
Moore s Law (source: wikimedia)
What is digitalisation? New technologies which underpin social and economic transformations: The Cloud Big Data Artifical Intelligence Internet of Things 3D Printing Etc.
What does digitalisation mean? Digitalisation means that most human activities (including research or innovation) have an "avatar" (a mirror image) in the digital sphere (the Internet, the cloud), from which they can be controlled developed, changed and oriented and which works a partly different way than the original in the real sphere, offering new opportunities and risks, and requiring adapted policies.
What does digitalisation mean?
What does digitalisation mean? Digitalisation of science and innovation = Open Science + Digital Innovation
What does open science mean? Open Science: New ways of doing research and sharing research inputs and outputs enabled by digital tools Gathering data (DNA, geolocation, archives) Analysing data (data analytics; AI) Sharing data (reuse allowing broader use of data and verification of research results) Publishing results (articles on the web; databases) Managing research (agenda setting; research projects; researchers careers; evaluation)
What does digital innovation mean? Digital Innovation: new innovation processes and products enabled by digital tools or embodied in data and software Connected objects e.g. the autonomous car New services (the sharing economy) Servitisation (goods turned into services: cars become transportation) Platforms and platformisation; multi-sided markets Cooperation in innovation open innovation
The Economics of Digital Innovation Digitalisation implies the convergence of information economics and "real world economics", notably innovation economics: the mechanisms and processes which govern this emerging, digital world are a mix of those governing the information world and the real world. Innovation (with tangible goods) Partial non-rivalry Digital Innovation Total non rivalry + Network effects Scale economies «Scale without mass» Creative destruction Winner take all Digital innovation economics is «innovation economics on steroids»
Policies for promoting Data Access Data has many properties of a public good Policy objective: Ensure the broadest access to data and knowledge (favouring reuse, reducing social cost) while respecting constraints in relation with: ethics (privacy etc.) and economics (incentives to produce and disseminate the data).
Policies for promoting Data Access (2) Private data Data are a source and a result of digital innovation, their ownership & rules of access govern value sharing and incentives: Private business & innovation data: owned by companies as part of their innovative process (tests etc.) all the more strategic in a context of global competition. Personal data on platforms: current deal is: free services in exchange for personal data => should users own (and possibly sell) their personal data?
Policies for promoting Data Access (3) Public research data Incentives and obligation for researchers of sharing data, e.g. in the funding arrangements, while preserving incentives for researchers to produce data - tragedy of the commons vs of the anticommons; Documentation and sharing of meta data, in order to ensure traceability and integrity of the data (reproducibility); Funding of data repositories (business models) and of functionalities facilitating use of the data; Publications: evolution in the business model of publishers, aimed at making published papers open and free access (gold, green etc.).
Policies for promoting Competition Objective: Ensure that markets keep fuelling innovation. Innovation dynamics: Big established players do "big innovations (indivisibilities) and are integrators of the many small ones (standards, platforms); young and small players do more radical and smaller innovations. The issue: digital innovation & platforms are subject to scale & network economies, hence favouring winner take all; but they also reduce entry cost to markets and costs of expanding businesses ( scale without mass ) Competition policies should not inhibit innovation and integration by dominant players, but they should bar them from exercising excessive market power vis à vis suppliers & customers need a new competition policy framework. A matter of particular interest is the systematic takeover of successful start-ups by the GAFA.
NPR will create new challenges for the intellectual property system Copyrights, patents, trademarks, designs T M
NPR will create new challenges for the intellectual property system Copyrights, patents, trademarks, designs For example: - We may need to rethink some aspects of IP rules in a world where machines can invent things. - Will 3D printing create challenges for patent eligibility, for instance T Mif printed human tissue improves on the original?
Adapting Science and Innovation Policies
in a context where many countries have reduced public spending on R&D R&D budgets 2008-2015, G7 and Korea, Index 2008 = 100 many countries have reduced public spending on R&D
Policies for Supporting Innovation Public research has been at the origin of most digital technologies: Government need to continue supporting fundamental research («Industries du Futur», «Industry 4.0» in Germany etc.) knowing that the distinction between fundamental & applied research or development is blurred in frontier technologies (e.g. AI) Leveraging open innovation and ecosystems: favouring the openness of public research, adapting competition policy to cooperation in innovation.
Agility and Speed in Policies The research agenda in advanced domains is shifting very quickly and impossible to predict; hence targeted support should be agile as well, adapting to changing targets, otherwise businesses will rather avoid participating in public programs (as it generates a risk of being locked in an outdated technology). Agility: unicorns grow very fast ("scale without mass"), government support, when needed, has to keep pace. For instance, «French Tech» has certain accelerated procedures for providing funds.
Transitioning to Digital Innovation Digitalisation often means changing business processes or even business model. That is complex for large firms (which sometimes fail: Kodak, Nokia), but it is even more difficult for the small ones (except start-ups which are born digital): lack of knowledge, funding, no right to fail:
Well designed institutions needed for technology diffusion and adoption Enterprises using cloud computing services by employment size class, 2014 As a percentage of enterprises in each employment size class % All enterprises 10-49 50-249 250+ 70 60 50 40 30 20 10 0 Source: OECD Science, Technology and Industry Scoreboard 2015.
We need well designed institutions for technology diffusion and adoption Enterprises using cloud computing services by employment size class, 2014 As a percentage of enterprises in each employment size class % All enterprises 10-49 50-249 250+ 70 60 50 40 30 20 10 0 We have to execute quickly, otherwise those who are already leading in digital will snatch the industrial production from us. Angela Merkel, World Economic Forum 2015 Source: OECD Science, Technology and Industry Scoreboard 2015.
We need well designed institutions for technology diffusion and adoption Enterprises using cloud computing services by employment size class, 2014 As a percentage of enterprises in each employment size class % All enterprises 10-49 50-249 250+ 70 60 50 40 30 20 10 0 - Only 4% of German businesses have implemented digitalised and networked production processes or have plans to begin doing so. We have to execute quickly, otherwise those who are already leading in digital will snatch the industrial production from us. ZEW Mannnheim 2015 survey of 4,500 German businesses Angela Merkel, World Economic Forum 2015 Source: OECD Science, Technology and Industry Scoreboard 2015.
Transitioning to Digital Innovation Companies which don t adapt to digitalisation will fail and disappear. It might be appropriate to let market selection operate in a number of cases and let go those who fail to adapt but cleansing means that valuable assets would simply vanish the social cost of creative destruction. Policies: Provide various types of support to SMEs in their transition to digitalisation, e.g. technology demonstrators, expert advice or financial support (e.g. loan guarantees) «Nouvelle France Industrielle» (France), «Smart Industry» (NL)
Digitalising Policies & Governance Digitalisation offers new opportunities for the design, implementation, monitoring and evaluation of policies => Government need to seize them
Research Information Management Systems (RIMSs) RIMS is a digital infrastructure for collection, management, and analysis of research inputs, outputs, and outcomes. RIMSs are based on the synergy between Current Research Information Systems (CRISs) and Institutional Repositories (IRs) RIMSs may be national or institutional.
Functions of RIMSs 1 to link research inputs, outputs, and outcomes 2 to study developments in research, research workforce, and student flows 3 4 5 to support dissemination of ideas and research collaboration to optimize state funding on research to reduce administrative burden of researchers 6 to support commercialization of research results 7 to justify the value of science to citizens
Institutional RIMSs RIMS Projects R&D personnel Results Organizations Equipment Publications CV Patents Partners Science field Location Funding Skills Events Language
Conclusion Digitalisation of science and innovation raises radically new conceptual and policy challenges, much policy learning is needed but innovation economists are well equipped to address them only if they can update their conceptual tool box!
Thank you dominique.guellec@oecd.org Innovationpolicyplatform.org