研究和創(chuàng)新政策與生產(chǎn)力增長

1、Table of contents HYPERLINK l _bookmark0 Introduction 2 HYPERLINK l _bookmark1 R&I and growth 2 HYPERLINK l _bookmark2 Main features of R&I 2 HYPERLINK l _bookmark6 Main enablers for R&I creation, adoption and diffusion 4 HYPERLINK l _bookmark7 Size of the market 4 HYPERLINK l _bookmark9 Competition

2、 and intellectual property right protection 5 HYPERLINK l _bookmark10 Human capital 5 HYPERLINK l _bookmark11 Financing of R&I 6 HYPERLINK l _bookmark12 Evidence on rates of return from R&I 6 HYPERLINK l _bookmark13 R&I policies 7 HYPERLINK l _bookmark14 The case for R&I policies: market failure 7 H

3、YPERLINK l _bookmark16 R&I policies: why it may not work: government failure 9 HYPERLINK l _bookmark18 The R&I policy toolkit 10 HYPERLINK l _bookmark19 R&I tax incentives 10 HYPERLINK l _bookmark20 Direct subsidies 11 HYPERLINK l _bookmark21 Loans and public venture capital 11 HYPERLINK l _bookmark

4、22 Public R&I and public-private partnerships 12 HYPERLINK l _bookmark23 Public procurement 12 HYPERLINK l _bookmark24 Patents 12 HYPERLINK l _bookmark26 Regulations 13 HYPERLINK l _bookmark27 Main R&I policies deployed in the EU 13 HYPERLINK l _bookmark28 R&I policy at the EU: Framework Programmes

5、13 HYPERLINK l _bookmark32 Major R&I Policy instruments at the EU and its Member States 15 HYPERLINK l _bookmark34 R&I policies and growth 20 HYPERLINK l _bookmark35 R&I policies: do they work to stimulate private R&I? Evidence from evaluations at the micro HYPERLINK l _bookmark35 level 20 HYPERLINK

6、 l _bookmark36 R&I tax incentives 20 HYPERLINK l _bookmark39 Direct subsidies 22 HYPERLINK l _bookmark40 The R&I Policy Toolkit 23 HYPERLINK l _bookmark41 R&I (policies) and their effect on (productivity) growth: evidence from macro-models 24 HYPERLINK l _bookmark44 NEMESIS 25 HYPERLINK l _bookmark4

7、5 QUEST III 27 HYPERLINK l _bookmark47 Conclusions and further steps 30 HYPERLINK l _bookmark48 References 32IntroductionBefore research and innovation (R&I) policies can be identified as instruments for economic growth, and therefore of smart government spending, several questions must be answered.

8、 First: does R&I contribute to growth? At present, it is widely acknowledged that innovation is an important force behind long-run economic growth. In particular, models that use an endogenous growth framework make a strong case for the growth power from R&I and innovation (eg Aghion, 2006). But thi

9、s does not yet make the case for R&I policies. Will R&I policies lead to sufficient innovation and growth to cover the opportunity costs of using public funds for R&I?To address these questions, we have reviewed the evidence on the impact of R&I policies. We first looked at the evidence of the impac

10、t of public intervention on private R&I and innovation, which is mostly from micro-analysis. To assess the impact from public R&I on growth, we needed to complement micro-results with a macro-perspective. To this end, we looked at how R&I and R&I policies perform in affecting GDP growth and jobs in

11、the applied macro-models most commonly used in European Union policy analysis. We start with a brief description of the critical features and enablers of the R&I process needed to appropriately account for its role HYPERLINK l _bookmark3 1 in economic growth, as well as the main policies aimed at pr

12、omoting R&I to boost growth. A proper understanding of the features of R&I, and its enablers and policies, is important to be better able to use the appropriate models and data to evaluate the impact of R&I policies. We conclude with some recommendations for improving R&I policy assessment for growt

13、h.R&I and growthBased on the literature on R&I, this section describes and analyses the critical features of the innovation process needed to account for R&Is role in economic growth and welfare.Main features of R&ITechnological knowledge is the sum of techniques employed in the production of goods

14、and services. It results from the cumulative aggregation of new technological ideas or innovations, resulting from a systematic research and development (R&I) effort and benefiting from multiple forms of knowledge spillovers. R&I refers to those economic activities undertaken with the purpose of imp

15、roving the actual1 See also Veugelers (2016) and European Commission (2020b) for more on R&I policies and their impact on growth.state of technology. As for physical and human capital, R&I is a form of investment that cumulates in the stock of technological knowledge. The stock of technological know

16、ledge also relates to intangible capital, which includes computerised information, innovative property and economic competencies HYPERLINK l _bookmark4 2.Innovations represent additions to the stock of technological knowledge, and result from R&I activity. In the pioneering work of Romer (1990) and

17、Aghion and Howitt (1998), R&I activities are undertaken with the purpose of improving the actual state of technology, which in Romer (1990) takes the form a new product, and in the Schumpeterian framework a better-quality version of an existing product (see Aghion and Howitt, 1998).During the R&I st

18、age, researchers benefit from their past experience and are conditioned by their state of knowledge and their past technological choices. This is where path dependency or technological cumulativeness comes into the picture. To quote Dosi and Nelson (2010), “ advances are likely to occur in the neigh

19、bourhood of the techniques already in use within the firm.” Following Griliches (1979), one way of modelling the cumulative nature of R&I is to consider not the flow of R&I expenditure but the accumulated stock of past and present R&I expenditure as the appropriate variable to affect productivity, o

20、r to enter as an input in an extended production function. Learning-by- doing and learning-by-using are other important elements in knowledge building HYPERLINK l _bookmark5 3.In a world of open innovation, firms may collaborate in their research efforts, exploiting knowledge complementarities. Firm

21、s may also benefit from scientific progress, coming from universities or public research labs. Instead of searching themselves for new ideas, firms may prefer to acquire technological knowledge, in embodied or disembodied form, eg by licensing the latest technology.A specific feature of technologica

22、l knowledge is its non-rival nature: it can be used by many agents at the same time. In this sense, it entails high generation costs but can be easily reproduced. However, technological knowledge does not get transmitted as easily. First, technological knowledge is partially tacit, that is, it canno

23、t be entirely explained in a manual by means of words, symbols or graphs, as opposed to codified knowledge. Second, it is cumulative, meaning that it cannot be understood without grasping prior knowledge. Third, it cannot be assimilated, adopted and reproduced without incurring substantial costs rel

24、ated to building the needed absorptive capacity. It bears the cost of2 Corrado et al (2005, 2009) extended, in the context of the economic growth literature, the definition of a production technology to include the stock of intangible capital. They found for the United States that the growth rate of

25、 output per worker increases more rapidly when intangible capital is included, that capital deepening (on physical and intangible capital) becomes the unambiguously dominant source of growth, diminishing the role of total factor productivity, and that the labour income share has significantly decrea

26、sed over the last 50 years because of the rise of intangible capital.3 The role of learning-by-doing in the growth process was first studied by Arrow (1962).learning, adaptation and reproduction. Technological knowledge is therefore not a free good that falls like manna from heaven for economic acto

27、rs wishing to use it, but requires a deliberate effort on the part of the actors to generate, adopt and use it. In summary, the process of building technological knowledge can considerably benefit from knowledge spillovers. In order to benefit from R&I spillovers, firms have to develop their absorpt

28、ive capacity HYPERLINK l _bookmark8 4. The absorptive capacity of a firm positively depends on the accumulation of its previous R&I investments. It helps understanding why private R&I and knowledge spillovers are likely to be complements in the creation of new technologies.New technological ideas, a

29、fter discovery, diffuse throughout the economy by affecting the design of products and production processes of firms in different industrial sectors and geographical locations. This process is referred as the diffusion of new technologies. Extensive empirical work was undertaken by Comin et al (2008

30、) and Comin and Hobijn (2004, 2010), among others, describing the long process of diffusion of new technologies across time and countries. The diffusion of technologies can take place through the various mechanisms by which knowledge spills over across space and time: trade relationships, foreign di

31、rect investment, mergers and acquisitions, movement of personnel, patents, patent citations, publications, research collaborations and networks. A new product or process is more likely to be widely and/or more quickly adopted if it uses existing or familiar technologies, or if complementary goods or

32、 services already exist. It is not always the superior technology that gets adopted.Main enablers for R&I creation, adoption and diffusionSize of the marketFirms may be more willing to do R&I if there is a large market, enabling them to recover their R&I investment expenditure quickly. The market ca

33、n be national or international depending on the presence or not of trade and non-trade barriers. Acemoglu and Linn (2004) found for the pharmaceutical industry that potential market size had a major effect on the entry of non-generic drugs and new molecular entities.4 For Cohen and Levinthal (1990),

34、 absorptive capacity was the “firms ability to recognize the value of new, external information, assimilate it, and apply it to commercial ends.” In more precise terms, Leahy and Neary (2007) defined a firms absorptive capacity as “its ability to absorb spillovers from other firms.” Cohen and Levint

35、hal (1989) “suggest that R&I not only generates new information, but also enhances the firms ability to assimilate and exploit existing information.”Competition and intellectual property right protectionPatent systems grant temporary monopoly rights to innovators in order to protect them from being

36、copied or imitated. By restricting competition, the patent system aims at solving the market distortion generated by the non-rivalry feature of technological knowledge; patents are expected to promote innovation, restoring the incentives to innovate.Competition, however, can be beneficial to innovat

37、ion too, since in competitive environments firms innovate with the hope of escaping competition (see Aghion et al, 2001). This argument relates to the so-called replacement effect: entrant firms have more incentives to innovate than incumbents since innovation allows them to steal at least partially

38、 the monopoly rents of incumbents (Arrow, 1962). In an extreme case when the prospect of entry is very low, incumbents have little incentive to innovate. However, if there is a risk of their monopoly rents being stolen, incumbents innovate to escape competition from potential entrants. Competition m

39、ay also promote innovation through the type of market size effect described above, if it results in increasing the size of the market by lowering prices.Theoretical and empirical research has found that the relationship between competition and innovation follows an inverted U-shape. When competition

40、 is low, incumbents do not need to protect their monopoly rents, having little incentive to innovate (replacement effect). When markets become more competitive, incumbent firms innovate more to escape competition (escape competition effect). However, when there is too much competition, the gains fro

41、m innovation dilute to the point that neither incumbents nor potential entrants have incentives to innovate (Schumpeters argument). Aghion et al (2005) found strong evidence of an inverted U-shape relationship between product market competition and innovation, and developed an endogenous growth mode

42、l to understand this evidence. Aghion et al (2009b) found that “the threat of technologically advanced entry spurs innovation incentives in sectors close to the technology frontier, where successful innovation allows incumbents to survive the threat, but discourages innovation in laggard sectors, wh

43、ere the threat reduces incumbents expected rents from innovating.”Human capitalInnovation requires an education system able to produce a large enough body of scientists and applied researchers capable of moving the frontier of knowledge. The adoption of new technologies also requires a large body of

44、 highly qualified workers able to easily understand and operate the frontier technology. There must also be a major effort to make the frontier technology user-friendly. Human capital and skills formation are endogenously determined by educational choices and training.The role of human capital forma

45、tion for economic development has been emphasised by Lucas (1988).Financing of R&IIn a world of imperfect information, innovation financing faces a problem of asymmetric information between the innovator and the fund provider. Because of the non-rival nature of knowledge, the innovator has no intere

46、st in sharing with the fund provider some of the information the latter would need to justify the funding. Therefore, R&I is, as much as possible, financed through internal funds. In the absence of sufficient self-funding possibilities, external funding will have to be accessed. This external fundin

47、g can be private or public. Major sources of private funding include bank financing, capital markets and venture capital. Major sources of public funding include grants, subsidies, tax incentives and public venture capital. In their survey of empirical evidence, Hall and Lerner (2010) concluded “tha

48、t while small and new innovative firms experience high costs of capital the evidence for high costs of R&I capital for large firms is mixed. Nevertheless, large established firms do appear to prefer internal funds for financing such investments and they manage their cash flow to ensure this.”Evidenc

49、e on rates of return from R&ITo assess the effects of R&I on innovation, the concept of knowledge production functions is used. Innovation is assumed to be produced by means of an R&I technology using as inputs labour, capital (tangible and intangible, including research infrastructure, computers an

50、d software) and other intermediate inputs (including research materials such as protein structures). R&I labour is to a large extent comprised of scientists and technicians, ie highly skilled and specialised labour, whose research skills are acquired through specific R&I education (typically PhD deg

51、rees) and training. The productivity of firm-specific R&I technologies benefits from multiple knowledge spillovers, the assimilation of which may require some form of absorptive capacity.To obtain evidence on rates of return from R&I knowledge, production functions are estimated. Hall et al (2009),

52、summarising results from these studies, concluded that the rate of return in developed countries has been strongly positive, with the estimated private rate of return to R&I usually exceeding that of physical capital. Most studies obtain rates of return within the range of 10 percent to 30 percent.

53、They also report a large heterogeneity in rates of return for R&I across firms, technologies, sectors and countries.The returns to R&I may depend on the existence of complementarities between R&I and information and communication technologies (ICT), which allow productivity gains from doing research

54、 (Mohnen et al, 2018), or from in-house research combined with purchased technology (Cassiman and Veugelers, 2006), or from process and organisational innovation (Bresnahan et al, 2002).Another important issue is whether returns to R&I are constant. The original Romer (1990) model assumed that retur

55、ns to R&I are constant, which generates the undesirable property known as scale effect: since returns are constant, the growth rate of the economy positively depends on the economys size. Jones (1999), among others, argued that scale effects are counterfactual since large countries dont grow on aver

56、age faster than small countries. This controversy gave rise to the so-called semi-endogenous growth models, under the assumption of decreasing returns to R&I. Bloom et al (2017) claimed that ideas are getting harder to get, pointing out that research effort is increasing sharply, while research prod

57、uctivity is declining substantially in a wide range of sectors. By contrast, returns may also increase because of intertemporal R&I spillovers; the so-called standing on the giants shoulders argument (Scotchmer, 1991; Caballero and Jaffe, 1993).R&I policiesThe case for R&I policies: market failureTh

58、e fundamental justification for government support for research is the classic market failure argument: markets do not provide sufficient incentives for private investment in research owing to the non-appropriable, public good, intangible character of knowledge and its risky nature. In addition, pub

59、lic research is needed to meet specific needs of public interest, common goods that the market would not supply on its own, such as defence, public health or a clean environment. Once invented, the new knowledge created from R&I is non-rival and only partially non-excludable. Others may learn and us

60、e the knowledge, without necessarily paying for it. It is these spillovers, which include pure knowledge spillovers, as well as pecuniary spillovers, that lead to social rates of return above private rates of return and private investment levels below socially desirable levels. This divergence betwe