FINANCE, COMPETITIVENESS & INNOVATION INSIGHT | FIRMS, ENTREPRENEURSHIP & INNOVATION Promoting Innovation in China: Lessons from International Good Practice © 2020 The World Bank Group 1818 H Street NW Washington, DC 20433 Telephone: 202-473-1000 Internet: www.worldbank.org All rights reserved. This volume is a product of the staff and external authors of the World Bank Group. The World Bank Group refers to the member institutions of the World Bank Group: The World Bank (International Bank for Reconstruction and Development); International Finance Corporation (IFC); and Multilateral Investment Guarantee Agency (MIGA), which are separate and distinct legal entities each organized under its respective Articles of Agreement. We encourage use for educational and non-commercial purposes. The findings, interpretations, and conclusions expressed in this volume do not necessarily reflect the views of the Directors or Executive Directors of the respective institutions of the World Bank Group or the governments they represent. The World Bank Group does not guarantee the accuracy of the data included in this work. Rights and Permissions The material in this publication is copyrighted. Copying and/or transmitting portions or all of this work without permission may be a violation of applicable law. The World Bank encourages dissemination of its work and will normally grant permission to reproduce portions of the work promptly. All queries on rights and licenses, including subsidiary rights, should be addressed to the Office of the Publisher, The World Bank Group, 1818 H Street NW, Washington, DC 20433, USA; fax: 202-522-2422; e-mail: pubrights@worldbank.org. Photo Credit: Shutterstock.com FINANCE, COMPETITIVENESS & INNOVATION INSIGHT | FIRMS, ENTREPRENEURSHIP & INNOVATION TABLE OF CONTENTS ACKNOWLEDGMENTS III ABSTRACT V INTRODUCTION 1 POLICIES TO SUPPORT TECHNOLOGICAL PROGRESS IN ADVANCED ECONOMIES 3 CHINA’S PROGRESS IN BUILDING INNOVATION CAPACITY 5 China’s Spending on its National Innovation System 6 China’s Innovation Policies 8 Research Capacity 11 Technology Upgrading and Diffusion 15 China’s Indicator—Based Innovation Rankings 16 INNOVATION REFORM AGENDA 17 Enhancing Government Capabilities 17 Balancing National and Regional Support for Innovations 22 Investing in Basic Research 23 Leveraging Open Innovation 25 Facilitating Technology Diffusion 27 SUMMARY OF POLICY RECOMMENDATIONS 33 ENDNOTES 37 REFERENCES 43 LIST OF BOXES Box 1: Innovative SOEs in China 11 Box 2: Public Expenditure Reviews of Innovation Policy 21 Box 3: Business Support Services In Singapore 30 LIST OF FIGURES Figure 1: Reducing Distortions, Accelerating Diffusion, Nurturing Discovery 2 Figure 2: The Innovation “Production Function” 2 Figure 3: China’s R&D Spending has Grown Rapidly, 1995-2017 6 Figure 4: China’s Share in Global R&D Spending, 1981-2015, PPP 7 Figure 5: Distribution of Innovation Policy Instruments by Year and Agency 8 Figure 6: Institutional Structure of Central Government Spending on Innovation 9 PROMOTING INNOVATION IN CHINA: LESSONS FROM INTERNATIONAL GOOD PRACTICE I Figure 7: Bachelor and PhD Degrees in Science and Engineering are Climbing Rapidly, 2000–14 12 Figure 8: Patent Applications in Percent of the World’s Total, 2018 13 Figure 9: Firms Have Become the Most Important Source of Patents in China 15 Figure 10: A Lack of Alignment Between Program Objectives and Budgets Hampers Outcomes 18 Figure 11: Priority Forms of National Innovation Policy Coordination, 2012 20 Figure 12. China’s R&D Spending for Basic Research vs. the OECD Average 25 ENTREPRENEURSHIP & INNOVATION FINANCE, COMPETITIVENESS & INNOVATION INSIGHT | FIRMS, ENTREPRENEURSHIP, ACKNOWLEDGMENTS T his paper has been authored by Denis Medvedev, Practice Manager, the World Bank, Marcin Piatkowski, Senior Economist, the World Bank, and Shahid Yusuf, World Bank Consultant and Chief Economist, Growth Dialogue, GWU, Washington, D.C. The paper builds on background papers and inputs Rogers, Qursum Qasim, Gregor Schueler, Hua Shen, commissioned for the joint World Bank Group and Hoon Soh, Lin Shi, Mingzuo Sun, Zhang Taolue, Ari Development Research Center of the State Council van Assche, Jo van Biesebroeck, Danqing Zhu, L. of the People’s Republic of China’s 2019 report Colin Xu, and Xiaobo Zhang. on Innovative China. New Drivers of Growth, The authors appreciate additional comments and including contributions by Stefan Beisswenger, suggestions from Martin Raiser, Country Director Philipp Boeing, for China and Mongolia and Director for Korea, and Xavier Cirera, Marcio Cruz, Banning Garrett, Irina Astrakhan, Practice Manager, East Asia and Song Hefa, Justin Hill, Victor Mulas, Anastasia Pacific, Finance, Competitiveness and Innovation Nedayvoda, Ruimin Pei, Dan Prud’homme, Juan Global Practice. PROMOTING INNOVATION IN CHINA: LESSONS FROM INTERNATIONAL GOOD PRACTICE III SECTION TITLE IV FINANCE, COMPETITIVENESS & INNOVATION INSIGHT | FIRMS, ENTREPRENEURSHIP & INNOVATION ABSTRACT C hina considers innovation be one of the key drivers of its future growth and convergence with more developed countries. It spends more than 2.2 percent of GDP on R&D, above the average for the European Union, is a global leader in domestic patents, and has developed ground- breaking advances in key sectors such as high-speed trains, e-commerce and mobile payments. However, the quality of patents has been slower to improve, Chinese firms remain dependent on foreign suppliers in a number of core high-tech components, and resources do not flow easily to more productive firms resulting in large productivity gaps between market leaders and remaining enterprises. In order to restart its productivity engine and support continued technological catch- up, China must revise its approach to innovation policy. This paper takes stock of China’s progress in building a modern national innovation (NIS) system, reviews international good practice in promoting innovation and shares policy recommendations to help China sustain its drive to become one of the global innovation champions. PROMOTING INNOVATION IN CHINA: LESSONS FROM INTERNATIONAL GOOD PRACTICE V SECTION TITLE VI FINANCE, COMPETITIVENESS & INNOVATION INSIGHT | FIRMS, ENTREPRENEURSHIP & INNOVATION INTRODUCTION T he growth of the Chinese economy is stabilizing around a “new normal”, with the government now attaching greater importance to the quality of growth than to its pace. To sustain this new growth path, productivity gains arising from technology adoption and innovation will need to displace investment as the principal driver of growth, and China will need to enhance the potential of its innovation system to fully leverage advances in technology. This paper assesses the country’s science, technology, and innovation capacity and research and development outputs, reviewing existing policies and identifying weaknesses. It then recommends measures that will aid China in its transition to an innovation- and productivity-led growth path by improving its national innovation system and by closing the distance to the technological frontier. Many Chinese firms have caught up technologically the global production frontier by nurturing the by first linking with lead firms in global value chains discovery and development of novel technologies and more recently through domestic applied research (“frontier innovation”), as in figure 1. and technology acquisition via outward foreign direct Chinese authorities have mobilized a wide range investment. Leading Chinese firms have joined their of policies to support improvements along the foreign competitors as generators of knowledge by entire innovation “production function”, from investing in scientific research and by mobilizing inputs to outputs to outcomes (figure 2). The global talent. But the closing of technology gaps current emphasis of the authorities is on financial by industry leaders has not been enough to stem a and supply-side support, with applied research and decline in aggregate productivity growth and, despite product development receiving primacy. Against ground-breaking advances in several key sectors, the backdrop of continued rapid increases in such as high-speed trains, e-commerce and mobile China’s science, technology, and innovation (STI) payments, China generally remains fairly distant expenditure, especially at the local level (whose from the global technological frontier. Annual total share in total STI spending has already surpassed factor productivity (TFP) growth fell sharply from the central government’s), policy priorities may 2.7 percent in the early 2000s to only 1.1 percent in need to be reordered to respond to pressing needs 2008–2017. TFP growth in manufacturing has nearly not adequately addressed by the private sector. ground to a halt (DRC-World Bank 2019). The key objective for China’s STI system should Restarting the productivity engine requires be to transition from quantity to quality in public resolving challenges related to strengthening spending. To achieve this shift, China ought to technology diffusion from leading firms to others, reform its STI system by enhancing government commercializing ideas, deepening managerial skills, capabilities, balancing national and regional creating an open innovation system, and building support for innovation, investing in basic research, basic research capabilities. China has considerable leveraging open innovation, facilitating technology scope for increasing productivity by i) eliminating diffusion, and encouraging firms to improve their market distortions that impede the sharing of best managerial practices. In addition, China would practices among domestic firms, ii) facilitating exit also be well-advised to continue to strengthen of poorly performing firms, and iii) closing some IPRs, promote a more bottom-up, market-based of the gap with the OECD production frontier by innovation policy, streamline innovation policies prioritizing the diffusion of existing technology and and expand the use of monitoring and evaluation in innovations (the so-called “technology absorption” assessing the effectiveness and efficiency of public strategy). And it can build the capabilities to extend support for innovation PROMOTING INNOVATION IN CHINA: LESSONS FROM INTERNATIONAL GOOD PRACTICE 1 Figure 1: Reducing Distortions, Accelerating Diffusion, Nurturing Discovery Focusing on the second D on accelerating diffusion of Services innovation and technology is Nurturing likely to result in the largest payoff Discovery for China’s foreseeable future Accelerating Diffusion China remains quite distant from the global technological frontier Reducing Distortions China could double its GDP by simply catching up to OECD China’s OECD countries in total factor productivity production production by adopting already existing frontier frontier technology and managerial practices Goods Source: DRC-World Bank 2019. Figure 2: The Innovation “Production Function” Innovation Inputs and Innovation Outputs Impact Knowledged Activities and Outcomes • Technology • New or improved products Firm Growth • Equipment and services (New demand or increased • New or Improved business market share due to enhanced • R&D processes quality or cost advantage) • Intellectual property use • New business models • Human capital • New or improved organizational Productivity Growth • Training (Improved businesss and managerial practices • Engineering and design processes and technology) • Patents and other intellectual • Software and databases property • Managerial and organizational capital and practices Economic Diversification Source: Cirera and Maloney 2017. INTRODUCTION 2 FINANCE, COMPETITIVENESS & INNOVATION INSIGHT | FIRMS, ENTREPRENEURSHIP & INNOVATION POLICIES TO SUPPORT TECHNOLOGICAL PROGRESS IN ADVANCED ECONOMIES A dvances in technology have been increasingly driven by the growing ability of technologies to build upon each other. Electronic, telecom, digital, and biogenetic technologies are at the forefront of this process of mutual amplification.1 In manufacturing circles, this change is known as “Industry 4.0”, and is characterized by robotics, artificial intelligence, and the industrial Internet of Things (IoT). Emerging technologies, including cloud storage and computing, have laid a foundation for generating, sharing, and applying data on a vast scale, permeating every corner of the economy and public management. These advances have dramatically reduced entry In mature national innovation systems in developed costs for start-ups based on information and digital countries, where the focus is on pushing the technology and increased the speed at which their global knowledge frontier, emphasis has recently services can be scaled-up. The internet and the been placed on accelerating time-to-market and global positioning system (GPS)—two “platforms encouraging bottom-up innovation to better take of platforms”—have enabled the creation of advantage of emerging technologies. These policies firms such as Google, Facebook, Baidu, Tencent, help commercialize scientific and technological Didi, and Mobike. Other technologies, including outcomes by fast-tracking innovation and accelerating 5G mobile networks, are helping develop smart patent review. In its Horizon 2020, the European energy, transport, and urban infrastructure and Commission launched a pilot to fast-track market more widely apply biomechanical, genomic, and entry of innovative products and services. Reforms in robotic technologies in agriculture. Blockchain—a the US sought to improve patent quality and shorten distributed digital ledger originally designed for the approval process, including the milestone 2011 accountancy of the digital currency Bitcoin2— America Invents Act, which stipulates that after March is spreading beyond the financial and scientific 16, 2013 the First Inventor-to-File Rule would apply domains to public service, digital identification, to all American patents. This replaces the long- and supply chain management. prevailing First to Invent System.3 The Act, which aligns U.S. practice with that of all other countries,4 Green technologies ranging from photovoltaic and encourages individuals or firms with potentially wind power supply, to next generation battery storage, lucrative inventions to not delay the patent filing, to low-power information and communications even if it is provisional, to avoid patent rights being technology (ICT) incorporated in “intelligent” awarded to a later inventor. It also establishes new manufacturing systems could, beyond spurring mechanisms, such as a post-authorization review economic growth, reduce greenhouse gas emissions. procedure to improve the quality of patents and They are reducing energy use and increasing the safeguard the rights of inventors. To implement the productivity of traditional industries. They are new Act, the court identified the scope of patentability expanding environment-friendly services and clean and guiding principles to reduce patent speculation industries like renewables, electric cars, and waste and abuse of litigation opportunities. The U.S. Patent management. And they are stimulating the emergence Office is striving to cut the average approval time of new, yet unknown, industries (World Bank, 2018). from 35 months to 20 months and to bring the most Advances in bio- and nanotechnologies also show valuable patented technologies to the market within promise in seeding new industrial activities. 12 months. PROMOTING INNOVATION IN CHINA: LESSONS FROM INTERNATIONAL GOOD PRACTICE 3 European countries have identified technologies of growth. Based on comprehensive and rigorous with large growth potential by leveraging the monitoring and evaluation, smart specialization entrepreneurial drive of the private sector, working helps governments support winners and cut off through discovery and experimentation. This losers, much in line with Rodrik’s concept of what sidesteps the problem created by weakening market constitutes a good industrial policy (Rodrik, 2008). and technology signals once countries operate close For most EU countries, smart specialization underlies to the technological frontier. Such a bottom-up, their regional and national innovation policies innovation-based industrial policy has been called (World Bank, 2016). Its adoption is a condition for “smart specialization” (Foray, David and Hall, EU states to tap billions of euros from the EU’s 2009; Foray, 2015). It prioritizes public support for Structural Funds in 2014–20. China may well find innovation in economic activities that may not yet be it useful to adopt such a bottom-up, market-oriented fully “visible” but have the largest developmental, approach, as a complement to its current top-down innovative, and spillover potentials. It relies on the policies. It could pilot this at a regional level and, “entrepreneurial discovery” to identify, select, and depending upon results, roll it out across the country. modify smart specializations and to provide public support based on the information from the markets. Other measures intensify the demand for business This process unites the private sector, academic innovation. U.S. Federal Government departments bodies, and public administrations in an interactive use technology awards to encourage innovation. process of mutual learning and knowledge sharing.4 The “Grand Challenges” of the U.S. Defense It also strives to consolidate all knowledge in one Advanced Research Projects Agency (DARPA) place to reduce coordination failures and better offer competitors a prize if they can achieve a guide private decisions. particular objective, such as perfecting a viable autonomous vehicle. Several EU countries A bottom-up approach also helps reduce the risk of (Ireland, for example) use public procurement capture of public innovation policies by incumbents, to advance green growth objectives, incorporate that is, companies and industries that already have a innovative products in procurement criteria, conduct strong market position, rather than emerging firms precompetitive procurement, and provide early that may still be too weak to affect public policy (and markets for innovative products. Such procurement which may not reach the critical mass of development is a key driver of development and technological without public support). The public sector then progress in architecture, health services, and concentrates public resources on the most promising public transport.5 activities and supports them at their critical junctures POLICIES TO SUPPORT TECHNOLOGICAL PROGRESS IN ADVANCED ECONOMIES 4 FINANCE, COMPETITIVENESS & INNOVATION INSIGHT | FIRMS, ENTREPRENEURSHIP & INNOVATION CHINA’S PROGRESS IN BUILDING INNOVATION CAPACITY A lthough China has reached the world technology frontier in some areas (Alibaba, WeChat payments, high-speed trains) and is approaching it in others, a wide gap with major developed countries in core technical capacity remains. Within China, technological, productivity, and managerial capacity differences between leading players and the majority of Chinese firms are also large. The country thus needs to further strengthen its innovation capacity and enhance its innovation capabilities, using initiatives that have delivered results in advanced economies but which are adapted to the local context. Innovation requires R&D spending, some on basic While a strong performer in some inputs (R&D research but the bulk on downstream development spending) and outputs (patents), China falls short of and activities that produce marketable outcomes. A leading economies across a range of other important simplified linear view sees money going in at one end inputs. Despite much recent progress, China lags of the research pipeline, and research papers, patents, in high-quality human capital (measured by R&D and innovative products and services emerging from professionals, PhD students, and tertiary enrollments the other, due to the efforts of an army of highly relative to the country’s population), internet access skilled workers aided by specialized measuring and use, and the overall business and regulatory and testing equipment.6 This perspective appeals to environment. Chinese firms also need to strengthen planners, policymakers, and business managers, and management quality to catch up with practices there is empirical support for the belief that returns to in frontier countries. Findings from the World R&D can be quite high if supported by a conducive Management Survey (WMS) suggest that Chinese environment.7 Scientific papers and patents can managers excel at meeting short-term production begin pouring out quite rapidly, as China shows.8 targets and at competing based on price, but among 40 countries surveyed, they are among the less But it is also apparent from the faltering productivity sophisticated in innovation-related practices, such performance of industry and of research9 in some as long-term strategic planning and staffing.12 western countries and in China, that a linear strategy Chinese firms’ managerial capabilities are therefore with its emphasis on R&D spending and on a few weaker than the sample’s average, with China metrics is unlikely to yield sustained productivity located between Greece and Turkey (Bloom, Sadun gains.10 A review of five prominent Chinese industries and Van Reenen (2017); Cirera and Maloney 2017). finds that Chinese firms “rely on their low-cost While the worst managed Chinese companies are advantage and mass production mode as their main only slightly worse than their U.S. counterparts, the competitive advantage instead of product innovation difference between the top Chinese and U.S. firms and product and service quality” (Cooke, 2018). in the quality of managerial practices is much larger A more comprehensive, multifaceted strategy is (DRC-World Bank 2019). This suggests that there therefore warranted, one that balances a strong is much room for improving managerial practices emphasis on measures that improve management across the whole spectrum of firms, with the payoff skills at the firm level—and that accelerate the greatest when upgrading the management of leading diffusion of technology and the commercialization Chinese firms. of innovations—with the basic research that leads to scientific discovery.11 PROMOTING INNOVATION IN CHINA: LESSONS FROM INTERNATIONAL GOOD PRACTICE 5 Figure 3: China’s R&D Spending has Grown Rapidly, 1995-2017 5.0 4.5 4.0 3.5 R&D/GDP in Percent 3.0 2.5 2.0 1.5 1.0 0.5 0 0 10,000 20,000 30,000 40,000 50.000 60,000 70,000 GDP Per Capita Japan Korea USA China Germany Source: DRC-World Bank 2019. China’s Spending on its In absolute numbers, China’s R&D spending (in National Innovation System purchasing power parity (PPP) terms) is now second only to that of the United States (figure 4) and was China has a comprehensive national innovation estimated to account for about 20 percent of global system (NIS), building on policies introduced in the R&D outlays in 2017. The rate of growth is higher early 1990s. In addition to the central ministries and than in the EU and the United States, suggesting that specialized agencies there are many more innovation- China’s share in global R&D spending will continue oriented institutions and agencies at regional and to rise. Reaching the R&D target of 2.5 percent of local levels. There is also a well-developed system of GDP by 2020 or so would increase China’s share incubators and high-tech and science and technology in global R&D spending to about one-fourth.14 The parks, estimated to exceed 3,000 in number.13 In United States and China alone would then account 2015, there were 146 national high-tech parks, which for more than half the world’s R&D spending. housed companies responsible for one-third of R&D enterprise spending in the country (OECD, 2017). National and regional R&D spending has risen rapidly, with regions now spending more than half STI spending grew from RMB 5 trillion in 2007 to of the total (Science and Technology Annals, 2018). RMB 18.7 trillion in 2016, but recently slipped as Even though all provinces promote R&D and a share of total public spending (China, Annals of encourage patenting by setting quotas and offering Science and Technology, 2017). Spending on R&D incentives to maximize filings, the disparities alone amounted to 2.13 percent of GDP in 2017, a among regions are wide on R&D intensity and little below the OECD average, which was 2.37 in patent filings—coastal regions dominate (Sun et al, 2017 but higher than expected given China’s per 2015). That would be in line with the same patterns capita income (figure 3). in advanced countries, where innovation activity is CHINA’S PROGRESS IN BUILDING INNOVATION CAPACITY 6 Figure 4: China’s Share in Global R&D Spending, 1981-2015, PPP 600 500 400 Billions of Current PPP Dollars 300 200 100 0 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 00 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 19 19 19 19 19 19 19 19 19 19 19 19 19 19 19 19 19 19 19 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 United States EU France Germany United Kingdom China Japan Rep. of Korea India Source: Authors based on National Science Board (2018)15 Note: Foreign currencies converted to current dollars through PPP. concentrated in a few regions, such as Silicon Valley There is ongoing debate about the overall returns to in the United States, Emilia Romagna in Italy, and China’s R&D. During 2006–10, returns were relatively Baden-Württemberg in Germany. low, seemingly due to a lack of complementary factors such as quality of scientific infrastructure, overall Most national and provincial public funding in functioning of the NIS, and firms’ capabilities (Goni China is directed to a few strategic industries, and Maloney, 2017). But there is also evidence that selected by the national and regional governments private spending has boosted output and productivity. as part of their top-down innovation strategy. In One study estimated that, among 12,000 Chinese firms 2015, rail and other transport equipment received in 120 cities during 2002–04, firm output increased more than one-fourth of all public support for by around RMB 0.4 for each additional RMB 1 spent innovation, followed by computers (17 percent) on R&D in the previous year, similar to Japan and and general-purpose machinery (7 percent) the United States in the 1970s (Goh, Li and Lixin, (OECD, 2017). Unlike the vast majority of other 2015). Although there is much evidence of public middle-income and even some high-income policy supporting several successful technologies countries, China’s enterprise sector accounts for (such as high-speed trains), the overall efficiency of three-fourths of total R&D spending (DRC-World public spending is mixed (Liu et al, 2017). This is Bank, 2019). While this share is above the vast likely because public support seems to “be strongly majority of other middle-income economies as biased in favor of SOEs, especially those owned well as the OECD average of two-thirds, as much by local governments” even though “private sector as one-third of the total in 2015 was due to R&D firms exhibit a higher rate of innovation per renminbi outlays of SOEs (down from 55 percent in 2005) invested in R&D than SOEs” (Wei, Zhuan and Zhang, (Molnar, 2017). 2017). PROMOTING INNOVATION IN CHINA: LESSONS FROM INTERNATIONAL GOOD PRACTICE 7 Figure 5: Distribution of Innovation Policy Instruments by Year and Agency 50 1 SASAC PDCC 1 45 MOC 1 MHRSS 1 40 MEP 1 CFDA 1 CBRC 1 35 1 ACFTU SAT 2 30 PBC 2 ODCC 2 MIIT 2 25 2 GAQSIQ CSRC 2 20 CAST 2 MOA 3 15 SIPO 4 NDRC 6 CCCPC 6 10 MOE 11 CAS 17 5 MOF 19 MOST 26 SC 56 0 2011 2012 2013 2014 2015 2016 2017 0 10 20 30 40 50 60 Source: Authors’ calculations based on data collected by Ruimin Pei and Hua Shen. China’s Innovation Policies technology transfer, and R&D and non-R&D innovation). And they use a wide range of instruments: China’s top-down approach to innovation policy fiscal incentives, grants, loan guarantees, vouchers, dates to the launch of the Medium and Long Term equity, public procurement, technology extension Plan (MLP) in the mid-2000s to “re-conceptualize the services, incubators, accelerators, competitive broader innovation policy framework”, “introduce grants and prizes, science and technology parks, the theme of indigenous innovation”, create a and collaboration and networks (many of these Leadership Small Group to guide and coordinate instruments are highlighted later in this chapter). technology development, and make it clear that implementation would be “with key point projects The number of instruments per period reflects and key point tasks”. The MLP committed China to the five-year plan cycles, with most toward the achieve four broad objectives by 2020: to allocate front end of each cycle (figure  5, left panel). The 2.5 percent of GDP to R&D, to source 60 percent instruments in the policy mix are spread across of growth from the “contribution of S&T progress”, 24 government entities, 16 of which implement to base 70 percent of production on homegrown between 1 and 3 instruments each (figure  5, right technologies, and to raise the share of strategic and panel). A full one-third of the number of instruments emerging industries to 15 percent of GDP (Ling and are implemented by the State Council, more than Naughton, 2016). twice the proportion of instruments of the next entity, the Ministry of Science and Technology To further these objectives, more than 170 policies (MOST). The State Council is therefore critical to supporting STI have been introduced since 2011.16 overall implementation of the policy mix, though They span socioeconomic objectives (productivity, the number of policies does not necessarily mesh diversification, human capital, entrepreneurship, and with budget allocations (data are not available on inclusion) and STI objectives (research excellence, the public resources allocated to each policy). CHINA’S PROGRESS IN BUILDING INNOVATION CAPACITY 8 Figure 6: Institutional Structure of Central Government Spending on Innovation Major Public Expenditure on STI Major S&T Programs Inter-ministerial Joint Council NSFC MOST MOE NDRC MIIT National Natural National S&T S&T Base and ‘Double First-Class Tech-Renovation Special Fund for the Science Fund Major Projects Talent Special Fund University Projects’ Special Fund development of SMEs Tech Innovation Special Fund for the Emerging Industry Guiding Special Fund development of VC Guidance Fund Electronic Info. Ind. National Key High-tech National IC Industry R&D Plan Industrialization Fund Invesment Fund Source: DRC-World Bank, 2019. The large number of entities with only one, Basic research is mostly funded through the NSFC two, or three instruments raises the question of and the China Academy of Sciences (CAS). Both integration of their interventions into the overall saw their budgets more than double between 2011 innovation policy. Despite the recent reform of and 2017 to more than RMB 30 billion. Overall the public R&D support system, which helped outlays on basic research increased to about 5.3 streamline STI spending by strengthening the role percent of total spending, with further spending of MOST in managing the flows of public R&D, increases envisaged.18 Spending on applied research, some researchers argue that coordination of policies mainly funded through MOST, MIIT, and CAS, has could be strengthened and the overlap among R&D also increased.19 programs could be significantly reduced (Liu et al., Policy instruments seem to favor financial instruments 2017). Moreover, the budget allocation system is over regulatory, advisory, and other types of support.20 multitiered and confusing because instruments and Such emphasis suggests that the main market failure budgets are ascribed to one entity at the general that the authorities seek to address has to do with level but the action plans may be carried out by funding and access to finance. Tax incentives, other agencies. matching grants, and funding for collaborative Spending by ministry indicate that the responsibility activities cover twice the number of instruments of for implementing R&D policy rests largely with the next four: i) business advisory and technology five agencies: MOST, the National Development extension services, ii) science parks, iii) incubators and Reform Commission (NDRC), the National and accelerators, and iv) quality infrastructure and Science Foundation of China (NSFC), the Ministry standards for innovation.21 In addition, the authorities of Industry and Information Technology (MIIT), provide substantial financial support through more and the Ministry of Economy (MOE) (figure  6). than 2,000 government-guided funds and a plethora Together, they account for almost one-third of total of venture capital and private equity funds co-funded public R&D spending.17 by central and local governments. PROMOTING INNOVATION IN CHINA: LESSONS FROM INTERNATIONAL GOOD PRACTICE 9 Supply-side interventions seem to dominate support Further inquiry would be needed to determine instruments on the demand side (in terms of the whether tax incentives and direct support policies number of instruments, not the allocated budgets), are addressing the constraints inhibiting high- and instruments that aid applied research outnumber productivity entrants. Such inquiry should be support to basic research by a factor of 2 to 1. Most based on (currently unavailable) data on the budget policies are targeted at young firms (pre-seed, seed, allocations for each support instrument, which and start-up) and micro and small firms. Of 42 would provide information for policymakers to instruments devoted to businesses along some stage determine whether the policy mix is appropriate for of their life cycle, 36 are focused on seed funding the needs of national and local innovation systems. or the start-up phase, while scaling up and mature Most developed countries provide such data as a stages are the objective of only three instruments matter of course to adjust their innovation policy each. This pattern is consistent with diversification, mix. China could take steps in this direction. one of the main economic objectives of Chinese Also important is analyzing the efficiency of innovation policy given the sharp decline in public support for innovation, which is rarely, if the contribution of new entrants to firm-level ever, monitored and evaluated for impact. Policies productivity growth in China (DRC-World Bank, thus proliferate and are hardly ever discontinued. 2019). But the imbalance discriminates against In addition, public support seems to channel a potential “gazelle” firms, which need to scale up, significant part of funding to SOEs. In 2013, public undertake organizational and process innovations, funding represented 9 percent of total intramural and connect to a larger network of suppliers and R&D spending among SOEs, three times more customers.22 It also goes against the experience than among private enterprises (Liu et al, 2017). It of countries like Japan or South Korea, which suggests a strong preference for SOEs even though supplemented financing support for firms with innovative outputs among state firms seem to be programs focused on building of firm innovation generally lower than among private firms, examples capabilities (Lee, 2016 and Malerba, 2006). of several innovative SOEs notwithstanding The volume of the government’s support for (box 1).23 Innovation policy instruments also business R&D is close to the OECD median, but emphasize potential innovators over non-innovators, is much higher than what China’s level of income who do not appear to be the focus of any instruments, would suggest. Tax incentives represent about even though this group may face the greatest barriers half of the value of the total support (OECD, 2017 to innovation. “R&D Tax Incentive Indicators, March). There is Industry distribution of public support for innovation evidence that tax incentives promote private R&D in China is quite concentrated. Railways and transport spending in China: a 10% reduction in R&D user equipment, mostly high-speed train technology, costs tends to boost R&D spending by about 4 received more than one-fourth of total innovation percent in the near term. This is a large impact, but support in 2015. The top five recipients—railways and it nevertheless remains lower than in developed transport equipment, computers and electronic, general countries, likely reflecting a less developed purpose machinery, special purpose machinery, and institutional environment that dampens the size of electrical machinery—were responsible for almost the returns (Jia and Ma, 2017). The impact of tax two-thirds of total support (OECD, 2017). In most incentives is also higher among private firms than developed countries, public support for innovation among private firms with “political connections” across industries is spread more equally, suggesting a and among SOEs. The latter seem to invest in R&D larger role for market-driven mechanisms. regardless of tax incentives. This suggests that there is scope for further increasing the efficiency of tax support for private R&D by focusing it on the private sector. CHINA’S PROGRESS IN BUILDING INNOVATION CAPACITY 10 Box 1: Innovative SOEs in China Over the past decade, Chinese SOEs have developed multiple innovations, including breakthroughs in introducing new technologies and industry standards, building new markets, and developing new business models. Such breakthroughs have likely had substantial positive spillovers on the rest of the economy and helped support productivity growth in the private sector. The most innovative SOEs include CRRC, a producer of widely acclaimed high-speed trains; SGCC, manufacturer of ultra-high voltage transmission technology and smart grid innovation; COMAC, producer of new commercial aircraft, which promises to break the global duopoly of Airbus and Boeing; CASTC, developer of the Beidou Navigation Satellite System; Sugon/ Dawning, the leader in high speed computing; and ZTE, one of the emerging global leaders in communications, including 5G networking and optical access technology. ZTE spends 12.2 percent of its revenue on R&D, almost on par with China’s leading private firms such as Alibaba or Baidu, suggesting that public ownership does not always mean less innovation. Source: Authors’ own. Research Capacity Training in western and Japanese research institutions and the connections forged there have contributed Complementing the large and growing R&D greatly to the quality of China’s research capital.27 spending has been investment in training students China strongly encourages researchers to publish in science and engineering (S&E), who represented in refereed scientific publications, and urges around half of all bachelor degrees. In 2014, S&E researchers and firms to patent their findings. It is graduates comprised a group of almost 1.5 million now second only to the United States in number people—more than the combined total of degrees of publications. The quality of papers—initially in S&E awarded in the United States and the a weakness—is improving,28 apparent from the EU-8 (figure  7).24 In addition, more than 30,000 scoring of scientists by the Nature Index, which received doctorates in S&E disciplines, more than shows that the Weighted Fractional Index (WFC) in the United States and three times more than in for Chinese researchers has risen and only the India.25 The number of graduates and scientists United States continues to outrank China.29 Forty has since increased further, especially as the data Chinese institutions were among the top 100 do not include the thousands of Chinese students institutions worldwide that registered the greatest receiving advanced training abroad, at least half of increase in their WFC scores between 2012 and whom return to China attracted by the Thousand 2015. The United States was second, with 11 Talents and similar programs introduced in the (Nature, 2016). Chinese researchers are engaged early 2000s to encourage the return of members of in cutting-edge research on stem cells, genomics, China’s scientific diaspora.26 protein and green chemistry, optoelectronics, Thanks to the large annual inflow, the pool of supercomputing, machine learning, ballistic and researchers climbed from 443 per million people in anti-satellite weaponry, and other areas, but citation 1996 to 1,113 per million in 2014, still well below counts suggest that the strengths of Chinese 3,122 in the United States and 4,947 in Japan. researchers are greater in physical than biological However, China still has only two researchers per sciences. China’s progress is all the more impressive 1,000 employed, well below 10 in the EU and the given the state of the country’s teaching and United States, 14 in Korea, and almost 18 in Israel research institutions in the early 1980s, and some (OECD, 2017). continuing challenges.30 PROMOTING INNOVATION IN CHINA: LESSONS FROM INTERNATIONAL GOOD PRACTICE 11 Figure 7: Bachelor and PhD Degrees in Science and Engineering are Climbing Rapidly, 2000–14 Bachelor’s degree 1,600 1,400 1,200 1,000 Thousands 800 600 400 200 0 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 China United States EU-Top 8 Japan Rep. of Korea PhD 60 50 40 Thousands 30 20 10 0 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 United States EU-Top 8 China India Japan Rep. of Korea Note: The Top 8 EU total includes aggregated data for the eight EU countries with the highest number of S&E bachelor and doctoral degrees awarded in 2014. Source: Authors’ own based on the National Science Foundation Science and Engineering Indicators 2018.31 China’s progress in building innovation capacity 12 Improvements in human capital and rapid growth in Figure 8: Patent Applications in STI spending have led to a boom in invention and Percent of the World’s Total, 2018 utility-patent applications, especially after 2008. By 2011, China led the world in the number of domestic patents filed.32 In 2018, China’s National 50 46.4 45 Intellectual Property Administration received 1.54 40 million patent applications, which accounted for almost half of the world’s total and almost three 35 30 times more than the United States (figure 8). 25 The number of international patent submissions 20 18 has also increased, as reflected in the growing 14.7 15 number of patent submissions to the U.S. Patent 10 9.4 6.3 5.2 and Trademark Office (PTO) and the European 5 and Japanese patent offices. Patents granted to 0 foreign parties by the U.S. PTO increased from China US Japan South EPO Other Korea 111,823 in 2010 to 178,184 in 2018, with applicants from China receiving 2,655 in 2010 and 16,759 Source: Author’s own based on WIPO 2019. in 2018—a steep increase (USPTO, 2018). The Note: EPO is European Patent Office. quality of patents has also gone up.33 In utility patent applications, China is now the fifth most patent-intensive country. Chinese international result of interfirm or business–science collaboration. patent applications (PCT) counted by the World Most R&D projects reside within individual firms, Intellectual Property Organization (WIPO) have with little input from universities or other firms also surged: China has now overtaken Japan as the (OECD, 2017). Foreign-owned entities are a second-largest source country of patent applications small source of invention patents but receive more and is likely to overtake the United States in 2020 one-third of the patents granted (DRC-World or so.34 Bank, 2019). The quality of patents merits further improvement. China’s renewable energy industry leads other Despite the surge in patenting and heavy spending countries in patent applications but only 1.4 percent by enterprises, only a small proportion of Chinese are submitted to overseas patenting offices. The patents are registered abroad at the patent offices citation rate for China’s photovoltaic (PV) patents in the United States, EU, and Japan, implying their is a low 14.9 percent against more than 60 percent lower quality (Molnad, 2017). 96 percent of patents for PV patents received by firms from Germany, are filed only in China and two-thirds of China’s Japan, and the United States. Chinese firms are most research-intensive companies file no patents the leading global manufacturers of wind turbines abroad. Moreover, the number of triadic patent (with five firms in the top 10 in 2015).35 But they filings is miniscule (Boeing and Mueller, 2019). have received very few international product or Forward citations received by patents abroad and process patents for new technologies, indicating claims on Chinese-owned patents are fewer than that innovations were modest and incremental those of the leading industrial economies, as is the (World Bank, 2018). Uneven patent quality may percentage of national patent stock that represents be one reason there is so little impact of increased truly sophisticated technologies (Prud’homme and patenting on productivity (OECD, 2017). Low Song, 2016, Prud’homme and Zhang, 2017). Most patent quality can become self-reinforcing as firms patents are also concentrated in few cities, including attempt to use them to create barriers to entry Beijing, Shanghai and Shenzen (Branstetter, for competitors, especially start-ups—blocking Glennon and Jensen, 2018), and few patents are the freedom to operate and increasing transaction costs PROMOTING INNOVATION IN CHINA: LESSONS FROM INTERNATIONAL GOOD PRACTICE 13 associated with “identifying and invalidating low- patents, and among these firms only 5.8 percent quality patent rights, licensing them, or subjecting had multiple patents, representing a remarkable them to lawsuits” (Prud’homme and Zhang, 2017). 91 percent of all patents (Fang, He and Li, 2016). The computer industry had the highest number Even with these concerns, however, China’s of patents per firm, 20 times more than the food patents are making an important contribution to industry, while the medicine industry had the economic growth. According to one estimate, highest ratio of firms with patents. Younger mid- higher patent density is associated with an increase sized firms, larger firms, and firms with more in the productivity of firms (Zhao and Liu 2011; Li, R&D spending tend in general to patent more, and 2012). Another estimate shows that a one-standard- exporting firms patent more than nonexporting deviation increase in patent density is associated firms. The share of patents attributable to SOEs with a 6-percentage point higher city GDP growth has declined substantially since the mid-1990s in a subsequent period (Sun et al, 2017). The effect (figure 9, right panel), even though SOEs are more is particularly strong for inland areas, suggesting heavily subsidized than private and foreign firms. patenting that promotes productivity could, in This may indicate that SOEs are less efficient in principle, enable lagging inland regions to narrow using their R&D budgets, or that many SOEs are the gap with the more advanced coastal zones (Sun in industries with low R&D intensity (Wei, Xie et al., 2017; Xu et al., 2017). And while neither and Zhang, 2017; Boeing, Mueller, and Sandner SOE presence nor college student density appears 2016). to have a direct effect on city-level growth, cities with higher college density and a more supportive There are concerns about how China’s intellectual local government tend to produce more patents property rights (IPR) regime affects the volume per capita (Sun et al., 2017). Lower SOE presence and quality of patents, given evidence indicating appears to be associated with better performance on the significant positive effects of stronger IPR overall patent and trademark density, but a greater protection on innovation and corporate R&D in SOE share is positively and significantly related to China (Fang, Lerner, and Wu 2016; Lin, Lin, invention patent density. This topic deserves to be and Song 2010). Despite recent improvements validated by more research as it could help shape in IPR laws, policies, patent evaluation,38 quality regional policy. thresholds for utility patents,39 and enforcement, more progress is needed. MNCs and local firms Although China’s investments in innovation continue to complain about IPR breaches and the have intensified and patents have proliferated, difficulty of enforcing claims given the complexity TFP growth has trended down since the turn of the IP system, the pressure to transfer technology, of the century and stagnated since 2012 (Wu, the protection afforded to Chinese firms, and the 2016). This may stem in part from the dominance limited deterrence value of the statutory damages of utility models of questionable value and awarded (OECD, 2017). lower-quality patents not derived from formal employment contracts.36 Another concern is China’s first-to-file system, which does not that, even though firms rather than institutions recognize trademarks registered in other countries, or universities have become the most important encourages trademark “squatting” by local firms, source of patents in China over the last two which is damaging for foreign companies. Other decades (figure 9, left panel), most patents are problems arise from gaps in regulations (as produced by just a handful of companies. To in China’s import-export regulations, the law facilitate greater rates of technology transfer governing science and technological progress, and and adoption in industries, there is ample room regulations governing standards involving patents) for tightening university-industry-research and in other IP laws (China has, for example, yet to collaboration and firm-level collaboration.37 accede to the UN’s 1991 convention on protection of Only 9  percent of all Chinese firms had any new varieties of plants) (Prud’homme 2012; Zhang China’s progress in building innovation capacity 14 Figure 9: Firms Have Become the Most Important Source of Patents in China 100 100 90 90 80 80 70 70 60 60 Percent Percent 50 50 40 40 30 30 20 20 10 10 0 0 95 19 6 97 98 20 9 00 01 20 2 03 04 20 5 06 07 08 20 9 10 11 20 2 13 14 95 19 6 97 98 20 9 00 01 20 2 03 04 20 5 06 07 08 20 9 10 11 20 2 13 14 9 9 0 0 0 1 9 9 0 0 0 1 19 19 19 19 20 20 20 20 20 20 20 20 20 20 19 19 19 19 20 20 20 20 20 20 20 20 20 20 Institution University Firm Individual SOE Non-SOE Source: Authors’ own based on the National Bureau Statistics of China. et al. 2017). Agencies and courts administering IP broadband mobile communications, the next- continue to suffer from capability constraints, and generation Internet, and quantum computing. coordination across some central and state-level The R&D intensity and market competitiveness bodies could be improved. Two-thirds of Chinese of Huawei, ZTE, Alibaba, Baidu, and Tencent firms believe that patents do not provide enough have made China a dominant force in developing protection against IPR infringement (OECD, international standards for 4G mobile telephony. 2017). Smaller, private companies are often most Chinese companies are also among the global affected, because they tend to lack the power and leaders in the development of the 5G mobile expertise that large firms and MNCs possess for standard. China’s application software and support fighting violations. platforms have reached an internationally advanced level. Patents in electrical machinery, computers, Technology Upgrading and Diffusion and digital communications are among the top three technology fields for the Chinese patent office.41 Chinese firms have made major progress towards upgrading their technological capacity through As a result of this continued progress, the diffusion of technology from the rest of the world composition of Chinese exports has rapidly shifted and, more recently, through their own advances.40 toward more skill-intensive industries (Amiti and In manufacturing, Chinese firms have demonstrated Freund, 2010) and its export mix increasingly the ability to reverse engineer, adapt, and improve overlaps with that of developed countries, even foreign production technologies and to scale up if they are concentrated at the bottom of the their production at short notice. This narrowed price–quality range.42 These outcomes have China’s estimated hypothetical output per worker been supported, at least in part, by Chinese firms using U.S. technology and actual output from acquiring advanced technology and moving up 7.79 in 1979 to 3.86 in 2008 (Shen, Wang, and the product value chain. In China, technological Whalley 2016). Chinese companies have made catch-up has been more prevalent in locations significant breakthroughs in technological research with greater industrial diversity (Zhang, 2015), in and industrial development of high-performance sectors with lower prevalence of SOEs (Guan and computers, Internet applications, new generation Pang, 2017), and among exporters and foreign- PROMOTING INNOVATION IN CHINA: LESSONS FROM INTERNATIONAL GOOD PRACTICE 15 owned firms (Olabisi, 2017). Firm performance China’s Indicator—Based has also been aided by the internet, even in the Innovation Rankings pre-Alibaba era, facilitating communication with buyers and input suppliers (Fernandes et China has reached—or is approaching—the al., 2017). global technology frontier in some areas. But its comparative advantage still resides mainly in low- Access to foreign inputs—a significant driver of cost and large-scale manufacturing. A wide gap productivity growth in many countries—has been in core technical capacity remains with the major important in facilitating technology adoption at developed countries. firm level. Lowered import tariffs in China made higher-quality imported intermediates attractive for Several innovation indexes that rank countries more firms, raised domestic product quality, and suggest that China is moving up the ladder, but stimulated incremental innovation (Fan, Li, and still has some way to go. Increased spending has Yeaple 2015).43 Indeed, imports of capital goods are strengthened STI capacity, but acquiring innovation a known channel for foreign technology transfers, capability comparable to some of China’s neighbors R&D spillovers, and product upgrading (Eaton and and western countries could take longer. Kortum, 2001). China’s performance varies widely on innovation Despite rapid progress in developing in-house rankings: The Bloomberg index of innovative technological capacity, industries such as economies ranks China in 15th place, the top place integrated circuits, high-precision sensors, basic for a non-high-income economy. China scores software, large-scale industrial software, and highly on patent activity (2nd) and tertiary efficiency biotechnology continue to require high levels (5th), while it lags behind in productivity (47th) of integration with producers elsewhere in the and researcher concentration (39th).44 The WIPO/ world. And foreign producers have 80 percent INSEAD/Cornell Global Innovation Index 2019 of the market for industrial robots, advanced shows that China moved from 17th to 14th place automatic control systems, computer numerical worldwide and is now the only upper-middle-income control machine tools, and computer numerical country in the global top 30 (WIPO/INSEAD/ control systems. Chinese companies also depend Cornell, 2019). Other rankings show similar results. on foreign suppliers for three core components— On balance, these assessments reinforce the controllers, servomotors, and reducers (Peng and perception that, except in a few sectors such as the Dongyu, 2015). Of four major technologies— digital industry, China’s innovation capabilities and computation, visualization, voice recognition, underlying learning and creative culture may yet and driving systems—Chinese enterprises lead take some time to mature fully.45 The differences only in voice recognition. Key targets for drug across indexes are also driven by varying R&D, cell lines for vaccine antibody production, methodologies, choice of indicator, and areas of and critical genes for animal and plant breeding focus. Some emphasize quantitative indicators and are largely monopolized by developed countries. areas where China has been a strong performer. In the development of new materials, China’s Others give greater weight to governance, financial, overall technical level in basic raw materials is political, business, and regulatory environments not high. And its consumption of material and and to ecological sustainability—areas where China energy per unit of GDP is well above the average performs less well than comparator countries. Still for industrialized countries, in part because it others downplay China’s strengths by ignoring lacks process technology know-how and is weak factors important for innovation, such as availability in supporting and engineering capabilities. of venture capital, strength of manufacturing, penetration of consumer ICT technologies, and size of the domestic market, where China has an advantage over countries ranked higher. China’s progress in building innovation capacity 16 FINANCE, COMPETITIVENESS & INNOVATION INSIGHT | FIRMS, ENTREPRENEURSHIP & INNOVATION INNOVATION REFORM AGENDA C hina seems to have already achieved tangible success in supporting innovation, as reflected in high R&D spending, development of cutting-edge technologies in several areas, and trailblazing development of digital economy. However, to become a global innovation champion, China will need to continue to improve its innovation capability in at least six other areas: enhancing government capabilities, balancing national and regional support for innovation, investing in basic research, leveraging open innovation, facilitating technology diffusion, and improving managerial practices, now discussed in detail. Enhancing Government Capabilities inefficient, leading to little alignment between the policy objectives, budget allocations, and final Public institutions that design, implement, and results. When the World Bank and Colombia, for evaluate public innovation-support policies can reduce instance, jointly reviewed the country’s innovation market failures and lower coordination failures. Yet policies, they found that the government’s objective the importance of such institutional capacity is often to leverage innovation to diversify the economy underestimated. from natural resources was not reflected in increased The capacity of innovation-supporting institutions budget allocations for diversification. In 2014, can be assessed across four dimensions (Cirera spending on diversification came to less than 10 and Maloney 2017, p. XXII):research that leads to percent of total innovation expenditures (figure 10, scientific discovery. left panel). In a similar vein in Poland, there was little alignment between the government’s intention • Policy design, including the ability to identify to support R&D spending in SMEs, where market market failures and key measures needed to failure is the largest, and the fact that more than address them. 40 percent of spending under one of the flagship innovation support programs ended up financing • Implementation quality, including the availability capital investment in large enterprises, where of robust M&E systems that can prove the “value market failure hardly existed (figure 10, right panel). for money” of public investment and of well- Most other countries face similar challenges. aligned performance incentives. Steps to avoid fragmentation, duplication, and • Policy coherence, especially the links between inefficiency would have a high payoff—given government strategies, policies, and instruments, the plethora of policies and initiatives supporting and the quality of coordination among the innovation by different levels of government key institutions. in China (more than 170 innovation policies at • Policy consistency and predictability, ensuring national level alone, as noted), and given that that public support policies do not change with China will continue to get closer to the technology each new government. frontier. One approach would be to empower an agency with oversight and decision-making In practice, many countries fail to build innovation- authority to serve as a national clearinghouse for support systems with enough institutional capacity innovation programs and projects above a certain to meet the government’s objectives. Public support amount of funding. This would be in line with policies are often fragmented, duplicative, and good international practice, with most developed PROMOTING INNOVATION IN CHINA: LESSONS FROM INTERNATIONAL GOOD PRACTICE 17 Figure 10: A Lack of Alignment Between Program Objectives and Budgets Hampers Outcomes Colombia: Structure of innovation spending, Poland: Public support for innovation under an EU-financed 2010–14, by type flagship program, by intervention type and firm size 1,400 100 90 1,200 137.5 80 Billions of Local Currency 1,000 132 70 800 487.5 90 60 153 363 Percent 50 600 161 40 400 586.5 666 30 336 625 605 20 200 203 10 110.5 144 0 0 2010 2011 2012 2013 2014 Private Capital R&D Investment Diversification of existing business and creation of new ones Innovation infrastructure and ecosystem Micro Small Medium Big Improving existing business Sources: Cirera and Maloney 2017. Cirera and Maloney 2017 and Kapil et al. 2012. countries concentrating their innovation policies responsibilities, but the bulk of the research effort is in a few institutions or coordinating the policies still managed by few agencies. Even in the United closely (Paic and Viros, 2019). In Korea, while States, just three institutions—the Department of many ministries are involved in innovation policy, Defense (DoD), National Institutes of Health (NIH), the Ministry of Science, ICT, and Future Planning and Department of Energy (DoE)—are responsible and the Ministry of Trade, Industry, and Energy for more than two-thirds of federal R&D spending. are responsible for around two-thirds of total R&D And the five top institutions are responsible for spending.46 In addition, the National Science and almost all spending. Similarly, support for basic Technology Commission, headed by the prime research (discussed later), is largely garnered by minister, coordinates national innovation policies; only a few agencies.48 it also helps analyze impacts of national R&D One way to streamline policies in China would programs and instruments to adjust R&D budgets be to follow most developed countries and set up (OECD, 2014). Support for enterprise innovation a new innovation agency, or upgrade an existing is largely concentrated in the Korean Institute for institution to focus on high-value-added support for Advanced Technology, established in 2009 after a enterprise R&D. While there is no single blueprint merger of six public institutes.47 for how an innovation agency ought to look—much Other developed countries have made similar depends on the circumstances of each country— attempts to streamline their innovation policies at most innovation agencies benefit from the ability to the national level, while decentralizing them to local concentrate top-notch know-how, human resources, levels (OECD, 2015). Finland, Ireland, Singapore, and advanced financial instruments that respond to and Israel concentrate decision-making and R&D the increasingly sophisticated innovation needs funding in only a few agencies. True, these are of enterprises (Glennie and Bound 2016; Kapil small economies and in larger countries there is and Aridi 2017). They also reduce red tape and greater dispersion of policymaking and allocative information and coordination failures by becoming INNOVATION REFORM AGENDA 18 one-stop shops for funding and for support to (Glennie and Bound 2016, p. 69). In Switzerland the enterprise sector. As research and innovative and Finland, innovation agencies decide on budget activities become more global, innovation agencies allocations for innovation support programs and are also natural partners for supporting international projects based mainly on the recommendations of cooperation. And such agencies can deal with the panels of experts from industry, academia, and civil “indivisibility” of R&D spending, which demands society, rather than the government. Innovation substantial investment outlays for selected R&D agencies also spend significant budgets with no projects to reach required scale. Given the already prior sectoral allocations. For instance, TEKES, significant institutional complexity in China, Finland’s leading innovation agency, spends around upgrading an existing institution to become a one- 40 percent of its annual budget on “responsive” stop-shop for enterprise innovation would likely investments—projects that do not fit any specific be the most beneficial option. government strategy. Israel’s Office of the Chief Scientist even goes a step further: regardless of Strategic, regulatory, and analytical aspects of sectoral provenance it bases its support for all projects innovation policy could remain vested among the solely on the capacity to generate new technologies existing institutions and coordinated within groups with substantial growth and export potential. and committees headed by the premier or a vice premier, such as the Leading Group of Science and Many larger countries also have specific arrangements Education of the State Council.49 The coordination for national–regional coordination. Spain, for mechanisms between all existing entities should instance, relies on collaboration agreements between also be strengthened, based on international good state and regional governments, which align the practice (OECD, 2014b). Many countries, such parties along the key innovation priorities, provide as Australia, Japan, and Korea, use roundtables targets for innovation policy, and coordinate budget or policy councils to coordinate innovation policy allocations. Most developed countries also use horizontally and vertically. Japan and Korea other forms of coordination, most often through use high-level policy councils to take strategic national strategies, a dedicated agency or ministry, decisions; in other countries, the policy council and policy evaluations—or a combination of all plays more of an advisory role. three (figure 11). China is well placed to complement its traditional Strengthening policy evaluations and reviews approach of targeting strategic industries with of innovation policy in China is another way to horizontal reforms. Policies to promote new enhance coordination and improve the quality and technologies (including R&D tax credits) should be impact of innovation policies. Policy evaluations open to all sectors, regardless of their “strategic” in particular should be based on rigorous M&E value. Because tax credits help mostly large and frameworks. The lack of such frameworks is one established companies that have enough taxable of the reasons for the misalignment between public income and administrative capacity to apply for policy and its results. credits, support programs for start-ups and SMEs Many countries, including China, have yet to develop should include other instruments such as grants, these frameworks, which can provide ongoing vouchers, and collaborative networks. feedback on the performance of the innovation This is the case in most developed countries, where support system and assess, to the extent possible, the flagship innovation agencies have considerable return on public investment. Experience gained by discretion over the focus of their support. In the United States in implementing the Government Sweden, 80 percent of spending by VINNOVA, Performance and Results Act of 1993 indicates that the country’s innovation agency, is administered research is inherently difficult to evaluate because its by the agency. The government decides only on purpose is to enlarge knowledge and understanding the remaining 20 percent, but even there it uses of a subject. Practical outcomes need not materialize “brushstrokes rather than detailed instructions”. for years, and some research may show that certain PROMOTING INNOVATION IN CHINA: LESSONS FROM INTERNATIONAL GOOD PRACTICE 19 Figure 11: Priority Forms of National Innovation Policy Coordination, 2012 National Strategies and Visions Dedicated Innovation Agency or Ministry Policy Evaluations and Reviews High-level Policy Council Informal Channels of Communication Inter-agency Joint Programming Intervention of the President’s or Prime Minister’s Office Job Circulation of Civil Servants, Experts and Stakeholders 0 1 2 3 4 5 6 7 Source: OECD Science Technology and Industry Outlook 2012. Note: Based on own-country ratings, where 7 = high importance, 1 = low importance and 0 = nonexistent. devices or processes being sought cannot actually There is a need also to incorporate explicitly the be produced—which, while negative, remains a development of human resources as an objective of valuable result. Government agencies in developed research and to continuously refine methodologies countries therefore rely on expert review to evaluate through the process of learning. And for the research while it is in progress, with respect to evaluation to improve decision-making, the findings quality, relevance, and whether the research is at need to be incorporated into the ways agencies plan the international technological frontier (when such their research activities, share information across criteria can be applied appropriately). agencies, and coordinate their activities to steadily enhance research productivity. A RAND Corporation review of several OECD country research frameworks and tools concluded that Evaluation based on rigorous methodologies has research evaluations are generally conducted under four proven helpful in measuring the impact of innovation rationales: “(1) Advocacy: to enhance understanding policies and their additionality (that is, the extent to of research and its processes among policymakers and which their impacts go beyond what would have the public; (2) Accountability: to show that money occurred without them). The United Kingdom was has been used efficiently and effectively, and hold the first to begin assessing research by universities researchers to account; (3) Analysis and Learning: through its Research Assessment Exercise, now the to build an evidence base on research effectiveness Research Excellence Framework, which evaluates and relevant support mechanisms to inform policy research by all 165 higher education providers in decisions; (4) Allocation: to determine where best to the country along three criteria: quality, impact, and allocate funds in the future to get value for money”. vitality of research environments, with the help of (Guthrie et al., 2013). peer reviews. The STAR Metrics framework50 in the United States has the tools and a repository for data Evaluation can be improved by delineating its on research projects that enable the NIH and the purpose and by establishing transparent procedures NSF, with assistance from the Office of Science and and rationales to verify the results obtained, Technology Policy, to weigh the impact of research including longer-term impacts with the help of while minimizing the burden of data collection. A longitudinal studies. The timing of evaluation framework developed by the Canadian Academy needs to be coordinated with performance reports. of Health Sciences is another example of how INNOVATION REFORM AGENDA 20 governments can track and measure the benefits points”. (Bruhn and McKenzie, 2016, p. 2). The from research (CAHS, 2011). support program also enhanced science–industry collaboration, boosted patent applications, and Other frameworks are in use, too—notably the increased the number of research publications public expenditure reviews for innovation, jointly —and helped commercialize project-related implemented by the World Bank and client inventions (ibid.). governments in middle and high-income countries (box 2.3). A recent study on Poland, which used a China could productively adopt an M&E framework regression discontinuity methodology to assess based on the experience of Canada, the United the impacts of the country’s program to support Kingdom, the United States, and several other innovation activities, found that it “largely funds developed countries. It could become part of a larger projects that would not otherwise get funded effort to monitor systematically the efficiency and by other agencies or by the business–science effectiveness of innovation policy,51 supplemented consortia themselves, increasing the probability of by public expenditure reviews of innovation policy a project being completed by almost 60 percentage (see box 2). Box 2: Public Expenditure Reviews of Innovation Policy Public expenditure reviews (PERs) can help countries improve the quality of their innovation policymaking, improve resource allocation across programs, achieve budget savings, and enhance program impact. PERs evaluate four stages of innovation policy interventions: • General evaluation of the quality and coherence of the policy mix based on the conditions of the country and its innovation system, including the portfolio mapping of STI programs and their assessment based on coherence with existing innovation policy objectives. • Evaluation of the quality of design, implementation, and governance (functional analysis) of existing instruments based on good practices. • Evaluation of the efficiency of existing instruments, meaning their ability to produce the expected outputs with reasonable levels of resources. • Evaluation of the effectiveness of existing instruments and the system, by analyzing their ability to generate the desired impact. Unlike traditional PERs, innovation PERs have the individual innovation policy instrument as the unit of analysis, which allows evaluators to identify what is spent with what objectives, and thus efficiency and effectiveness at a more detailed level. Their main objectives are to: • Support the process of redesigning and shaping public innovation policies by using data and information on existing instruments. • Improve the ability of governments to coordinate innovation policies by evaluating the design and implementation process and assessing the quality of the M&E system. • Support the adoption of good practices in design, implementation, and coordination of innovation policy instruments by benchmarking instruments across countries (where appropriate). • Formulate policy recommendations to strengthen the innovation policy mix by eliminating redundancies and leveraging complementarities across the portfolio of instruments. Source: Cirera and Maloney 2017. PROMOTING INNOVATION IN CHINA: LESSONS FROM INTERNATIONAL GOOD PRACTICE 21 Balancing National and Regional their close geographic proximity to universities Support for Innovations and research institutions.55 These agglomerations benefit from urbanization economies56 and a variety As China moves up the technological ladder and of service providers that are important innovators productivity displaces capital investment as the contributing an increasing share of the value added primary driver of growth, a national innovation by manufacturers.57 strategy can provide the ground rules and a large part of the financing for innovation. But given Many and varied are the makings of an RIS. A start- the uneven distribution of innovation across the up that grows into a major firm and that creates a country, every region will need to complement cluster of suppliers and competitors can launch a national initiatives by crafting its own policies to regional system, as can government policy that foster innovation—that is, to develop or deepen a locates research and manufacturing activities in a regional innovation system (RIS).52 particular area. The defense industry in the vicinity of Chengdu is at the core of an RIS; Shenyang serves With so many administratively delineated regions as the axis of another. An RIS begins acquiring in China trying to create functioning RISs, there is traction once key firms that compete on the basis an urgent need for a precise characterization of what of innovation start scaling up, generate spillovers, an RIS is, the enabling conditions that facilitate its and crowd in other activities. This is more likely formation, and policies that can stimulate innovation in a large, diversified industrial centers and, as by firms and their supporting institutions. Shenzhen and other southern special economic zones show, a strategically located technology zone An RIS, as a subset of the NIS, can be described as a can become the hub of an RIS. “production structure embedded in an institutional structure in which firms and other organizations European and American RISs also identify research are systematically engaged in interactive learning”, universities as building blocks because they are (Doloreux and Partob, 2005) or as “a place the principal sources of human capital and conduct where innovation activities are concentrated and much of the upstream research.58 Most effective networked, creating a context-specific environment are universities and other research institutions that fosters the production of knowledge with integrated into a “triple helix” that ties together a systemic configuration”, (Prodi, Frattini, and firms, universities, and businesses engaged in Nicolli, 2016) or as “sticky place in slippery space” manufacturing and allied producer services that nurtures, attracts, and retains firms that innovate (Etzkowitz, 2003). An interactive helical structure and grow (Markusen, 1996).53 Almost invariably— lends added impetus to regional innovation. Such as in China—an RIS is anchored in major urban a structure has emerged in Beijing, Shanghai, and centers that may also host special economic or Shenzhen, and policy incentives nudge firms in technology zones.54 Cities with colocated—and other regions to link with universities, though this possibly interlinked—clusters are advantaged. So remains a work in progress. are cities that house clusters at different stages of the life cycle, allowing new clusters to succeed A second factor that affects the dynamism of an mature or declining ones and take advantage of RIS is foreign direct investment (FDI), which in their learning and workforces. China has introduced new technologies through vertical spillovers, helped build ecosystems of local The more advanced RISs in China began to suppliers, accelerated accumulation of intangible germinate in special economic zones such as capital that undergirds innovation, and enabled Shenzhen and in “open cities” such as Shanghai China to become a leading exporter of medium- and and Beijing. Large urban agglomerations have high-tech manufactures (Enright, 2017). Regional proven especially conducive to forming industrial innovation has also benefited from policy initiatives clusters and ecosystems, sometimes a fertile source that improve the business environment by pruning of commercially viable innovation because of regulations, providing stronger IP protection, using INNOVATION REFORM AGENDA 22 public procurement to launch new products and searchable databases, national and international services, and setting standards that build product market intelligence, measuring and testing quality, reliability, marketability, and brand image. equipment, and a means for entering joint purchasing partnerships that reduce their costs. As the pace of innovation activity quickens in Standards for products and their certification can China, an increasing share of overall spending also help firms—especially smaller ones—meet will be subnational, with provincial and municipal quality requirements and thereby lower market authorities playing a larger role. In the recent past, entry barriers (OECD, 2008). Many of China’s there has been a marked tendency for provincial budding “gazelles” lack the capacity to produce authorities to take their cues from the central high-quality products and achieve large-scale government and to set their priorities accordingly. distribution, particularly in markets abroad. This Regional research thus tends to be duplicative, is where public–private partnerships (PPPs) and wasting resources and human capital. This needs collaboration across regions would be valuable. to change. RISs in China now tend toward insularity, and Regional innovation in China is more likely to make nationally mandated efforts to localize technologies the best use of research financing and deliver growth if could make RISs even more inward looking. A each regional entity takes the broad central government more open innovation system would more likely directives as the point of departure, and tailors its lead to smart specialization of R&D. Chinese cities innovation strategy to its existing urban industrial base, that have been more open to FDI, promoted joint research capabilities, and entrepreneurial potential.59 ventures, and encouraged learning by exporting Not every region can engage in cutting-edge research, have proven more innovative. They have used however, and not every region needs to focus on MNC-linked GVCs to acquire and improve new the most advanced technologies. Specialization and technologies, scale up production, and enlarge their collaboration with researchers in other regions and global footprint. The lesson for RISs in the making abroad can make research more productive and lead is clear. Specialize in the region’s strengths; use to innovations that local firms can commercialize. PPPs to enhance entrepreneurial activity; coordinate The trick is to recombine available resources in new research activities; and keep the system open so that ways. Smart specialization, implemented regionally local researchers benefit from research elsewhere in the EU, could be a source of inspiration for Chinese and collaborate with researchers worldwide. regions and help them identify regional endogenous strengths around which to build their support (World Bank, 2016). Investing in Basic Research Better coordinating private entrepreneurship and Governments and the research community know public research could yield more usable patents and that large private firms—with the time horizon and profitable innovations. An organization comparable the resources to conduct basic research—have been to the Georgia Research Alliance60 would enhance the cutting back on their in-house activities and looking productivity of subnational research by coordinating to publicly supported universities, research institutes, the activities of public and private entities. And it and start-ups to close the gap (Arora, Belenzon and could increase the contribution of universities to Patacconi, 2015). Indeed, much of the EU’s research applied research and product development, directly to expand the frontier of discovery now takes place at assisting firms in commercializing their ideas (Chen universities and in public laboratories.61 But research and Guan, 2011). can generate large social returns.62 Basic and applied research are intertwined, so pushing hard on applied What else can translate research findings into research to the neglect of basic research—as many products and services? Local support services, firms and universities are now prone to do in pursuit sponsored and in part financed by provincial and of near-term commercial payoffs—can lead to municipal authorities, can provide firms with diminishing returns.63 PROMOTING INNOVATION IN CHINA: LESSONS FROM INTERNATIONAL GOOD PRACTICE 23 Some argue that the most productive years of 7 percent of the R&D spending (OECD, 2017). Bell Labs were when R&D was conducted under Most of the growth in R&D spending is from the same roof by a critical mass of exceptional “experiment and development” spending by large researchers with considerable leakage between the firms that typically do very little basic research two sides of R&D and with researchers “having (figure  2.27, left panel). At the same time, the the flexibility to move between research domains United States allocates 17.6 percent of its R&D to [and] between R&D when it was productive to do basic research, Japan 12.6 percent, and France 24.1 so”. (Narayanamurti and Odumosu 2016, pp. 46- percent. OECD countries spent about 17 percent of 47; Gertner, 2012). Researchers also enjoyed the total R&D outlays on basic research, 21 percent on freedom to pursue projects of their own choosing applied research, and 62 percent on experimental because the monopoly enjoyed by the parent development. China’s spending on basic research company AT&T meant that there was little pressure was much lower (figure 12). to restrict research to projects that promised That basic research is beginning to receive more a commercial return. The DARPA model also attention and funding in China is apparent with promotes parallel advances in pure science and in several research facilities coming online. The technology development by identifying a high-risk/ synchrotron light source in Shanghai is enabling high-payoff objective, as in information technology, researchers to better understand the structure of sensor systems, cyber security, robotics, new proteins and cell metabolism (Nature, 2017a, materials, and directed energy systems.64 2017b). The Spallation Neutron Source and China’s Basic research is essential in, for example, furthering Advanced Research Reactor can further basic the development of quantum computing and research in important areas. China is rightly moving communications in enhancing cybersecurity,65 in at a measured pace in scaling up basic research as it fighting infectious diseases increasingly resistant to builds human research capital of the highest caliber. the most potent antibiotics, in finding new catalytic Investment in expensive equipment that sits idle is materials (possibly using protein-derived enzymes) best avoided. And while the social returns to R&D for improving industrial catalytic processes,66 remain high, the “bang for the buck” from research and in discovering new and commercially viable is on a downward slope, with research productivity battery technologies (such as replacing lithium declining 4–6 percent a year (Bloom et al. 2017). with sulfur).67 Spillovers from the Diamond Light Consider the EU’s Horizon 2020, a seven-year Source—a third-generation synchrotron in the €75 billion program managed by the European United Kingdom—increased the productivity of Commission to enhance innovation in the EU and scientists (measured by publications) within a 25 develop research capital. The research activities km radius of the facility by 11 percent over 10 years have the material support of the private sector and (Helmers, Christian and Henry Overman, 2017). engage researchers from 130 countries, building an Research published in Science provides further extra-European network that taps expertise from empirical support for the links between patenting, across the world. Horizon 2020 has supported innovation, and earlier scientific discoveries 17 Nobel prize winners and the publication of (Ahmadpoor and Jones, 2017). numerous scientific papers, patent applications, China has 1,000 research institutes and as many and key research findings, such as the discovery of as 1,000 universities that can, in principle, new exoplanets, the Higgs boson, and gravitational conduct basic research (Yang, 2016). Yet, China’s waves. Because it restrains red tape and transaction contributions to original, frontier-pushing innovation costs, it caters to SMEs, which consume a quarter of to build the base of scientific discovery received the funding. A recent internal evaluation established only about 5.6 percent of total spending on R&D in that 83 percent of the projects financed would not 2018.68 Research universities, which could do more have gotten off the ground without funding from of the groundbreaking research, accounted for just Horizon 2020. It is also estimated that each €1 from INNOVATION REFORM AGENDA 24 Horizon 2020 generated €6–8.5 of GDP and that the Figure 12. China’s R&D Spending for benefits accrued to the EU could amount to €600 Basic Research vs. the OECD Average billion by 2030 (European Commission, 2017b). China could also review this experience and that 100 of other large countries such as the United States, 90 which finances basic research mainly through the 80 NIH and NSF, responsible for more than 60 percent 70 62 of annual spending on basic research.69 While the 60 84 NIH has a focus on biomedical research, the NSF 50 supports all types of research involving science 40 and engineering. Both institutions independently determine which areas of science are the most 30 21 promising and would provide the largest economic 20 payoff. The NSF identifies research projects 10 17 11 through a bottom-up approach, based on a constant 5 0 OECD China monitoring of the changing research trends in the Basic Research Applied Research United States and in the world. Project selection Experimental Development draws on merit review, considered the global “gold standard” for scientific review.70 Source: Colin 2017 and OECD 2017. Leveraging Open Innovation “hottest” technological developments. The reasons? Even in today’s world, the most advanced scientific Many innovative ideas are not the result of a knowledge diffuses slowly, and a great deal of it can given firm’s own efforts but instead come from remain tacit, diffusing by word of mouth (Adams, outside: from customers, partner firms, suppliers, Clemmons, and Stephan 2006). universities, and others. Almost any complex product today incorporates inputs from researchers With the advantages of openness more widely and firms in different economies—the iPhone is appreciated, R&D in advanced countries has tended one frequently cited example, with firms from toward greater cross-border collaboration (Hale, China, France, Japan, Korea, Taiwan China, and 2012). Close to a fifth of spending on research by the United States producing key components national governments is on international projects assembled in China by Foxconn (Gould and Villas- and in some instances it can be more than one- Boas, 2016).71 ICT advances have made it far easier half (Wagner and Jonkers, 2017). International to network, collaborate, and crowdsource, with the collaboration enhances the citation impact of emergence of pockets of specialized scientific and research, because researchers of the highest caliber engineering expertise in developing economies that and productivity are thinly distributed and among can now be drawn into collaborative ventures or the ones more likely to collaborate with the best tapped through crowdsourcing channels. researchers in other countries (Adams, 2013). Two- thirds of the scientific output of Switzerland— By dispersing R&D activities across countries, consistently ranked among the world’s most MNCs can shave costs, achieve around-the-clock innovative and competitive economies by the operations, and harness expertise that may be World Economic Forum—involves collaboration scarce in their home countries. For companies— with outsiders. Open economies produce the most or countries—that want to stay at the frontier of high-impact research, while in less open ones, scientific research, collaborative research and including several large emerging economies, almost interaction among researchers is the only sure three-quarters of research tends to be domestic, way to stay abreast of the latest thinking and the generating much less impact (Fleming, 2017). PROMOTING INNOVATION IN CHINA: LESSONS FROM INTERNATIONAL GOOD PRACTICE 25 China has been one of the foremost beneficiaries of and taking steps toward becoming technologically the open and global innovation system, but Chinese autonomous. These two goals may become firms are now viewed as competitors by their increasingly incompatible, especially as China is counterparts in high-income countries. Chinese no longer perceived as a “developing country” and researchers trained in the United States, Europe, and as many of its industries, companies and regions Japan have been conduits for knowledge transfers, (especially at the coast) are seen (and acknowledged and collaboration with foreign scientists has helped by the Chinese authorities themselves) to be closing Chinese scientists climb the citation indexes and in on the global technology frontier. By dismantling narrow technology gaps. China has been strikingly the barriers that shelter domestic firms from foreign advantaged by co-ethnic cooperation between the multinationals (a protection that is rapidly becoming almost 100,000 holders of doctoral degrees in redundant if not counterproductive), China would STEM disciplines in the United States of Chinese in one stroke stimulate the capacity for innovation origin and scientists based in China (Gupta, 2016). of Chinese enterprises, increase the level of productivity by enhancing market competition and Collaboration is by no means a one-way street. eliminate a major source of friction with its trading With Chinese researchers gaining experience, the partners. Developments during the past three years quality of equipment in Chinese labs is improving indicate that China’s partial globalization has markedly. And with research funding becoming outlived its utility. An industrially mature China can more plentiful, U.S. researchers, especially confidently embrace globalization in its entirety by younger ones struggling to mobilize grant funding, matching the openness of the advanced economies are collaborating with Chinese scientists to further with which it is now drawing abreast. their own research (Gupta, 2016). By opening satellite campuses in China, western universities What can China do to keep the global innovation such as NYU, Monash, Liverpool, and Duke72 system open and continue to benefit from the global have facilitated even more collaboration. And technological progress? While some developments the connectedness of China with its international are driven by exogenous factors that China cannot diaspora is tight: Chinese scientists are linked to control, there is much that China could do on others in 94 countries (Nature, 2015). its own. In the future, an open global innovation system First, it would be in China’s interest to continue will remain critical to sustaining China’s economic to strengthen international norms that govern growth and helping to meet the country’s two international economic relationships and promote “centenary goals”. Given the diminishing returns dialogue. China could help these efforts by fully of additional public investment and gradual aligning itself with WTO rules, including those decline in the size of the labor force, productivity on industrial subsidies, treatment of SOEs, and growth will need to become the main engine of regulations that give rise to non-tariff barriers to China’s growth. Faster productivity growth will in trade and strengthening enforcement mechanisms turn need to be supported by an open innovation (Grossman, McCalman, and Staiger, 2019). Policies system that promotes the absorption, diffusion and that may be in conflict with WTO SCM Agreement development of new technologies. or otherwise against the principle of competitive neutrality should be reviewed, modified or phased But in many ways China seems to have come to a out as appropriate. China could also spearhead crossroads in its development based on absorbing efforts to further develop international norms foreign technology and participating in the open governing cyberactivity. global innovation system. It could be argued that under the current, open system China has been Second, China could “lead the world by example” able to gain from the open global innovation by fully opening its markets, ensuring a level- system, while keeping its market largely protected, playing field for all companies, including for private progressing with indigenization of technologies and foreign-owned companies, strengthening INNOVATION REFORM AGENDA 26 effective IP protection, and making public support Facilitating Technology Diffusion policies transparent. All these measures would not only help reinvigorate the growth of the domestic Adapting to constantly changing technology economy, including that of the private sector, but and preparing for growing disruption requires a at the same time also lessen other nations’ fears of combination of policies aimed at technology transfer, an “unequal playing field” and thus help keep the science–industry collaboration, and R&D- and non- global innovation system open. R&D-based innovation. It also requires the skills and managerial competencies to maximize firms’ Third, China could further promote global cooperation capacity to innovate and absorb technology. The on science and innovation. This could include majority of potential productivity gains in developed mechanisms to increase investment in international countries (55 percent) appear to be derived from collaboration in basic research on global public adopting best practices and technologies rather goods (such a green technologies, medicines and than by developing new ones (Dobbs, Manyika, others), funded by, for instance, a newly established and Woetzel, 2015). Mechanisms to support the global innovation fund. Another fund could help diffusion of new and existing technologies range finance global scientist exchange programs modeled from broad management extension services to more on, for instance, the European Union’s successful sophisticated R&D institutions, and more recently to Erasmus student exchange program. technology information and collaboration platforms Finally, China might reconsider its indigenizing (Youtie and Shapira, 2017). policies to capture as much of the manufacturing The broader management and dedicated field value chain as possible. Some technological extension services, such as the Manufacturing autonomy is desirable for a major economic power, Extension Partnership in the United States (see as with technology development catering to major below), tend to follow a model similar to traditional infrastructure projects heavily oriented to a large agricultural extension services, and support building domestic market, such as high-speed rail (Liu and the capacity of SMEs to absorb existing technologies. Cheng, 2011). But much like import substitution One step above, at least on technological focus, carried to an extreme, it can rapidly become are technology extension services that concentrate counterproductive, a waste of resources spent in more narrowly on technology adoption, for rediscovering known technologies, and break down example of specific digital technologies, and not trust and collaboration with other countries— on broad absorptive capacity. For more specialized aside from potentially triggering trade disputes.73 technologies, OECD countries have technology While in a static view the advantages for China of and R&D centers. Some technology centers, such mastering all technologies and promoting global as those in Catalonia, Spain, arose from PPPs to technological champions may promise larger solve technological problems and develop solutions benefits than those proffered by an open global for specific sectors or clusters. The centers tend to innovation network, in a dynamic view where other be very specialized in specific technologies and countries react to such an indigenization strategy, are managed as a network, as with the Fraunhofer the apparent benefits may be outweighed by the Institutes in Germany, a network of 60 private losses resulting from the reactions of other players, nonprofit research institutes that do contract especially the leaders of the global innovation research for the government and for business network. The win-win solution could be to engage organizations (Shapira, Kwon and Youtie, 2017). other countries in China’s efforts to develop new But the capabilities of the Fraunhofer Institutes technologies, including core technologies, and are inseparable from the strength of Germany’s ensure that other countries do not see China as a engineering sector. Without strong capabilities in country that aims to monopolize the commanding the private sector, implementing these models in heights of global innovation and exclude other other contexts may be difficult. nations in the process. PROMOTING INNOVATION IN CHINA: LESSONS FROM INTERNATIONAL GOOD PRACTICE 27 Empirical evidence on the effectiveness of the various provide leading examples of creating integrated mechanisms is scarce, however. There are just too approaches to deploying policy instruments (BMBF few evaluations, and those that exist have focused 2017; GTAI 2017). on a few policy instruments, such as some business The United States, for example, launched the advisory or R&D schemes, rather than broader Advanced Manufacturing Partnership 2.0 (AMP 2.0) evaluations. A few successful experiences have been in 2011. This is a “national effort bringing together documented in East Asia—Japan, Korea, Singapore, industry, universities, and the federal government” Taiwan China, and China to certain extent—where a (White House 2011). Comprised of 19 industry, significant share of the private sector has upgraded academia, and labor representatives, the AMP its technologies. steering committee published a report in 2014 titled Effective technology diffusion requires going “Accelerating U.S. Advanced Manufacturing”, which beyond process innovation to include other types of advances policy recommendations on skills building, innovation, as in business models or in marketing, research, and technology adoption. Overseeing the encouraged by grants. Although often more implementation of these recommendations is the related to product innovation by helping to finance Advanced Manufacturing National Program Office activities ranging from development, prototyping, as part of the National Institute for Science and viability, or commercialization, the support has Technology, with its Manufacturing.gov platform. also financed the introduction of new production It coordinates multi-stakeholder initiatives, such as processes. Many countries use matching grants the Advanced Manufacturing Technology Consortia and technical assistance to facilitate digitization of (AMTech), a grant program for research consortia; SMEs, sometimes by partnering SMEs with large MForesight, a mechanism for soliciting forward- ICT companies. Similarly, vouchers can incentivize looking private sector input on R&D priorities; the collaboration of SMEs and knowledge providers the Manufacturing Extension Partnership (see in starting the implementation of Industry 4.0 below); and Manufacturing USA, a network of processes. Grants, usually smaller, have financed 15 Manufacturing Innovation Institutes, of which the collaboration of SMEs with local consultants or nine have been established (and a further six were universities in supporting technological needs within scheduled to open in 2017).75 These institutes are the SME. Again, however, the evidence is limited on PPPs and received US$600 million in federal and the effectiveness of these approaches, particularly US$1.2 billion in nonfederal funding in 2016. They the adoption of complex technologies. Studies in provide access to state-of-the-art facilities and to OECD countries show positive effects of input workforce training and skills development so as additionality—additional innovation activities—and to shepherd technologies from the research to the positive but very scarce impacts in relation to outputs adoption stage (Manufacturing.gov 2017). and outcomes—introductions of new processes and Governments also support R&D via tax credits, increases in productivity (García-Quevedo, 2004; innovation vouchers, public procurement programs, Zúñiga-Vicente et al., 2014; Becker, 2015). and access to testing facilities. Japan and the United Today, France, Germany, Japan, and the United States provide research tax credits (OECD 2016), States are leading the efforts to deploy existing and while Europe and the United States have provided new policy instruments to facilitate the adoption and incentives for firms to purchase R&D services and diffusion of advanced technologies, though Brazil expertise though innovation vouchers (Shapira and China are also making deep inroads.74 The first and Youtie 2017). The United States has a public steps in these efforts are to set up a central secretariat procurement program run by the Small Business or platform that orchestrates all efforts around Innovation Research program, spending about Industry 4.0, to conduct foresight studies, and to US$2 billion per year on as many as 4,000 contracts create national plans involving multi-stakeholder with SMEs to spur technological innovation and working groups. Germany and the United States commercialization (Rigby et al. 2013). Testing INNOVATION REFORM AGENDA 28 facilities require more adaptation, as they are Kohsetsushi Centers provide free advisory services technology-specific and need to provide the right to SMEs as well as access to laboratories on a cost- tools and equipment. In Germany, for example, there sharing basis (Ezell and Atkinson 2011). are now more than 500 “test beds” providing access The United States supports technology diffusion to digital and physical technology relevant to Industry through the Manufacturing Extension Partnership, 4.0 as well as expert researchers (Industrie 4.0 2017). a network of local centers in all U.S. states at 400 U.S. companies can access testing facilities through locations and with a staff of 1,600 that is funded several of the Manufacturing Innovation Institutes, from federal, state, and industry sources (Ezell and each focusing on a specific technological aspect of Atkinson 2011). In 2017, the federal funding portion advanced manufacturing (President’s Council of of the budget was increased from one-third to one- Advisors on Science and Technology 2014). half, a sign that the government is committed to A growing literature documents high-quality expanding the program (Youtie and Shapira, 2017). managerial practices as a key input and The locations serve SMEs with services for process complementary factor for R&D outlays and firm- improvement, product development, marketing, level innovation (Bresnahan, Brynjolfsson, and Hitt, training, and sustainability; and they connect 2002, Bloon, Sadun and Van Reenen, 2012, Bloom manufacturing SMEs to other private and public et al., 2017). Managerial quality can generally be assistance sources. strengthened by a range of measures including The Korean government’s Centers for Creative enhancing market competition, encouraging Economy and Innovation in 17 regions serve as one- participation in GVCs and cooperation with MNCs, stop shops for entrepreneurs interested in starting a providing additional education to managers, business. They provide mentoring and incubating increasing exposure to international markets, and services for start-ups. Some of the centers are improving bankruptcy procedures to facilitate exit backed by strong local businesses (for example, of the weakest performers. Diffused ownership of Hyundai Motor with the center in Gwangju), while firms also helps, as it provides stronger incentives to others include big data (Gangwon), bio-beauty managers than state or family ownership. (Chungbuk), and carbon (Jeonbuk). Box 3 looks at Governments also directly support improvements support services for businesses in Singapore. through management extension programs. Technology adoption is also supported by standards, These involve advisory services that benchmark which improve compatibility of assets and reduce the companies according to an agreed benchmark of uncertainty of investments. Germany and Japan have the quality of management practices (such as the prioritized the development of standards for Industry World Management Survey or Management and 4.0. In Japan, the Open Robot/Resource interface for Organizational Practices Survey, with the latter the Network (ORiN) is being developed to make it increasingly incorporated as a module in national firm easier to adopt smart connected factory technology surveys), provide a business plan on how to improve (ORiN Forum 2016).76 Three major German management skills over a given period, and support industry associations involved with the Industrie firm management in implementing the business plan. 4.0 Platform are advancing the development of the Such extension activities, inspired by the successful “Referenzarchitektur Industrie 4.0” (RAMI 4.0) or diffusion of agricultural technology, have been reference architecture, which is a first step toward widely adopted. In Germany, Mittelstand-Digital standardization (ZVEI 2015; BMBF 2017). helps SMEs understand the advantages of digital applications and take concrete steps toward Industry Complementary efforts that ensure the supply of 4.0. It does so primarily through more than twenty technical skills are necessary to maximize firms’ “SME 4.0 competence centers” that demonstrate how capacity to innovate and absorb technology. digitization, innovative networking, and Industry Germany and the United States have conducted 4.0 can be used in business practice. In Japan, 162 reviews to determine the future workforce needs that PROMOTING INNOVATION IN CHINA: LESSONS FROM INTERNATIONAL GOOD PRACTICE 29 Box 3: Business Support Services In Singapore Singapore provides a suite of instruments to SMEs that want to upgrade their capabilities, including managerial practices. SMEs start the engagement at the SME Centre, which is generally run by the trade association (chamber of commerce), where the SPRING agency covers the costs of the business development services that they undertake. As a first step, a business advisor visits the firm and provides a diagnostic suggesting the upgrading area and matching the SME to a local private sector consultant. As firms grow more sophisticated, they may be matched with an account manager in SPRING who acts as their “general practitioner”. Singapore provides a broad menu of instruments to support upgrading through SPRING and A*STAR, depending on the sophistication of the SME’s capabilities. The former provides support to management upgrading to those firms primarily at the bottom of the escalator, whereas the latter combines instruments for upgrading technology in firms with more sophisticated capabilities. SPRING offers the following to support and finance upgrading: • Online support and toolkit. Enterprise One is a single-window portal that helps firms navigate the government’s programs. SPRING also has an online toolkit that includes self-help guides on customer service, financial management, human resource capability, marketing, and productivity • Innovation and capability voucher. The entryway to government programs is a S$5,000 voucher to upgrade and strengthen core business operations through consultancy projects in innovation, productivity, human resources, and financial management. • Loans. These include the Micro Loan Program, Local Enterprise Fund Scheme, and Loan Insurance Scheme. The government pays 50 percent of the cost of insurance through a third party for trade credit or working capital. The Local Enterprise Finance Scheme helps SMEs secure financing for productive assets, working with a partner financial firm to get SMEs loans of up to S$15 million for up to 10 years. • Capability and development grant. This program covers from 0 to 70 percent of costs—up to S$100,000—for technical upgrading. Eligible activities include consultancy, manpower, training, certification, upgrading productivity and developing business capabilities for process improvement, product development, and market access. There are 10 areas: brand development, business innovation and design, business strategy development, quality and standards enhancement, financial management, human capital development, intellectual property, franchising and productivity improvement, services excellence, and technology. Firms may apply to SPRING, and may be recruited by SPRING directly or through the chambers of commerce, which are on the lookout for promising firms. • Productivity and innovation credit. Firms may either deduct 40 percent of investment or receive a cash payout of 60 percent up to S$100,000 for each year in all six qualifying activities of investment, including acquisition of information technology and automation equipment, training of employees, registration of patents, trade mark designs and plant varieties, acquisition and in-licensing of intellectual property rights, project design, and R&D. Source: Cirera and Maloney (2017, p. 175–76). INNOVATION REFORM AGENDA 30 are now informing policymakers’ efforts at shaping At the subnational level, ecosystem-based education and training systems.77 The German approaches are growing in prominence. For example, Federal Institute for Vocational Education is funding New York City responded to more than 90,000 in the digitization of training centers and is running employment losses during the 2008–09 financial a two-year program with the Federal Ministry for crisis by introducing programs to harness the laid- Education and Research on “vocational training 4.0” off talent to help build the second-largest tech start- to establish an impact screening of occupations and up ecosystem in the country (after Silicon Valley), sectors, and a high skill labor forecasting system. with almost US$6 billion venture capital investment Countries like Estonia and Finland are integrating in start-ups and over 14,500 listed start-ups in digital skills into basic school education. 2015. The city government introduced mentorship programs, accelerators, incubators, coworking The Skills for America initiative has partnered spaces, events, skills training programs, and other with the Aspen Institute and the Manufacturing supporting services (Mulas and Aranguez, 2016). Institute—a nonprofit body affiliated with the This boom in the urban tech ecosystem evolved National Association of Manufacturers—to help half around the traditional local industries of the city (for a million community college students pursue training example, finance, advertising, media, fashion, and and credentials, making it easier for them to find health care), which not only enabled New York to work with manufacturing companies (White House retain specialized talent at a moment of crisis, but 2011).78 Additional specialized training is provided also increased the competitiveness and innovation of by the Manufacturing Innovation Institutes. its traditional base. PROMOTING INNOVATION IN CHINA: LESSONS FROM INTERNATIONAL GOOD PRACTICE 31 FINANCE, COMPETITIVENESS & INNOVATION INSIGHT | FIRMS, ENTREPRENEURSHIP & INNOVATION SUMMARY OF POLICY RECOMMENDATIONS C hina has achieved a remarkable success in promoting technology absorption and—for a few industries—spurring frontier innovation. This has helped support China’s rapid economic growth and transformation over the last three decades. To continue to develop at a fast pace and become a leading global power in innovation, in line with its ambitious plans as outlined in the Made in China 2025 10-year plan, China needs to address the existing shortfalls in its NIS and improve the translation of innovation inputs and outputs to productivity growth (table 2.4). Each of the table’s specific recommendations can could be based on the public expenditure reviews be implemented by learning from examples of in Colombia. Centralizing support for enterprise international efforts. For instance, putting greater innovation could learn from DARPA in the United emphasis on diffusion of technology could be States and Israel’s Office of the Chief Scientist. informed by the experiences of the Manufacturing And stronger involvement of the private sector in Extension Partnership in the United States or Japan’s the design of innovation policies could be modeled Kohsetsushi Centers. Public support programs to on the EU’s smart specialization. Each of these promote improvement in management skills could recommendations would, however, need to be learn from the experience of Singapore’s SPRING adjusted to China’s economic, institutional, and agency. Efforts to streamline innovation policies cultural context. One size will not fit all. PROMOTING INNOVATION IN CHINA: LESSONS FROM INTERNATIONAL GOOD PRACTICE 33 Table 3. China’s NIS Shortcomings and Policy Recommendations General NIS Shortcoming Specific Recommendations Recommendations Slowing productivity Put greater emphasis • Develop a nationwide system for helping SME growth and widening on facilitating diffusion of manufacturers upgrade their technology and capabilities disparities between technology • Design targeted support programs for SME productivity collaboration with research institutes and universities, performance of top including effective industry liaison programs firms and the rest of the distribution • Review the effectiveness of existing policies and institutions, strengthening those that work and eliminating those that do not • Promote market competition • Use public procurement to create markets for innovative products, especially for SMEs Weak managerial Highlight continuous • Expand management-practice support programs based capacity, particularly management upgrading on international good practice compared as a core innovation • Introduce a support program to include awareness with top firms in competency and best practice campaigns among firms, detailed advanced economies benchmarking information for firms, and technology use, particularly in lagging regions Limited collaboration Nurture a more open • Expand support for interfirm, business–science, and across firms, and and globally integrated international cooperation in research and patents between firms and other innovation system • Further enhance the mobility of researchers, including NIS actors; challenges travel and exchange programs in acquiring advanced technology Greater need to Expand public R&D • Increase the share of public R&D support devoted to develop own knowledge toward basic and “blue basic research, in step with the improvement in the and increase quality of sky” research, and quality of human capital and the governance framework, research outputs move from quantity to including through greater support for universities and quality of research and national basic research programs, such as the National patenting Natural Science Foundation of China • Reduce the emphasis on targeting specific fields and technologies in favor of a more decentralized and bottom-up approach • Strengthen evaluation of research, making greater use of objective peer reviewing • Reward high-quality, rather than high-output, research • Redirect public support to high-quality domestic and international patents SUMMARY OF POLICY RECOMENDATIONS 34 Market and institutional Strengthen IPRs, • Expand the capacity and resources of the relevant failures in creating including stronger administrative agencies and the court system, including the right incentives enforcement specialized IP courts for innovation and • Strengthen enforcement of court orders and formal entrepreneurship, judgments, including the levying of fines sufficient to especially for SMEs change behavior • Encourage moving court cases to third-party jurisdictions in certain circumstances to avoid conflict of interest and further central initiatives to monitor and limit unfair court rulings • Promote nongovernment arbitration and mediation for IP cases Prevalence of linear, Promote a more • Complement the current approach to promoting top-down approaches bottom-up, market- innovation with a more market-based and private sector– to supporting innovation based innovation policy driven innovation policy (for example, one based on the to deal with increasing EU smart specialization concept), especially at regional unpredictability levels of technological • Make selected public support programs open to all development industries and enterprises • Enhance support of ecosystems and expand the network of incubators organized around existing science and technology parks, with links to existing businesses and professional networks Fragmentation of Consolidate and • Streamline innovation policies and programs by carrying innovation policymaking streamline innovation out a public expenditure review of STI, and restructure or and implementation policies eliminate programs that do not perform well • Consider upgrading an existing institution to centralize financing for enterprise innovation • Increase coordination of innovation policy by adapting global good practice to the Chinese institutional context Limited penetration of Expand the use of M&E • Increase the reliance of policymaking on rigorous, rigorous M&E in design quantitative evaluations of policies, including evaluations and implementation of of the innovation policy instrument mix, its functional innovation policy design, efficiency, and effectiveness Limited alignment Promote closer Expand dialogue with the private sector, including by of innovation policy engagement of the using government–business policy platforms and firm instruments with firms’ private sector in opinion surveys needs designing innovation Experiment with new public support instruments, with an policies and support embedded expiry date to ensure periodic evaluation instruments Expand involvement of private sector professionals in the selection process for public innovation support PROMOTING INNOVATION IN CHINA: LESSONS FROM INTERNATIONAL GOOD PRACTICE 35 FINANCE, COMPETITIVENESS & INNOVATION INSIGHT | FIRMS, ENTREPRENEURSHIP & INNOVATION ENDNOTES 1. It took 30 years for electricity and 25 years for telephones to reach 10 percent adoption in the United States, but less than five years for tablet devices to achieve the same rate of adoption. It took an additional 39 years for telephones to reach 40 percent penetration and another 15 before they became ubiquitous. Smartphones, on the other hand, have accomplished a 40 percent penetration rate in just 10 years. According to some futurologists, we are at the threshold of a new technological epoch characterized by the endless acceleration of technical progress. See, for instance, Kelly (2016), Kurzweil (2005) and Diamandis and Kotler (2012 and 2015). 2. Bitcoin was created by an individual inventor or a group of inventors calling themselves Satoshi Nakamoto. Blockchain, as a technology, is essentially a repeatedly updated electronic ledger recording transactions of any kind. 3. https://www.uspto.gov/sites/default/files/aia_implementation/aia-effective-dates.pdf 4. See, for instance, World Bank (2016) for an example of such interaction through Smart Labs in Poland. 5. France, for instance, has revised its Government Procurement Law to strengthen support for innovation in medical instruments and help reduce costs. See, for instance, Georghiou et al. (2014). 6. The actual outcomes can be highly nonlinear. Kancs and Silverstovs (2015) show that initially the impact of R&D on firm productivity can be low—an elasticity of 0.15. However, as the critical mass of expertise and experience accumulates, productivity rises and the elasticity of response can go up to 0.33. 7. See Freeman, Marschke and Wang (2009) and Hall, Mairesse, and Mohnen (2009). 8. However, large outputs of patents do not necessarily follow from increased R&D spending. Hu and Jefferson (2009) found that the correlation between R&D spending in China in 1995–2001 and patenting was on the whole fairly weak. Hu and Jefferson also cite studies using U.S. and Japanese data that point to a weak relationship between patenting and innovation. Eberhardt et al. (2016) arrive at similar findings. 9. Bloom et al. (2017) point to declining returns from and productivity of R&D activity. 10. Boeing, Mueller, and Sandner (2016) argue that the increase in China’s R&D has yet to deliver TFP growth. See also Cirera and Maloney 2017. 11. Andrews et al. (2015) observe that a major reason why productivity is stagnating in Western countries is because the majority of firms in an industry are slow to adopt new technologies that firms at the frontier are already using, and lagging firms are also slow to exit. 12. The World Management Survey sample in China is not very large and comprises large manufacturing firms. Further research in this important area is needed. 13. There were at least 2,500 officially registered tech incubators in China (McKinsey, 2017). 14. National Medium and Long-Term Program for Science and Technology Development. PROMOTING INNOVATION IN CHINA: LESSONS FROM INTERNATIONAL GOOD PRACTICE 37 15. From: https://www.nsf.gov/statistics/2018/nsb20181/report/sections/research-and-development-u-s-trends -and-international-comparisons/cross-national-comparisons-of-r-d-performance. However, PPP measures can inflate R&D spending of China. Nonetheless, nominal measures would show similar trends. 16. These policies have been identified by searching for terms “science”, “technology”, and “innovation”, using the search method of “theme” and “exact match” in the law and policy database of Peking University. 17. According to Molnar (2017), the central government is responsible for about half of total public R&D spending, while SOEs are likely responsible for the rest. 18. http://www.chinadaily.com.cn/a/201802/13/WS5a827ffea3106e7dcc13c829.html 19. Based on the State Council’s ‘Opinions on enhancing basic scientific research’ issued in January 2018. See also http://english.gov.cn/policies/latest_releases/2018/01/31/content_281476032021752.htm 20. The data on the number of policy instruments are subject to two major caveats: first, they count the number of support instruments rather than their budgets (which were not available), making it hard to judge their relative importance; the data may also count policies, which despite a similar official title may in practice have a different focus. 21. 264 national engineering research centers provide technical assistance to SMEs in specific technology areas (Li, 2012). 22. Gazelle firms are midsized, fast-growing firms that frequently produce tradables, engage in exporting, and base their competitiveness on product and process innovation. 23. Preferential treatment of SOEs is likely to result from high risk aversion among the public support institutions, which prefer not to run the risk of dealing with the private sector; privileged access of SOEs to information on the funding instruments; predominance of SOEs in the priority sectors; and the ability of SOEs to influence government policy directly. These findings stand in contrast with the aim of government policy to shift to a more market-oriented approach to support innovation. 24. China had 4.7 million new graduates with degrees in STEM disciplines out of a total of 7.0 million graduates in 2016 (Stapleton, 2017). 25. There were only 18 students enrolled in PhD programs in 1978. Enrollment has increased annually by 23 percent since. The quality of the degrees awarded by many Chinese universities remains an issue as the number of qualified academics to supervise doctoral candidates has risen far more slowly such that, in 2010, the ratio of supervisors to students was 1:6, much lower than in Western countries. Further, it takes an average of three years to receive a doctoral degree in China against an average of five to seven years in Europe and the United States (Majumdar, 2014). 26. On defining and measuring the talent pool, see Simon and Cao (2006). Close to 3.5 million Chinese students have gone abroad to study and the government estimates that about one half have returned, including 110,000 holders of doctoral degrees (Zhou, 2015). According to China’s Ministry of Education, in 2016, 432,500 Chinese graduates returned to China, an increase of 60 percent over 2015 (China Daily, 2018, December 19, p. 16). 27. Freeman and Huang (2015) maintain that overseas training and networking has facilitated the closing of technology gaps. 28. Of the top 10 percent most cited publications, China’s share now stands at 14 percent (OECD, 2017). ENDNOTES 38 29. An author publishing in the 68 highest rated scientific journals is given a score based on the number of articles to which the author contributed. This is known as the Fractional Index. This is then weighted with reference to disciplines to arrive at the Weighted Fractional Index. China’s strength is greatest in chemistry, which contributes 61 percent of the total Weighted Fractional Index (Zhou, 2015). 30. China’s STI system has suffered from fragmentation, research conducted by specialized institutes subject to top-down planning, and a lack of commercial orientation. Cultivating commercial linkages and changing mindsets have been slow processes (Klochikhin, 2013). Moreover, selection processes in academia have suffered from favoritism bias, with, for example, hometown ties to fellow selection committee members increasing the likelihood of selection to the Academies of Sciences and Engineering by 39 percent and subsequently distorting the allocation of university leadership appointments and research grant allocation (Fisman et al., 2018). 31. https://www.nsf.gov/statistics/2018/nsb20181/report/sections/higher-education-in-science-and- engineering/international-s-e-higher-education. 32. Leading the push to patent are Chinese companies such as Huawei, ZTE, Lenovo, Shenzhen Huaxing Optoelectronic, Hongfujin Precision Industry, Sany, BYD, Tencent, SMIC, Mindray Medical, and the Alibaba Group. In 2018, Huawei was the world’s number one filer of international patent applications, https://www.wipo.int/edocs/infogdocs/en/ipfactsandfigures2018/. 33. The citing of Chinese patents by foreigners as an indicator of quality has been growing at an annual rate of 51 percent between 2005 and 2014 (Wei, Xie and Zhang, 2017). 34. https://www.wipo.int/edocs/infogdocs/en/ipfactsandfigures2018/. 35. Goldwind, include Ming Yang, Guodian, Envision, and CSIC. From: http://nawindpower.com/top-10- wind-turbine-makers-for-2015-the-year-of-china. 36. Nonservice/employee generated patents are less likely to be commercialized. Sixty percent of the utility patents and 56 percent of the design patents were of this sort (Prud’homme and Zhang, 2017). Although utility patents are easier to obtain and facilitated technological learning, their contribution to productivity is diminishing as Chinese industry moves up the learning curve. 37. The World Bank (2013) notes that research-industry collaboration is difficult to achieve even in high- income OECD economies. The authors cite firm-level evidence from Eastern Europe and Central Asia, which shows that while 75 percent of firms consider the acquisition of machinery and equipment to be an important channel for acquiring knowledge, less than 1 percent of firms felt the same way about collaborating with universities and public research institutes. However, evidence also shows that grants provided to consortia of firms and research entities can be an effective vehicle for incentivizing collaboration and generating more knowledge outputs through increased likelihood of patenting and commercializing products (Bruhn and McKenzie, 2017). 38. Although the shortage of qualified examiners and legal personnel with the requisite expertise persists. 39. Publication of a validating search report prior to the granting of a patent could be another step to raising quality. 40. This section is based on van Assche and van Biesebroeck (2017) and Misra (2017). 41. http://www.wipo.int/edocs/pubdocs/en/wipo_pub_943_2016.pdf. 42. See Rodrik (2006), Jarreau and Poncet (2012); Schott (2006); Wang and Wei (2010). PROMOTING INNOVATION IN CHINA: LESSONS FROM INTERNATIONAL GOOD PRACTICE 39 43. Similar results have been documented in countries outside of China. Halpern et al. (2009) find that a higher share of imported inputs increased the productivity of firms in Hungary, while Amiti and Konings (2007) obtain similar results for Indonesia. Goldberg et al. (2010) show that easier and cheaper access to imported intermediates in India led domestic firms to introduce additional product varieties, while Colantone and Crino (2014) find that a higher share of newly imported varieties in an industry raises the share of new domestic products in that industry in 25 European countries. 44. From: https://www.bloomberg.com/news/articles/2020-01-18/germany-breaks-korea-s-six-year-streak -as-most-innovative-nation. 45. The technology gaps clearly vary by industry and segment: some Chinese firms (for example, Alibaba, BGI) are among the top in their fields and McKinsey (2017) show that in 2016, China was in the global top three for venture capital investments across a range of leading technologies: fintech, virtual reality, autonomous driving, wearables, education technology, robotics and drones, 3D printing, big data, and artificial intelligence and machine learning. 46. From: https://www.slideshare.net/seogwon/innovation-in-korea-20131023. 47. From: https://www.kiat.or.kr/site/engnew/index.jsp. 48. From: https://www.aaas.org/programs/r-d-budget-and-policy/historical-trends-federal-rd#Agency. 49. Such groups include a Leading Group on Science and Education, a Leading Group on Science and Technology System Reform and Innovation System Development, an Expert Consultation Committee on Innovation-driven Strategies, and other agencies. 50. Descriptions of STAR Metrics: https://www.starmetrics.nih.gov. 51. See, for example, http://www.oecd.org/sti/inno/impact-assessment-public.htm. 52. Regions where the labor force is ageing and shrinking more rapidly because of low fertility and out- migration will be more dependent upon gains in productivity to sustain growth although these gains will be harder to realize. 53. Markusen is referring more to industrial districts or clusters; however, her observation has a bearing on the success of an RIS. 54. Zhang, Wu, and Cooke (2010) note that until recently, China’s technology zones were oriented more toward manufacturing than R&D. 55. Muro and Liu (2017) examine the growth of high-tech metro regions in the United States and show the dominance of the top 20. Liu (2013) shows that geographic proximity to universities has a positive impact on corporate patenting in China but not on the quality of patents. 56. Urbanization economies are defined in http://www.oxfordreference.com/view/10.1093/oi/ authority.20110803114851504. 57. The servicification of manufacturing is now a well-known phenomenon and key to the profitability of many complex products. 58. Research universities and technology parks can encourage innovative entrepreneurial activity by setting up incubators and accelerators, as many have done. See, for instance, http://www.scmp.com/week-asia/ business/article/2085464/business-incubators-look-china-tech-worlds-next-big-thing. ENDNOTES 40 59. McCann and Ortego-Argiles (2016): “New entrepreneurial actions must be based largely on existing capabilities, skills-sets or knowledge-bases, such that diversification takes place in an incremental manner using existing knowledge and drawing on local strengths. These general principles highlight the importance of fostering development trajectories which are both connected to the existing knowledge ecology but at the same time attempt to re-orient the existing trajectories. In order to achieve this it is essential to ensure that local connections and synergies between institutions and actors are as strong as possible and policy actions draw on all of the available local resources in order to build both scale and concentration”. 60. http://gra.org. 61. An interesting finding reported by the report monitoring the EU’s Horizon 2020 Eurostars Program indicated that “The involvement of universities and research institutes did not prove to have a positive contribution to the approval of the applications; the highest success rates corresponded to consortia with only R&D SMEs, with a higher number of R&D SMEs, and with main partners located in the United Kingdom, France and Sweden”. (European Commission, 2017). This applies to all projects and not just those having to do with basic research, but it nevertheless raises a question regarding the contribution of universities in Europe. 62. See references cited in Cirera, Goni and Maloney, 2017. The division of research into basic and applied acquired traction after the publication of Vannevar Bush’s landmark study, “Science, the Endless Frontier”. 63. Acemoglu, Akcigit, and Kerr (2017) conclude that the flow of knowledge tends to be asymmetric, with patenting responding to upstream research. 64. Narayanamurti and Odumosu (2016, pp. 54-55) illustrate the interplay between the two categories of research by showing how the discoveries responsible for the award of six Nobel Prizes and Draper Prizes straddled invention and innovation. “The discovery of the transistor effect…relied on the invention of the bipolar junction transistor and led to the processors and chips that [are at the heart] of computers and cars to the invention of the integrated circuit… The invention of fiber optics combined with the materials engineering and invention of heterostructures made the physical environment and speed of global communications networks possible. In fact, the desire to improve the electrical conductivity of heterostructures led to the unexpected discovery of fractional quantization in two-dimensional systems and a new form of quantum fluid. Each of these could be classifies as basic or applied research, but such classification elides the complexity and multiple nature of the research”. 65. From: https://dc.mit.edu/sites/default/files/Future%20Postponed.pdf. 66. Nanoscale catalysts could transform the production of energy. From: http://www.futurepostponed.org/ blog/2015/9/22/the-ultimate-clean-energy-strategyv. 67. From: http://www.futurepostponed.org/blog/2015/5/6/batteries; this research entails probing the working of new chemistries using advanced nanotechnologies and exploring ways of manufacturing new batteries. Thus, research on storage cannot be neatly compartmentalized into basic, applied, and developmental. 68. From: https://www.scmp.com/news/china/science/article/2189427/chinas-funding-science-and-research -reach-25-cent-gdp-2019. 69. From: https://www.aaas.org/programs/r-d-budget-and-policy/historical-trends-federal-rd#Agency. PROMOTING INNOVATION IN CHINA: LESSONS FROM INTERNATIONAL GOOD PRACTICE 41 70. From: https://www.nsf.gov/about/. 71. If all the parts of the iPhone 6 were manufactured in the United States it would add about US$100 to its price (Kakaes, 2016). 72. From: https://www.insidehighered.com/news/2016/02/26/research-suggests-ways-western-universities- succeed-programs-china. 73. This effort at mastering key technologies extends, it seems, to producing the metal ball casing of ballpoint pens. Early in 2017, the Taiyuan Steel announced that five years of research had finally resulted in a technological breakthrough. This SOE was now able to mass-produce the steel for the “ball socket containing the freely rotating ball at the tip of the pen” (Ng, 2017). 74. From: https://www.staufen.ag/fileadmin/HQ/02-Company/05-Media/2-Studies/STAUFEN.-studie-china -industrie-4.0-index-2015-de_DE.pdf. 75. A recent third-party assessment of Manufacturing USA by Deloitte concludes that it effectively addresses the gap between research and commercialization by connecting members who work in different parts of the R&D spectrum and by de-risking investments. According to the Deloitte report, participating institutes are achieving high degrees of network connectivity and strong member recruitment. The institutes are also becoming integrated into existing regional clusters and strengthening those clusters, thus tying innovation efforts to places with strong advanced manufacturing workforces and enabling R&D knowledge spillovers. 76. The Technology Research Institute of the Japan Society for the Promotion of Machine Industry is supporting the creation of applications for machine tools through ORiN (METI Journal 2015). 77. Fraunhofer, the National Academy of Technology Sciences, and the Industrie 4.0 platform published in 2016 a study titled “Competencies for Industry 4.0 – Qualification Needs and Approaches” that discusses both what enterprises can do to continually train and upgrade their workforce as well as how the German education system needs to adapt. 78. Finland’s government has taken a top-down approach, requiring mandatory teaching of coding to all students as part of basic education, although local municipalities have flexibility in implementing the program. Estonia has not made coding mandatory, but has taken a more bottom-up approach by encouraging PPPs to teach coding and robotics and to engage in other experimental initiatives to advance skills education. The United States has taken a “hybrid” approach of a top-down initiative from the Federal Government to fund the development of a bottom-up network of PPP digital skills vocational training programs for selected populations, including minorities and those needing job retraining. ENDNOTES 42 FINANCE, COMPETITIVENESS & INNOVATION INSIGHT | FIRMS, ENTREPRENEURSHIP & INNOVATION REFERENCES Acemoğlu, Daron, Akcigit, Ufuk and William Kerr. 2017. “The Innovation Network”, Voxeu, 20 January. Adams, Jonathan. 2013. “The fourth age of research”, Nature, volume 497, pages557–560. Adams, James D., Clemmons, Roger and Paula E. Stephan. 2006. “How Rapidly Does Science Leak Out?”, NBER Working Papers 11997, National Bureau of Economic Research, Inc. Ahmadpoor, Mohammad and Benjamin F. Jones. 2017. “The dual frontier: Patented inventions and prior scientific advance”, Science, 11 August. Amiti, Mary and Caroline Fruend. 2010. “The Anatomy of China’s Export Growth”. A chapter in China’s Growing Role in World Trade, National Bureau of Economic Research, Inc., pp 35-56 Andrews, Dan, Criscuolo, Chiara and Peter N. Gal. 2015. “Frontier firms, technology diffusion and public policy: micro evidence from OECD Countries”, OECD Productivity Working Papers November, No. 02 Arora, A. S. Belenzon, and A. Patacconi, 2015. “Killing the Golden Goose? The Decline of Science in Corporate R&D”, NBER Working Paper 20902. Bloom, Nicholas, Brynjolfsson, Eric, Foster, Lucia, Ron S. Jarmin, Megha Patnaik, Itay Saporta-Eksten, and John Van Reenen. 2017. “What Drives Differences in Management?” NBER Working Paper No. W23300, National Bureau of Economic Research, Cambridge, MA. Bloom, Nicholas, Sadun, Roberta, and John Van Reenen. 2017. “Management as a technology”, CEP Discussion Paper 1433. Bloom, Nicholas, Jones, Charles I., Van Reenen, John and Michael Webb. 2019. “Are Ideas Getting Harder to Find?”, NBER Working Paper 23782. Bloom, Nicholas, Raffaella Sadun, and John Van Reenen. 2012. “Americans Do IT Better: US Multinationals and the Productivity Miracle”. American Economic Review no. 102 (1):167–201 Boeing, Phillip. 2016. The Allocation and Effectiveness of China’s R&D Subsidies: Evidence from Listed Firms”, Research Policy 45: 1774–789. Boeing, Philipp, and Elisabeth Mueller. 2019. “Measuring Chinas Patent Quality: Development and Validation of ISR Indices”. China Economic Review, 57, 101331. Boeing, Phillip, Mueller, Elisabeth and Paul Sandner. 2016. “China’s R&D Explosion: Analyzing Productivity Effects Across Ownership Types and Over Time”. Research Policy 45 (1) 159–76. Branstetter, L, B Glennon, and J. B. Jensen. 2018. “The IT Revolution and the Globalization of R&D”, in: Josh Lerner and S. Stern (eds) Innovation Policy and the Economy, Volume 19, University of Chicago Press. Bresnahan, T., Brynjolfsson, Erik, and Hitt, L. 2012. “Information Technology, Workplace Organization, and the Demand for Skilled Labor: Firm-Level Evidence”. Quarterly Journal of Economics, 117 (1), pp.339-376. [31]; PROMOTING INNOVATION IN CHINA: LESSONS FROM INTERNATIONAL GOOD PRACTICE 43 Bruhn, Miriam and David McKenzie. 2017. “Can grants to consortia spur innovation and science-industry collaboration?: regression-discontinuity evidence from Poland”. Policy Research Working Paper; no. WPS 7934; Impact Evaluation series. Washington, D.C. World Bank Group.  CAHS (2011) ‘CAHS Menu of Preferred Indicators and Metrics of Impact’. As of 2 August 2012: http:// www.cahs-acss.ca/wp-content/uploads/2011/09/ROI_Indicators.pdf. Chen, Ling and Barry Naughton. 2016. “An institutionalized policy-making mechanism: China’s return to techno-industrial policy”, Research Policy, vol. 45, issue 10, 2138-2152 Chen, Kaihua and Jiancheng Guan. 2011. “Mapping the functionality of China’s regional innovation systems: A structural approach”, China Economic Review 22(1):11-27 · March 2011  Cirera, Xavier, and William F. Maloney. 2017. Innovation Paradox: Developing-Country Capabilities and the Unrealized Promise of Technological Catch-Up. Washington, D.C.: World Bank Group. Cirera, Xavier, Goñi, Edwin and William Maloney. 2017. “Why poor countries invest too little in R&D”, Voxeu, 29 November 2017 Cooke, Fang Lee. 2008. Competition, Strategy and Management in China. Basingstoke, United Kingdom: Palgrave Macmillan. Diamandis, Peter, and Steven Kotler. 2012. Abundance: The Future is Better than You Think. New York: Free Press. Diamandis, Peter, and Steven Kotler. 2015. Bold: How to Go Big, Create Wealth, and Impact the World (New York: Simon and Schuster). Dobbs, Richard, Manyika, James and Jonathan Woetzel. 2015. No Ordinary Disruption: The Four Global Forces Breaking All the Trends (PublicAffairs, May 2015). Doloreux, David and Saeed Partob. 2005. “Regional innovation systems: Current discourse and unresolved issues”, Technology in Society 27, 133–153 Eaton, Jonathan and Samuel Kortum. 1999. “Trade in Capital Goods”. NBER Working Paper Nos. 8070. Eberhardt, Markus, Helmers, Christian, and Zhihong Yu. 2017. “What can explain the Chinese patent explosion?” Oxford Economic Papers, Volume 69, Issue 1, January, Pages 239–262,  Enright, Michael J. 2017. Developing China: the Remarkable Impact of Foreign Direct Investment, Abingdon and New York: Routledge, 2017 Etzkowitz, Henry. 2003. “Innovation in Innovation: The Triple Helix of University-Industry-Government Relations”, Social Science Information 42(3):293-337 · September European Commission. 2017a. “Analysis of impact of completed Eurostars-1 projects. Final Report”, European Commission, EUR 2017.3358. European Commission. 2017b. “Horizon 2020. Key Findings from the Interim Evaluation”. Fang, Jing, Hui He, and Nan Li. 2016. “China’s Rising IQ (Innovation Quotient) and Growth: Firm-Level Evidence”. IMF Working Paper 16/249, International Monetary Fund, Washington DC. Fang, Lily, Josh Lerner and Chaopeng Wu. 2016. “Intellectual Property Rights Protection, Ownership, and Innovation: Evidence from China”. NBER Working Paper Nos. 22685. REFERENCES 44 Fernandes, A., A. Mattoo, H. Nguyen, and M. Schiffbauer. 2017. “The Internet and Chinese Exports in the Pre-Alibaba Era”, World Bank Policy Research Working Paper 8262. Fisman, Raymond, Shi, Jing, Wang, Yongxiang and Rong Xu, 2018. “Social Ties and Favoritism in Chinese Science”, Journal of Political Economy, vol. 126(3), pages 1134-1171. Fleming, Nic. 2017. “Big science in a small country”, Nature volume 550, pages S49–S51. Freeman, Richard B. and Wei Huang. 2015. “China’s ‘Great Leap Forward’ in Science and Engineering”, NBER Working Paper No. 21081, April. Freeman, Richard B., Marschke, Gerald, and Andrew Wang. 2009. “Estimating the Payoff to R&D”, mimeo. Foray, Dominique, Paul A. David, and Bronwyn Hall. “Smart specialisation–the concept”.  Knowledge economists policy brief 9.85 (2009):100. Foray, Dominique. 2015. Smart Specialization: Opportunities and challenges for regional innovation policies; Routledge Press. Georghiu, Luke, Edler, Jakob, Uyarra, Elvira and Jillian Yeow. 2014. “Policy instruments for public procurement of innovation: Choice, design and assessment”. Technological Forecasting and Social Change, Volume 86, July 2014, Pages 1-12 Gertner, Jon. 2012. The Idea Factory: Bell Labs and the Great Age of American Invention. Penguin Books. Glennie, Alex, and Kirsten Bound. 2016. “How Innovation Agencies Work: International Lessons to Inspire and Inform National Strategies”. Nesta, London, United Kingdom. Goh, Chorching, Li, Wei; Xu, and Colin Li. 2015. “R&D returns, spillovers, and firm incentives : evidence from China”. Policy Research working paper; no. WPS 7191. Washington, D.C.: World Bank Group Goni, Edwin and William Maloney. 2014. “Why don’t poor countries do R&D? Varying rates of factor returns across the development process”, European Economic Review, vol. 94, issue C, 126-147 Gould, Skye and Antonio. Villas-Boas. 2016. “Here’s where all the components of your iPhone come from”, Business Insider, April 13. Grossman, Gene, McCalman, Phillip and Robert Staiger. 2019. “The ‘new’ economics of trade agreements”. Voxeu,  23 September 2019 Guan, JianCheng and Lanxin Pang. 2017. “Industry Specific Effects on Innovation Performance in China”. China Economic Review. vol. 44(C), pages 125-137. Gupta, Sujata. 2016. “Science superpowers find common ground”, Nature volume 539, pagesS6–S9. Guthrie, Susan, Watu Wamae, Stephanie Diepeveen, Steven Wooding and Jonathan Grant. 2013. “Measuring research: A guide to research evaluation frameworks and tools”. RAND Corporation. Hale, Katherine. 2012. “Collaboration in Academic R&D: A Decade of Growth in Pass-Through Funding”, National Science Foundation, National Center for Science and Engineering Statistics. Hall, Bronwyn, Mairesse, Jacques, and Pierre Mohnen. 2010. “Returns to R&D and productivity”, in: B. H. Hall and N. Rosenberg (eds.), Handbook of the Economics of Innovation, Vol 2. Elsevier. Helmers, Christian and Henry Overman. 2017. “The impact of big research infrastructure facilities on local scientific output”, Voxeu, 25 November. PROMOTING INNOVATION IN CHINA: LESSONS FROM INTERNATIONAL GOOD PRACTICE 45 Hu, Albert Guangzhou and Gary H. Jefferson. 2009. “A great wall of patents: What is behind China’s recent patent explosion?”, Journal of Development Economics, Volume 90, Issue 1, September 2009, Pages 57-68 Jarreau, Joachim and Sandra Poncer (2012). Export Sophistication and Economic Growth: Evidence from China. Journal of Development Economics. 97(2): 281-29 Jia, J., and G. Ma. 2017. “Do R&D tax incentives work? Firm-level evidence from China”. China Economic Review, 46, 50-66. Kakaes, Konstantin. 2016. “The All-American iPhone”, MIT Technology Review, June 9. Kancs, d’Artis and Boriss Siliverstovs. 2016. “R&D and non-linear productivity growth”. Research Policy, vol. 45, issue 3, 634-646 Kapil, Natasha, and Anwar Aridi. 2017. “Innovation Agencies: Takeaways from Emerging Innovation Systems”. Unpublished Report, World Bank, Washington, D.C. Kapil, Natasha, Piatkowski, Marcin, Radwan, Ismail, and Juan Julio Gutierrez. 2012. Poland Enterprise Innovation Support Review: From Catching Up to Moving Ahead. The World Bank. Kelly, Kevin. 2016. The Inevitable: Understanding the 12 Technological Forces that Will Shape our Future (New York: Viking) Klochikhin, Evgeny. 2013. “Innovation system in transition: Opportunities for policy learning between China and Russia”, Science and Public Policy, Volume 40, Issue 5, October, p. 657–673 Kurzweil, Ray. 2005. The Singularity is Near (New York: Penguin Book) Lee, Keun. 2016. Economic Catch-Up and Technological Leapfrogging: The Path to Development and Macroeconomic Stability in Korea, Edward Elgar; Li, Y. 2012. ERAWATCH Country Reports 2012: China. Brussels: ERAWATCH Li, Xibao. 2012. “Behind the recent surge of Chinese patenting”. Research Policy; 41, pp. 236–249 Lin, Chen, Ping Lin and Frank M. Song. 2009. “Property Rights Protection and Corporate R&D: Evidence from China”. Journal of Development Economics. 93(1):49-62 Liu, Xielin, Schwaag Serger, Sylwia, Tagscherer, Ulrike and Amber Chang. 2017. “Beyond catch-up—can a new innovation policy help China overcome the middle-income trap?” Science and Public Policy, 1-14. Liu, Xielin and Peng Cheng. 2011. “Is China’s Indigenous Innovation Strategy Compatible with Globalization?”, Policy Studies, No. 61 Liu, Wan-Hsin. 2013. “The role of proximity to universities for corporate patenting: provincial evidence from China”, The Annals of Regional Science volume 51, pages 273–308. Majumdar, Kakoli. 2014. “Where does China stand in the global PhD boom?” Editage Insights, August 18. Markusen, Ann. 1996. “Sticky Places in Slippery Space: A Typology of Industrial Districts”, Economic Geography, Vol. 72, No. 3, pp. 293-313 Malerba, Franco. 2006. “Sectoral systems: how and why innovation differs across sectors”, in: Jan Fagerberg and David C. Mowery (eds.). The Oxford Handbook of Innovation. REFERENCES 46 McCann, Philip & Ortega-Argilés, Raquel, 2016. “Smart Specialisation: Insights from the EU Experience and Implications for Other Economies”, INVESTIGACIONES REGIONALES - Journal of REGIONAL RESEARCH, Asociación Española de Ciencia Regional, issue 36, pages 279-293. McKinsey. 2017. “China’s Digital Economy. A Leading Global Force”, McKinsey Discussion Paper, August. Mulas, Victor; Gastelu Iturri Aranguez, Mikel. 2016. “New York City: Transforming a city into a tech innovation leader”. Washington, D.C.: World Bank Group. Narayanamurti, Venkatesh and Toluwalogo Odumosu. 2016. Cycles of Invention and Discovery. Harvard University Press. Cambridge Mass. 2016. Nature. 2017a. “A Big Science Centre in Shanghai to Lead Innovation. Nature 545 (25 May 2017) Nature. 2017b.  “How China could make the most of its beamlines”, November 14, 2017 Nature. 2016. “Catch Them If You Can”, Nature volume 535, pagesS68–S76. Nature. 2015. From: https://www.nature.com/articles/527S68a Ng, Eric. 2017. “Why cracking the secret of the ball-point pen is such a big deal for Taiyuan Steel”. South China Morning Post, Feb 1. Molnar, Margit. 2017. “Boosting firm dynamism and performance in China”, OECD Economics Department Working Papers, No. 1408, OECD Publishing, Paris. Muro, Mark and Sifan Liu. 2017. “Tech in metros: The strong are getting stronger”, Brookings, March 8 OECD. 2017. “China Economic Survey 2017”. OECD Publishing, Paris. OECD 2015. The Innovation Imperative: Contributing to Productivity, Growth and Well-Being. OECD Publishing, Paris. OECD 2014. “Industry and Technology Policies in Korea”, OECD Reviews of Innovation Policy, OECD. Olabisi, Michael. (2017. “The Impact of Exporting and FDI on Product Innovation: Evidence from China”. Contemporary Economic Policy. 35(4): 735-750. Rodrik, Dani. 2008. ‘Normalizing Industrial Policy’. Commission on Growth and Development Working Paper; no. 3. Washington, D.C.: World Bank. OECD. 2014b.Centre Stage – Driving Better Policies from the Centre of Government. OECD. OECD. 2008. “OECD Reviews of Innovation Policy: China”, OECD. Paic, A. and C. Viros (2019), “Governance of science and technology policies”, OECD Science, Technology and Industry Policy Papers, No. 84, OECD Publishing, Paris, Prodi, Giorgio, Federico Frattini and Francesco Nicolli. 2016. “Regional Innovation Systems in China: A long-term perspective based on patent data at a prefectural level”. SEEDS Working Paper Series, 3/2016. Prud’homme, Dan, and Han Song (Eds). 2016. Economic Impacts of Intellectual Property-Conditioned Government Incentives. Springer 2016. PROMOTING INNOVATION IN CHINA: LESSONS FROM INTERNATIONAL GOOD PRACTICE 47 Prud’homme, Dan and T. Zhang. 2017. “Evaluation of China’s Intellectual Property Regime for Innovation Authors”, a background paper to DRC-World Bank (2019). Rodrik, Dani. 2006. “What’s So Special about China’s Exports?”, NBER Working Paper No. 11947. Shen, Keting, Wang, Jin, and John Whalley. 2015. “Measuring Changes in the Bilateral Technology Gaps between China, India and the U.S. 1979 – 2008”, NBER Working Paper No. 21657 Simon, Denis F. and Cong Cao. 2006. “China’s Emerging Science and Technology Talent Pool: A Quantitative and Qualitative Assessment”, mimeo. Stapleton, Catherine. 2017. “China now produces twice as many graduates a year as the US”. World Economic Forum, April 13. Sun, Zhen, Wright, Brian, Cohen, Mark and Zhen Lei. 2015. “Are Chinese Patent Applications Politically Driven? Evidence from China’s Domestic Patent Applications”, New Perspectives on Innovation and Intellectual Property in China USPTO-UCSD Workshop, May 19, 2015 Sun, Xiuli; Li, Haizheng; Ghosal, Vivek. 2017.“Firm-level Human Capital and Innovation: Evidence from China”, CESifo Working Paper, No. 6370, Center for Economic Studies and Ifo Institute (CESifo), Munich.  USPTO. 2018. From: https://www.uspto.gov/web/offices/ac/ido/oeip/taf/st_co_18.htm Wagner, Caroline S. and Koen Jonkers. 2017. “Open countries have strong science”, Nature, 23 March. Wang, Zhi and Shang-Jin Wei. 2008. “What Accounts for the Rising Sophistication of China’s Exports”. NBER Working Paper Nos. 13771. Wei, Shang-Jin, Zhuan Xie, Xiaobo Zhang. 2017. “From ‘Made in China’ to ‘Innovated in China’: Necessity, Prospect, and Challenges”. Journal of Economic Perspectives 31(1), 49–70. WIPO. 2019. “IP Facts and Figures 2019”, World Intellectual Property Organization. WIPO/INSEAD/Cornell. 2019. “Global Innovation Index 2019”. World Bank. 2018. “China Green Innovation Study”, background paper to DRC-World Bank (2019). World Bank. 2016. Toward an Innovative Poland: Entrepreneurial Discovery Process and Business Needs Analysis. Washington D.C., World Bank. World Bank. 2013. “Western Balkans Regional R&D Strategy for Innovation”; World Bank. Xu, Bin and Jianmao Wang. 1999. “Capital Goods Trade and R&D spillovers in the OECD”. Canadian Journal of Economics. 32(5): 1258-1274. Yang, Wei. 2016. “Policy: Boost basic research in China”, Nature, 22 June. Youtie, Jan and Philip Shapira, 2017. “Exploring public values implications of the I-Corps program”, The Journal of Technology Transfer, Springer, vol. 42(6), pages 1362-1376, December. Zhang, Fangzhu, Wu, Fulong, and Philipp Cooke. 2010. “Changing National and Regional Innovation Systems and Regional Distribution of R&D”, Geography Compass 4(6):532-544. June 2010  Zhao, Yanyun and Siming Liu. 2011. “Effect of China’s Domestic Patents on Total Factor Productivity: 1988-2009”, Renmin University. School of Statistics. Zhou, Yingying. 2015. “The rapid rise of a research nation”, Nature, volume 528, p. S170–S173 REFERENCES 48