The economic benefits of publicly funded basic …Research Policy 30 2001 509–532Ž. The economic benefits of publicly funded basic research: a critical review Ammon J. Salter),

Ž .Research Policy 30 2001 509–532www.elsevier.nlrlocatereconbase

The economic benefits of publicly funded basic research:a critical review

Ammon J. Salter), Ben R. MartinSPRU — Science and Technology Policy Research, UniÕersity of Sussex, Falmer, Brighton BN1 9RF, UK

Accepted 9 February 2000

Abstract

This article critically reviews the literature on the economic benefits of publicly funded basic research. In that literature,three main methodological approaches have been adopted — econometric studies, surveys and case studies. Econometricstudies are subject to certain methodological limitations but they suggest that the economic benefits are very substantial.These studies have also highlighted the importance of spillovers and the existence of localisation effects in research. Fromthe literature based on surveys and on case studies, it is clear that the benefits from public investment in basic research cantake a variety of forms. We classify these into six main categories, reviewing the evidence on the nature and extent of eachtype. The relative importance of these different forms of benefit apparently varies with scientific field, technology andindustrial sector. Consequently, no simple model of the economic benefits from basic research is possible. We reconsider therationale for government funding of basic research, arguing that the traditional ‘market failure’ justification needs to beextended to take account of these different forms of benefit from basic research. The article concludes by identifying someof the policy implications that follow from this review. q 2001 Elsevier Science B.V. All rights reserved.

Keywords: Economic benefits; Basic research; Government funding

1. Introduction

The relationship between publicly funded basicresearch and economic performance is an importantone. Considerable government funds are spent onbasic research in universities, institutes and else-where, yet scientists and research funding agenciesconstantly argue that more is needed. At the sametime, governments face numerous competing de-mands for public funding. To many, the benefitsassociated with public spending on, say, health oreducation are more obvious than those from basic

) Corresponding author.Ž .E-mail address: [email protected] A.J. Salter .

research. However, as this article will show, there isextensive evidence that basic research does lead toconsiderable economic benefits, both direct and indi-rect. Those responsible for deciding how the limited

Žpublic funds available are to be distributed and forensuring public accountability in relation to that

.expenditure should therefore be familiar with thefull range of relevant research. To this end, wereview and assess the literature on the economicbenefits associated with publicly funded basic re-search.

As we shall see, although the existing literaturepoints to numerous benefits from publicly fundedbasic research, there are many flaws or gaps in theevidence. These stem from a variety of sources.

0048-7333r01r$ - see front matter q 2001 Elsevier Science B.V. All rights reserved.Ž .PII: S0048-7333 00 00091-3

( )A.J. Salter, B.R. MartinrResearch Policy 30 2001 509–532510

Some are related to conceptual problems regardingthe nature of basic research and how this may bechanging, and the form of its outputs — whether this

Žis information or knowledge and whether the latter.is codified or tacit , or whether other types of output

such as trained people and new instrumentation areat least as important. There are also methodologicalissues about the approaches employed for analysingand assessing the benefits from research — forexample, whether one can legitimately apply tradi-tional economic tools such as production functions toscience, or the validity of using scientific paperscited in patents as a measure of the links betweenscience and technology. These conceptual andmethodological problems point to areas where fur-ther research is needed.

In what follows, we first define the area of re-search covered in this study before examining inSection 3 the nature of the economic benefits ofbasic research and the different methodological ap-proaches to measuring them. The next two sectionsthen critically review and synthesise the main typesof academic literature of relevance here. Section 4deals with econometric studies on the relationshipbetween research and productivity, the rates of returnto research and ‘spillovers’. Section 5 distinguishessix main types of economic benefit from basic re-search and discusses empirical findings on each ofthese. The final section identifies the main lessonsfrom the literature reviewed and the policy conclu-sions to be drawn.

2. Definitions and scope

The review is concerned primarily with basicresearch including both ‘curiosity-oriented’ researchŽundertaken primarily to acquire new knowledge for

. Žits own sake and ‘strategic’ research undertakenwith some instrumental application in mind, although

. 1the precise process or product is not yet known .

1 This definition should not be taken as implying a simplelinear model of innovation. Basic research is just one of manyinputs to technology and innovation, and new technologies orinnovations, in turn, can have an impact on basic research. Itshould also be noted that the concept of ‘strategic’ research is

Ž .very similar to the OECD category of ‘ application oriented’basic research.

However, much of the literature reviewed uses otherterms such as ‘science’, ‘academic research’ or just‘research’, categories that are not identical with‘basic research’ although they overlap considerably.2

We have used the terminology adopted by authorssince to rephrase everything in terms of ‘basic re-search’ would risk distorting their arguments or con-clusions. The use of an overly strict definition ofwhat is meant by ‘basic research’ would needlesslyrestrict the scope of this review. Indeed, the reviewsuggests that simple definitions of research under-play the variety and heterogeneity of the links be-tween research and innovation. Research can havedifferent objectives depending on the perspective ofthe observer. It is more appropriate to think of thedifferent categories of research and development asoverlapping activities with gradual rather than sub-stantial differences.

The study focuses on the economic benefits frombasic research rather than the social, environmentalor cultural benefits. However, there is a fuzzyboundary between the economic and non-economicbenefits; for example, if a new medical treatmentimproves health and reduces the days of work lost toa particular illness, are the benefits economic orsocial? Given this uncertainty, we define ‘economic’quite broadly. Moreover, the study considers notonly economic benefits in the form of directly usefulknowledge but also other less direct economic bene-fits such as competencies, techniques, instruments,networks and the ability to solve complex problems.Although it may be extremely difficult to quantifythese benefits with precision, this does not mean theyare not real and substantial.

Lastly, the study concentrates on publicly fundedbasic research.3 This includes much of the basic

2 In the United States, for example, about two-thirds of theresearch in universities is classified as ‘basic’, although this variesconsiderably across disciplines. Most analyses therefore focus on

Žpublicly funded research in general. We are grateful to one of the.referees for this point.

3 The study’s scope was set by the UK Treasury who commis-sioned the work on which this article is based. It is also based onwork conducted in association with David Wolfe for The Partner-

Ž .ship Group on Science and Engineering PAGSE in CanadaŽ .Wolfe and Salter, 1997 . We are grateful to our co-authors inthese two projects.

( )A.J. Salter, B.R. MartinrResearch Policy 30 2001 509–532 511

research conducted in universities, government re-search institutes and hospitals. Again, however, theboundary is somewhat indistinct since some publicfunds go to support research that is conducted on thebasis of collaboration between universities and in-dustry. The focus on publicly funded research in thisreview does not imply that public research is sepa-rate or disconnected from private sector research.There is often considerable mutual interaction be-tween public and private research activities.4 In manyindustries, as we shall demonstrate, there is a divi-sion of labour between public and private activities.

3. Conceptual and methodological overview

3.1. The economics of publicly funded basic research

Many of the problems in assessing the benefits ofpublicly funded basic research stem from limitationsof the models used to evaluate those benefits. Underthe traditional justification for public funding ofresearch, government action serves to correct a‘market failure’. The concept of market failure,rooted in neo-classical economic theory, is based onthe assumption that a purely market relation wouldproduce the optimal situation and that governmentpolicy should be limited to redressing situationswhere market failures have developed. As MetcalfeŽ .1995, p. 4 notes, this is a daunting task for sciencepolicy-makers:

future markets for contingent claims in an uncer-tain world do not exist in any sense sufficientlyfor individuals to trade risks in an optimal fashionand establish prices which support the appropriatemarginal conditions. Because the appropriate pricestructure is missing, distortions abound and thepolicy problem is to identify and correct those

w xdistortions. Yet the innovation process both gen-erates and is influenced by uncertainty and thisaspect of market failure is particularly damagingto the possibility of Pareto efficient allocation of

4 Business-funded research also allows industry to build ontheir own research through absorbing and deriving benefits fromother research.

w xresources to invention and innovation . . . T husinnovation and Pareto optimality are fundamen-

Ž .tally incompatible ibid., p. 4 .

Metcalfe offers the evolutionary approach as analternative to justifying the case for governmentfunding of basic research. In evolutionary theory, thefocus of attention ceases to be Amarket failure per seand instead becomes the enhancement of competitiveperformance and the promotion of structural changeBŽ . 5ibid., p. 6 . The broader perspective afforded byevolutionary theory, with its focus on both the publicand private dimensions of the innovation system,

Žappears to offer a more promising approach Nelson,.1995 .

The traditional ‘market failure’ approach to theeconomics of publicly funded research centres on theimportant role of information in economic activity.

Ž .Drawing on the work of Arrow 1962 , it underlinesthe informational properties of scientific knowledge,arguing that this knowledge is non-rival and non-ex-cludable. Non-rival means that others can use theknowledge without detracting from the knowledge ofthe producers, and non-excludable means that otherfirms cannot be stopped from using the information.The main product from government-funded researchis thus seen to be economically useful information,freely available to all firms. In this context, scientificknowledge is seen as a public good. By increasingthe funds for basic research, government can expandthe pool of economically useful information. Thisinformation is also assumed to be durable and cost-less to use. Government funding overcomes the re-

Žluctance of firms to fund their own research to a.socially optimal extent because of their inability to

appropriate all the benefits. With government fund-ing, new economically useful information is createdand the distribution of this information enhancedthrough the tradition of public disclosure in science.

Relatively few economists today would supportthe purely informational approach. Yet in certaineconomic writing on the relationship between pub-licly funded research and economic growth, there

5 For an evolutionary perspective on science and technologyŽ . Ž . Ž .policy, see Lundvall 1992 , Nelson 1993 and Edquist 1997 .


remains a presumption of the informational proper-Ž .ties of basic research. For example, Adams 1990

has developed a series of industry measures of thestock of knowledge by looking at articles in aca-demic journals and the employment of scientists. Hefound a 20–30 year lag between scientific publica-

Ž .tion the knowledge stock and productivity growth.He suggested that the decline in the productivity ofscientists and the subsequent fall in the stock of

Ž .knowledge measured by total papers was related tothe Second World War and speculated that 15% ofthe economic slowdown in the 1970s could be ex-plained by this earlier decline in the knowledge stockŽ .ibid., p. 699 .

The evolutionary approach to the economics ofpublicly funded research suggests that the informa-tional view of knowledge substantially undervaluesthe extent to which knowledge is embodied in spe-cific researchers and the institutional networks withinwhich they conduct their research. It also misrepre-sents the nature of the innovation process, implyingthat scientific knowledge is Aon the shelf, costlessly

Ž .available to all comersB Rosenberg, 1990, p. 165 .Callon argues that scientific research is therefore nota public good because of the investment required tounderstand it. Scientific knowledge is not freelyavailable to all, but only to those who have the righteducational background and to members of the scien-tific and technological networks. The informationalview fails to appreciate the extent to which scientificor technical knowledge requires a substantial capa-bility on the part of the user. To paraphrase the

Ž .OECD 1996, p. 231 , knowledge and informationabound, it is the capacity to use them in meaningfulways that is in scarce supply. Often this capacity is

Žexpensive to acquire and maintain Pavitt, 1991,.1998 . In an influential study, Cohen and Levinthal

Ž .1989 suggest that one can characterise the internalR&D efforts of firms as having two faces: theirR&D both allows firms to create new knowledgeand enhances their ability to assimilate and exploitexternal knowledge.6 They refer to this second di-mension as the firm’s ‘absorptive capacity’.

6 In their paper, Cohen and Levinthal refer to AinformationBrather than AknowledgeB. We have replaced information withknowledge here for the sake of consistency with other discussion.

The newer approach based on evolutionary eco-nomics has generated two strands of research. Thefirst assumes that, despite the limitations of the oldapproach, publicly funded research can still be use-fully seen as yielding information. For example,

Ž .Dasgupta and David 1994 regard the informationalproperties of science as a powerful analytical tool forstudying the payoffs to publicly funded basic re-search. Drawing on information theory, they suggestit is possible to develop a Anew economics of sci-enceB. They focus on changes in the properties ofknowledge brought about by developments in infor-mation and communication technologies such as theInternet, arguing that these allow for an expansion ofthe informational or codified component of scientificknowledge. They call on policy makers to focus onexpanding the distributive power of the innovationsystem through new information resources such as

Želectronic libraries ibid.; see also David and Foray,.1995 .

The second strand in the new approach focuses onthe properties of knowledge not easily captured bythe information view described above. Influential

Ž . Ž .here are Rosenberg 1990 and Pavitt 1991, 1998 ,who stress that scientific and technological knowl-edge often remains tacit — i.e. people may knowmore than they can say.7 Moreover, the developmentof tacit knowledge requires an extensive learningprocess, being based on skills accumulated throughexperience and often years of effort. This perspectivestresses the learning properties of individuals andorganisations. Focusing on the learning capabilitiesgenerated by public investments in basic researchmakes it possible to apprehend the economic benefits

Ž .of such investments ibid., p. 117 . Of crucial impor-tance in this approach are skills, networks of re-searchers and the development of new capabilities onthe part of actors and institutions in the innovationsystem. The approach we follow here owes more tothis second strand of research. The information the-ory approach is still quite new and has yet to beempirically validated, whereas the RosenbergrPavittapproach is grounded in a growing body of science

7 Ž .Polanyi 1962 distinguished between the two dimensions ofknowledge — tacit and explicit. For an application of this concept

Ž .to innovation, see Nonaka and Takeuchi 1995 .


policy research and seems to offer a more productiveapproach to the issues under discussion.

3.2. Methodological approaches

In studies of the benefits of publicly funded scien-tific research, three main methodological approaches

Ž . Ž .have been adopted: 1 econometric studies; 2 sur-Ž .veys; and 3 case studies. Econometric studies focus

on large-scale patterns, and are effective in providingan aggregate picture of statistical regularities amongcountries and regions, and in estimating the rate ofreturn to research and development. The results can,however, be misleading. Econometric approaches in-volve simplistic and often unrealistic assumptionsabout the nature of innovation. It is also very diffi-cult to trace the benefits of the research componentof a new technology through the process of innova-tion and commercialisation.

Surveys have opened up a productive line ofresearch, analysing the extent to which government-funded research constitutes a source of innovativeideas for firms. Surveys have examined how differ-ent industries draw upon the supply of publiclyfunded research. They have helped us understand theways in which different industries utilise the researchresults of different scientific fields. Surveys never-theless suffer from several limitations. In particular,survey respondents from firms may have a biastowards the internal activities of their own compa-nies and rather limited knowledge of their sectorsand technologies.

Case studies afford the best tool to examine di-rectly the innovation process and the historical roots

Ž .of a particular technology Freeman, 1984 . Theygenerally provide support for the main findings fromeconometric studies and surveys. For example, theTRACES study by the National Science Foundationshowed the substantial influence of government-

Ž .funded research in key innovations NSF, 1969 .However, case studies are expensive to administer,can take a long time to analyse, and yield only anarrow picture of reality.

4. Relationship between publicly funded researchand economic growth

Econometricians have tried to calculate that por-tion of economic growth accounted for by technolog-

ical innovation in general, and by research in particu-lar. Efforts to assess the role of technology haveadopted the technique of ‘growth accounting’,analysing the contributions of production factors toeconomic development. Most growth models focuson the substitution of labour by capital, suggestingproductivity growth occurs through the steady re-placement of labour by fixed capital investments.Early growth models said little about technology, letalone the benefits of basic research. Solow and otherpioneers treated technological change largely as aresidual — as the portion of growth that could not

Žbe explained by labour and capital inputs e.g. Solow,.1957, Abramowitz, 1986 . Technical change was

deemed to be part of the general productivity in-crease and played no independent role in explaininggrowth.

Newer models in growth theory have attempted totake account of technology more directly, with

Ž .Romer’s 1990 contribution having spawned a newgeneration of research. Yet these models remainsomewhat simplistic in their treatment of technologyŽ .Verspagen, 1993 . They suggest that, by introducinga variable for ‘technical progress’, one can indirectlyaccount for the portion of growth created by techno-logical development. The models vary in their con-clusions but all suggest a key role is played bytechnology in generating economic developmentŽLucas, 1988; Grossman and Helpman, 1991, 1994;

.Romer, 1994; Aghion and Howitt, 1995 . However,they usually rely on simplified assumptions about theproperties of information or technology, such as itsdurability. As yet, no reliable indicator has beendeveloped of the benefits derived from publiclyfunded basic research. The models are more effective

Žin showing that technology however measured or.treated does play a substantial role in the growth ofŽ .firms Verspagen, 1993 .

Some attempts have been made to measure theeconomic impact of universities or publicly funded

Ž .R&D e.g. Bergman, 1990; Martin, 1998 . Thesestudies show a large, positive contribution of aca-demic research to economic growth. Yet, as GrilichesŽ .1995, p. 52 has stressed, the relationship betweentechnological change and economic growth remainsproblematic for economic research; it is difficult tofind reliable indicators of technological change andthere is the econometric problem of drawing infer-


ences from non-experimental data. Furthermore, asŽ .Nelson 1982, 1998 pointed out, these models do

not explain the link between publicly funded basicresearch and economic performance in a direct way;

Ž .they simply look at inputs such as papers andŽ .outputs firm sales without analysing the process

linking them. Nelson suggests that new growth the-ory models ought to treat technological advance as adis-equilibrium process. In order to gain a fullerappreciation of innovation, these models should in-corporate a theory of the firm, including differencesacross firms and in capabilities among firms. Newgrowth models also need to take into account therole of institutions such as universities in supporting

Ž .economic development Nelson, 1998

4.1. Measuring the social rate of return to inÕest-ments in basic research

Studies of the rate of return to research take twoforms. Some focus on the private rates of return —i.e. the return on investments in research that flowfrom an individual research project to the organisa-tion directly involved. Others examine the socialrates of return to research — i.e. on Athe benefits

Žwhich accrue to the whole societyB Smith, 1991, p..4 . The difference between the two arises because

the benefits of a specific research project, or even afirm-based innovation, generally do not accrue en-tirely to one firm. The scientific benefit of a basicresearch study may be appropriated by more thanone firm — for example, by imitators who replicatethe new product without bearing the cost of theoriginal research. By lowering the costs of develop-ing new technologies or products through investingin basic research, publicly funded projects generatebroader social benefits. Hence, estimates of the pri-vate rate of return to research and development tendto be much lower than those for the social rate ofreturn. This difference underscores the importance ofestimating the social rates of return for investmentsin scientific research, despite the severe methodolog-ical problems involved.

As Table 1 shows, estimates of private and socialrates of return to privately funded R&D are large,most of them falling in the range between 20% and

Ž .50%. In a review, Hall 1993 calculated that the

Table 1Estimates of private and social rates of return to private R&Dspending

Studies Private rate Social rateŽ . Ž .of return % of return %

Ž .Minnasian 1962 25 –Ž .Nadiri 1993 20–30 50

Ž .Mansfield 1977 25 56Ž .Terleckyj 1974 27 48–78

Ž .Sveikauskas 1981 10–23 50Ž .Goto and Suzuki 1989 26 80

Ž .Mohnen and Lepine 1988 56 28Ž .Bernstein and Nadiri 1988 9–27 10–160

Ž .Scherer 1982, 1984 29–43 64–147Ž .Bernstein and Nadiri 1991 14–28 20–110

Ž .Source: Griliches 1995, p. 72 .

gross rate of return on privately funded R&D in theUnited States is 33%. He also suggested that theprivate return to R&D is not as profitable as it oncewas and that there may be a decline in the effect ofscience on productivity. However, the use of firm-level R&D spending statistics in studies such asthese is a somewhat limited approach to understand-ing the economic benefits of investments in innova-

Žtion since many firms do no formal R&D Baldwin.and Da Pont, 1996 . More generally, R&D spending

is only a small portion of society’s investment inactivities that generate innovation. Many processinnovations involve ‘grubby and pedestrian’ incre-mental processes within the firm and are not cap-

Ž .tured by figures for R&D Rosenberg, 1982, p. 12 .Ž .Indeed, Dennison 1985 has suggested that R&D

accounts for only 20% of all technical progress.Studies that rely on R&D spending at the firm levelhave to be considered in the light of these limita-tions.

Until quite recently, few attempts had been madeto measure the rates of return to publicly fundedresearch and development. Most of these have fo-cused on government R&D projects rather than ba-sic research and they have not been very successful

Ž .or convincing OTA, 1986, p. 14 . Nevertheless, thelimited evidence gathered to date indicates that pub-licly funded basic research does have a large positivepayoff, although this is perhaps smaller than thesocial rate of return on private R&D — see Table 2.


Table 2Estimates of rates of return to publicly funded R&D

Studies Subject Rate of return toŽ .public R&D %

Ž .Griliches 1958 Hybrid corn 20–40Ž .Peterson 1967 Poultry 21–25

Schmitz-Seckler Tomato harvester 37–46Ž .1970

Ž .Griliches 1968 Agricultural research 35–40Ž .Evenson 1968 Agricultural research 28–47

Ž .Davis 1979 Agricultural research 37Ž .Evenson 1979 Agricultural research 45

Davis and Agricultural research 37Ž .Peterson 1981

Huffman and Agricultural research 43–67Ž .Evenson 1993

Ž . Ž .Source: Griliches 1995 and OTA 1986 . Many authors of thesestudies caution about the reliability of the numerical results ob-

Ž .tained cf. Link, 1982 .

The studies cited in Table 2 focus on relativelyAsuccessfulB government R&D programmes. Theyassume Ano alternative method could have generatedthe economic returns associated with the productsand processes attributed to the basic research in

w xquestion . . . Yet most economists would find thisassumption to be an uncomfortable one, inasmuch asthere are few new products and processes completely

Ž .lacking substitutesB David et al., 1992, p. 77 . Thecosts and benefits of government-funded R&D pro-jects need to be compared with those of alternative

Ž .solutions ibid. . Tracing the benefits of a particularproject involves looking retrospectively at a technol-ogy, and does not take into account investments incomplementary assets needed to bring the technol-

Ž .ogy to market Teece, 1986 . Consequently, the re-sulting return on research investment may underesti-mate the true costs of technological development.

Using industry-level productivity growth ratesas an indicator of the social rates of return to go-vernment-funded basic research is also problematic.Although studies based on this method have demon-strated a statistically significant impact for govern-ment-funded basic research on productivity growthat the sectoral level, most have been at a high degreeof aggregation, rarely controlling for inter-industrydifferences. AMoreover, they do not reveal how the

Žeconomic returns of basic research and develop-. w x Žment are actually realisedB David et al., 1992, p.

.79 . Other econometric studies have reached intrigu-Ž .ing conclusions. For example, Hall 1993 showed

that one impact of publicly funded basic researchŽmay be to increase a firm’s own R&D spending cf.

.Cohen and Levinthal, 1989 .Despite the above problems, Mansfield made sub-

stantial progress in measuring the benefits of basicresearch. He focused on ‘recent’ academic research— i.e. research within 15 years of the innovation

Ž .under consideration Mansfield, 1991 . Using a sam-ple of 76 US firms in seven industries, he obtainedestimates from company R&D managers about whatproportion of the firm’s products and processes overa 10-year period could not have been developedwithout the academic research. He found that 11% ofnew products and 9% of new processes could nothave been developed without a substantial delay inthe absence of the academic research, these account-ing for 3% and 1% of sales, respectively. He alsomeasured those products and processes developedwith ‘substantial aid’ from academic research overthe previous 15 years; 2.1% of sales for new prod-ucts and 1.6% of new processes would have beenlost in the absence of the academic research. Usingthese figures, Mansfield estimated the rate of return

Ž .from academic research to be 28% ibid., p. 10 .In 1998, Mansfield published the results of a

follow-up study. He found that academic work wasbecoming increasingly important for industrial activi-ties. On the basis of a second survey of 70 firms,Mansfield estimated that 15% of new products and11% of new processes could not have been devel-

Ž .oped without a substantial delay in the absence ofacademic research. In total, innovations that couldnot have developed without academic research ac-counted for 5% of total sales for the firms. Mans-field’s second study also suggests that the time delayfrom academic research to industrial practice hasshortened from 7 years to 6. Mansfield made noattempt in this paper, however, to estimate a rate ofreturn to academic research. He suggested that in-creasing links between academic research and indus-trial practice may be a result of a shift of academicwork toward more applied and short-term work andof growing efforts by universities to work moreclosely with industry.

Mansfield recognised the limitations of his ap-Ž .proach: the time lag 15 years is short; it is assumed


that no benefits accrue to firms outside the US andthat there are no indirect benefits from research, suchas skilled researchers; the estimates rely on the opin-ions of managers in large firms; and they do not

Žconsider the full costs of commercialisation CBO,.1993, p. 15 . Moreover, the approach yields only the

average rate of return, not the marginal rate, so itcannot inform policy makers about the marginal

Žbenefits of additional research funding OTA, 1986,. 8p. 4; David et al., 1992, p. 79 . Mansfield’s figures

are also hard to compare with other data on rates ofreturn on investments. If the benefits are so great,why do governments and firms not invest more inresearch? The lack of investment might be related tothe riskiness of R&D. If so, these estimates cannotbe compared directly with other figures on rates of

Ž .return e.g. on capital equipment .Ž .Beise and Stahl 1999 have replicated Mansfield’s

survey in Germany with a much larger sample of2300 manufacturing firms. They found that approxi-mately 5% of new product sales could not havedeveloped without academic research. They alsoshowed that academic research has a greater impacton new products than new processes and that smallfirms are less likely to draw from universities than

Ž .large firms ibid., p. 409 . This study shares many ofthe difficulties of Mansfield’s early study and, unlikeMansfield, does not take into account sectoral differ-ences in the importance of academic research toindustrial innovation.

Ž .Narin et al. 1997 have developed a new ap-proach to evaluating the benefits of publicly fundedresearch based on analysing scientific publicationscited in US patents. Examining the front pages of400,000 US patents issued in 1987–1994, they tracedthe 430,000 non-patent citations contained in thesepatents, of which 175,000 were to papers publishedin the 4000 journals covered by the Science Citation

8 In a review of Mansfield’s work, the Congressional BudgetOffice noted that his findings could not guide policy makers onthe allocation of funds nor be used to determine the amount of

Ž .funding to devote to R&D CBO, 1993 . This did not stop theBush Administration from citing Mansfield’s work as a justifica-tion for an increase in basic research funding.

Ž .Index SCI . For 42,000 papers with at least one USauthor, they determined the sources of US and for-eign research support acknowledged in the papers.Their findings on the increasing number of scientificreferences cited in patents suggest that over a periodof 6 years there has been a tripling in the knowledgeflow from US science to US industry. US govern-ment agencies were frequently listed as sources offunding for the research cited in the patents. Narin etal. suggest that this indicates a strong reliance by USindustry on the results from publicly funded researchŽ .ibid. .

One possible methodological limitation of thiswork is that it focuses on the citations to the scien-tific literature made by the patent examiner ratherthan those made by the applicant. The three-foldincrease of scientific citations in US patents maypartly reflect a policy at the US Patent Office9 topromote scientific citations, changes in patent law, orsimply the availability of relevant data from newCD-ROMs listing academic papers by subject. Itseems surprising that there could have been such adramatic shift in the relationship between US indus-try and science over a period of just 6 years.

Ž .As noted by David et al. 1992 , measuring theeconomic benefits to basic research is complicatedby industry differences. A summary table developed

Ž .by Marsili 1999 illustrates the patterns within anddifferences across industries in the relationship be-tween academic research and industrial innovation.Table 3 is based on a statistical analysis of the Pace

Žsurvey of European industrial managers Arundel et.al., 1995 , US R&D data, employment patterns in

different industries, and patent citations.10 Using thePace survey, Marsili classified industries in terms ofthe contribution of academic research to innovationin each sector from very high to low. The underlyingscientific knowledge that industries draw upon intheir innovation activities was also described usingPace survey data.

9 Patents issued by the European Patent Office do not appar-ently exhibit the same dramatic increase in the number of scien-tific references.

10 Ž .A similar table is produced in Godin 1996 .


Table 3The role of academic research in different industries

Contribution of Development activities Research-based activitiesŽ . Ž .academic research engineering disciplines mainly tacit basic and applied science mainly codified

Very high Computers PharmaceuticalsHigh Aerospace Petroleum

Motor vehicles ChemicalsTelecommunications and electronics FoodElectrical equipment

Medium Instruments Basic metalsNon electrical machinery Building materials

Low Metal products TextilesRubber and plastic products Paper

aRelevant scientific fields Mathematics, computer science, mechanical and Biology, chemistry, chemical engineeringelectrical engineering

Ž .Source: adapted from Marsili 1999 .a Physics is important for both research and development activities. In the statistical analysis, physics was not highly significant in

discriminating between the two groups and therefore it has not been included in the table.

Using US R&D data, Marsili estimated the per-centage of research undertaken in each industry whichis basic, applied and development in orientation.These results were compared with data on employ-

Žment patterns of technical personnel e.g. scientists,.engineers and technicians across different industries.

As one might expect, the distribution of R&D iscorrelated with the distribution of employment —for example, industries with high levels of basicresearch employ large numbers of scientists. MarsiliŽ .1999 also analysed the degree of codification in theknowledge base of each industry, using the numberof academic papers cited in patents as a measure of

Ž .that codification cf. Narin et al., 1997 . The resultsindicate that firms and industries draw from publiclyfunded science in a heterogeneous fashion. In somesectors, the link is quite direct, with numerous cita-tions to scientific papers in patents and a closeinterest in scientific research. In other sectors, suchas automobiles, firms draw from the public basemore indirectly, mostly through the flow of studentswho help the firm overcome technological chal-lenges. These differences in the ways in which indi-vidual sectors derive their benefits suggest that anyattempt at a simple calculus of the benefits of gov-ernment-funded basic research is likely to be mis-leading.

Ž .As Meyer-Krahmer and Schmoch 1998, p.837suggest, Aa weak science linkage of a technologydoes not imply low university–industry interactionB.

Using a combination of European Patent Office dataand a survey of universities on their linkages withindustry, they show that there is a Atwo-wayB inter-action between universities and industry. Collabora-tive research and informal contacts are the mostimportant forms of interaction between universitiesand industry. Academic researchers gain funding,knowledge and flexibility through industrial funding.Collaborative research between universities and in-dustry almost always involves a two-directional flowof knowledge and informal discussion is preferred topublications for contacts. The strength ofuniversity–industrial interactions is dependent on the‘absorptive capacity’ of the industry and the innova-

Ž .tion system ibid.; see also Schmoch, 1997 . Meyer-Krahmer and Schmoch’s findings show that it isalmost impossible to measure the extent to which asector like automobiles gains economic benefits fromthe publicly funded research infrastructure. Only inpharmaceuticals, where the links are direct and oftenvisible, might some measurement of the benefits befeasible.

4.2. SpilloÕers and localisation

One prominent line of recent research into thebenefits of publicly funded research has been theinvestigation of the spillovers from governmentfunding to other activities such as industrial R&D.The existence of these spillovers augments the pro-


ductivity of a firm or industry by expanding thegeneral pool of knowledge available to it. Two main

Ž .forms of spillover have been identified: 1 geo-Ž .graphical spillovers and 2 spillovers across sectors

Ž .and industries Griliches, 1995 . The former implybenefits for firms located near research centres, otherfirms and universities. Evidence from bibliometricstudies indicates a strong tendency for basic research

Ž .to be localised. Katz 1994 has shown that researchcollaboration within a country is strongly influencedby geographical proximity; as distance increases,collaboration decreases, suggesting that research col-laboration often demands face-to-face interaction.

Ž .Hicks et al. 1996 also found that research acrosscountries is localised.

Jaffe has attempted to measure geographicalspillovers in the US employing a three-equation

Žmodel involving patenting, industrial R&D and uni-.versity research . Using patents as a proxy for inno-

vative output, he examined the relationship betweenpatents assigned to corporations in 29 US states in1972–1977, 1979 and 1981, industrial R&D anduniversity research. His results demonstrate that thereare spillovers from university research and industrialpatenting. There is also an association between in-dustrial R&D and university research at the statelevel. It appears that university research encourages

Ž .industrial R&D, but not vice versa Jaffe, 1989 . InŽ .a similar study, Acs et al. 1991 found that the

spillovers between university research and innova-tion are greater than Jaffe described.11 Anselin et al.Ž .1997 also observed significant spillovers from uni-versity research and ‘high technology’ innovations atthe level of metropolitan units or cities. Feldmann

Ž .and Florida 1994 developed a four-variable modelŽbased on distribution of university research, indus-trial R&D expenditures, distribution of manufactur-

.ing, and distribution of producer services to test for

11 Acs et al. used a database of innovations prepared by the USSmall Business Administration in 1982. The database contains

Ž .innovations reported in the literature for one year 1982 brokendown by city and state. Such databases are inherently subjective,relying on innovations cited in technical journals. The databasefocuses on a limited number of product innovations for a singleyear. The date of the database collection also raises questionsabout the reliability of the findings given the changes in theeconomy over the past 17 years.

geographical effects. Using the same data as Acs,they showed that geography does matter in the pro-cess of innovation, with the variables being highlycorrelated.12 These findings are supported by the

Ž .work of Mansfield and Lee 1996 who found thatfirms close to major centres of academic researchhave a major advantage over those located at adistance:13

While economists and others sometimes assumethat new knowledge is a public good that quicklyand cheaply becomes available to all, this is farfrom true. According to executives from our sam-

w xple of 70 major US companies , firms located inthe nation and area where academic research oc-curs are significantly more likely than distantfirms to have an opportunity to be among the first

Žto apply the findings of this research ibid., p..1057 .

Ž .Similarly, Hicks and Olivastro 1998 have shownthat US company patents tend to cite papers pro-duced by local public-sector institutions, with over27% of ‘state-of-the-art’ references in patents beingto institutions within the US state in which the patentwas taken out. They suggest that Apapers andpatents . . . were written precisely to make explicit . . .

Ž .complex, tacit knowledgeB ibid., p. 4 . There is alsoevidence for geographical effects at the national

Ž .level, with Narin et al. 1997 finding national pat-terns in the public research cited in industrial patents.For example, patents taken out by German firms inthe US are 2.4 times more likely to cite Germanpublic scientists among their scientific referencesthan other nationalities, and similar results are ob-tained for other major countries.

However, these geographical effects are not nec-Ž .essarily universal. Beise and Stahl 1999 found that,

while firms in Germany tend to cite local public

12AIn the modern economy, locational advantage in the capacity

to innovate is ever more dependent on the agglomerations ofspecialised skills, knowledge, institutions, and resources that make

Ž . Žup the underlying technological infrastructure of a place B Feld-.mann and Florida, 1994, p. 12 .

13 Among the limitations of this study are that it was based on arelatively narrow sample of firms and that it only asked industrial-ists to list the five most important academics for their firm’sactivities.


institutions, especially polytechnics, there were nosignificant differences between firms with innova-tions drawing upon public research and all otherfirms in the distribution of distances of academicscientists cited by the firms they surveyed. Theysuggest that this finding indicates that it is Ahard tobelieve that closeness to research institutions has aneffect on the probability of public research-based

Ž . 14innovationsB ibid., p. 411 . More important thandistance, they argue, is the willingness of firms toinvest in in-house R&D. For polytechnics and smallfirms distance still matters, but for large firms anduniversities distance appears to be much less impor-tant.

Recent work in economic geography also stressesthe importance of geographical agglomerations and

Ž .spillovers. Saxenian’s 1994 study of Silicon ValleyŽand Route 128 concluded that local institutions in-

.cluding the research infrastructure profoundly shapeŽ .a region’s capacity to innovate. Storper 1995, 1997

suggested that the development of geographical ag-glomerations is a result of the person-embodied na-ture of much technological knowledge and the con-sequent importance of face-to-face interactions. Sincethese personal interactions are essential to deal withthe uncertainty inherent in the future development oftechnologies or markets, firms and individuals tendto cluster. Their interactions are often untraded andthis helps to create a social environment that allowsand indeed encourages individuals to share knowl-edge and ideas. The consequent interdependenciesare place-specific and context-dependent, resultingfrom continuous interactions among firms and indi-viduals as they go about developing technology and

Žsolving common problems Dosi et al., 1988; Stor-.per, 1995, 1997; Cooke and Morgan, 1998 .

The value of geographic spillovers and untradedinterdependencies varies over time. They may beparticularly important when the technological trajec-

Ž .tories Dosi, 1982 are highly indeterminate — inother words, when a wide range of possible paths ofdevelopment increases the importance of tacit

14 Some of these differences in findings are probably linked tothe considerable geographical differences between Europe and the

Ž .United States. We are grateful to a referee for pointing this out.

knowledge to the innovation process, thus raising thevalue of direct interactions in interpreting and apply-ing new information. These untraded interdependen-cies form the collective property of the region andhelp the regional actors expand their range of activi-

Žties, drawing one another forward cf. Lundvall,.1988 . All this suggests that each nation or region

needs to maintain its own capability in research anddevelopment. Personal links and face-to-face interac-tions are essential not only for the research processbut also for sharing and transferring knowledgequickly and effectively. Policies designed to supportgeographical agglomeration should help facilitate this

Ž .interaction Wolfe, 1996 .Spillovers are also common among research-re-

lated activities: Athe level of productivity achievedby one firm or industry depends not only on its ownresearch efforts but also on the general pool of

Ž .knowledge accessible to itB Griliches, 1995, p. 63 .Ž .Using US patent data, Scherer 1982 constructed

development measures of the direction of spilloversby classifying a large sample of patents in terms ofthe industry where the innovation occurred and ofthe industry where it was expected to have an im-

Ž .pact. Los and Verspagen 1996 have expanded thetreatment of spillovers, looking at the locationalorigin of patents and papers cited in US patents todetermine the degree of spillover of domestic sourcesof science and technology. They found that spilloversdo exist but they vary across sectors and countries.15

Work by economists on new growth theory high-lights the spillover effects of technological develop-ment. Indeed, growth theorists tend to see spilloversas the main mechanism underlying growth patternsŽ . 16Romer, 1994, Grossman and Helpman, 1994 .These models suggest that the encouragement ofspillovers through government institutions may be

Ž .fruitful from a policy perspective Romer, 1990 .

15 Los and Verspagen’s approach faces similar methodologicalŽ .problems to that of Narin et al. 1997 .

16 The case for spillovers is strong but needs to be tempered byan understanding of the importance of firm-level dynamics. As

Ž .Rosenberg 1990 has emphasised, technological developmenttakes place within the firm; external relations among firms, suchas spillovers, help to constitute these internal firm dynamics.


These models do, however, rely largely on the theo-retical elaboration of production functions and makelimited use of empirical data. Most models alsofocus on industrial R&D rather than publicly fundedbasic research. These models show that knowledgeand technologies spill over across sectors and fieldsbut it is difficult to develop useful measures of theextent of these spillovers. Often the links betweengovernment-funded basic research and productionare varied and indirect. Simple measures such assales or cross-patent citations only capture thesespillovers to a limited degree.

5. The main types of benefit from publicly fundedresearch

Despite the methodological problems discussedabove in estimating the economic returns to publicinvestment in basic research, one can distinguishvarious types of contributions that publicly funded

Žresearch makes to economic growth Martin et al.,.1996 :

1. increasing the stock of useful knowledge;2. training skilled graduates;3. creating new scientific instrumentation and

methodologies;4. forming networks and stimulating social inter-

action;5. increasing the capacity for scientific and tech-

nological problem-solving;6. creating new firms.

Although these six categories of benefits areclearly interrelated and overlap,17 it is useful toseparate them analytically. In what follows, we drawupon recent science policy research to analyse the

17 For example, the category of ‘increasing the capacity forscientific and technological problem-solving’ is obviously quiteclosely related to that of ‘training skilled graduates’. However, wehave chosen to separate them analytically here partly because the

Žtwo categories are not identical problem-solving may also draw.upon knowledge, methodologies and networks, for example and

partly because of the emphasis given to the problem-solvingcomponent by industrialists when surveyed about the benefits ofbasic research.

benefits that flow from government funding of basicresearch in each category. It should be emphasisedthat these six types of benefits are not limited topublicly funded basic research; privately funded ba-sic research can yield similar benefits.

5.1. Increasing the stock of knowledge

The traditional justification for public funding ofbasic research is that it expands the scientific infor-mation available for firms to draw upon in theirtechnological activities. However, this underplays

Ž .the substantial efforts and associated costs requiredfrom users to exploit such information. The difficultywith the information theory of basic research is thatthe commercial value of scientific findings is notalways immediately evident. An authoritative reviewŽ .OTA, 1995, p. 43 noted numerous examples ofscientific advances whose commercial applicationcould not fully be conceived of at the time of their

Ž .discovery e.g. lasers . Yet, despite the difficulties intracing the path from scientific discovery to practicalapplication, firms apparently rely quite heavily onpublicly funded research as a source of new ideas or

Ž .technological knowledge Narin et al., 1997 . Publicand private research systems tend to complementeach other. The two systems are interlinked by com-mon interests, institutional affiliations and personalconnections. As Meyer-Krahmer and SchmochŽ .1998 suggest, there is ‘two-way interaction’ be-tween public and private knowledge generation anddiffusion.

Ž .Nelson and Rosenberg 1994, p. 341 argue thatpublicly funded basic research often stimulates andenhances the power of R&D done in industry, ratherthan providing a substitute for it. Klevorick et al.Ž .1995 suggest that one can think of governmentfunding for basic research as expanding the techno-logical opportunities available to society. They usethe analogy of firms drawing balls from an urn in theprocess of technological development. Governmentfunding for scientific research adds more balls to theurn, thus increasing the chances for firms to draw out

Ž .a winner. Mowery 1995, p. 521 argues that theinformational outputs of basic research can Aofferrules for empirical generalisation from specific indi-cations that can improve the efficiency of technology


Ž .developmentB see also Dasgupta and David, 1994 .Ž .Steinmueller 1994, p. 59 suggests that scientific

knowledge may affect the ‘options value’ of applieddevelopment by Areducing the expected returns ofsome lines of applied research . . . or increasing thereal returns in other areasB.

Much confusion exists in the literature over whatit is that firms draw from public sources — ‘infor-mation’ or ‘knowledge’. In many innovation sur-veys, these terms are used interchangeably and formost firms the distinction between the two is some-what academic. However, the difference is importantfor understanding the role played by basic research.The traditional justification for government-fundedbasic research relies on the ‘public good’ qualities ofinformation. Yet the evidence from science policystudies indicates that what firms draw upon is notinformation but knowledge. To understand informa-tion almost always requires knowledge. Individualsand organisations need a complex set of skills andmust expend significant resources to absorb andunderstand information. Without these investments,firms would be unable to use the information avail-

Žable. Information only becomes knowledge and.therefore valuable when users have the capabilities

to make sense of it; without these, information isŽ .meaningless Nightingale, 1997; Pavitt, 1998 .

Here, it important to distinguish between codifiedŽ .and tacit knowledge. Faulkner and Senker 1995

argue that codified knowledge on its own is capableof providing only limited information since the ap-plication of such knowledge requires other tacitknowledge and more personal interaction. A related

Ž .finding is that of Hicks and Katz 1997 who haveshown that firms are publishing an increasing num-

ber of scientific journal articles, apparently under-mining the informational view of economic benefitsaccording to which firms should be reluctant tocodify their knowledge for appropriability reasons.

Ž .Hicks 1995 suggests that firms use their publishedpapers not only to present codified knowledge butalso to signal to others the presence of tacit knowl-edge and expertise. Likewise, publications by otherscan be interpreted as an indication of the existence of

Ž .relevant tacit knowledge Godin, 1996 .Ž .The Pace survey Arundel et al., 1995 of large

European firms in 16 industrial sectors shows thatpublications remain the most common source forcompanies to learn about public research — seeTable 4. This and other evidence on the ways firmsuse publicly available knowledge stocks suggeststhat government funding provides an important meansfor expanding the technological opportunities opento firms. At the same time, firms must invest sub-stantial resources in acquiring and using this infor-mation since publications on their own rarely containeconomically useful information. Government fund-ing for science, to the extent that it leads to publica-tions, therefore helps to identify relevant tacit knowl-edge. Open publication is consequently an essentialaspect of publicly funded science. Publications ex-pand the opportunities for firms to access the knowl-edge and skills base in the scientific communitycreated by public investment in basic researchŽ .Dasgupta and David, 1994 .

This argument differs from the old information-based view of the benefits of publicly funded re-search. Firms use publications to network, to developcontacts and to signal expertise. Codified and tacit

Ž .knowledge are inextricably linked Hicks, 1995 .

Table 4Importance of different sources for learning about public research

Ž .Source % rating as important High scoring industries scores in percentages

Ž . Ž . Ž . Ž .Publications 58 Pharmaceuticals 90 , basic metals 64 , GCC 62 , utilities 61Ž . Ž . Ž . Ž .Informal contacts 52 Pharmaceuticals 88 , GCC 68 , utilities 67 , aerospace 60Ž . Ž . Ž . Ž .Hiring 44 Pharmaceuticals 85 , computers 56 , aerospace 52 , chemical 48Ž . Ž . Ž . Ž .Conferences 44 Pharmaceuticals 85 , utilities 56 , computers 56 , telecom 48

Ž . Ž . Ž . Ž .Joint research 40 Aerospace 70 , basic metals 68 , utilities 67 , pharmaceuticals 51Ž . Ž . Ž . Ž .Contract research 36 Utilities 72 , pharmaceuticals 51 , basic metals 48 , plastics 46

Ž . Ž . Ž . Ž .Temporary exchanges 14 Pharmaceuticals 27 , computers 22 , electrical 20 , basic metals 20

Ž .Source: Arundel et al. 1995 . 640 respondents rated the importance of each source on a 7-point scale. The figures indicate the percentage ofrespondents rating each source 5 or higher on that scale.


Accessing and using codified knowledge are costly,time-consuming and difficult, yet frequently hold thekey to the development of innovations by the firm.The publications resulting from government-fundedresearch make these processes easier.

5.2. Skilled graduates

Many studies of the economic benefits of publiclyfunded research identify skilled graduates as theprimary benefit that flows to firms. New graduatesentering industry bring not only a knowledge ofrecent scientific research but also an ability to solvecomplex problems, perform research and developideas. The skills developed during their educationwith advanced instrumentation and techniques may

Ž .be especially valuable. Gibbons and Johnston 1974 ,in their analysis of the role of science in technologi-cal innovation, pointed to trained students as a keybenefit from research funding. Martin and IrvineŽ .1981 showed that even students from very basicfields such as astronomy may move into industry and

Ž .make major contributions. Nelson 1987 noted thatacademics may teach what industrial actors need toknow without necessarily doing ‘relevant’ researchfor industry. For example, basic research techniquesare often essential for young scientists to learn toparticipate in the industrial activities within the firm.

Ž .As Senker 1995 pointed out, graduates bring toindustry an ‘attitude of the mind’ and a ‘tacit ability’to acquire and use knowledge in new and powerfulways.

Even in applied areas of science and engineering,however, the transfer of students into industry israrely a smooth process. Often firms have to makelarge investments in training new graduates. Studentsmay come prepared to learn but they need to betaught industrial practice before they can be used byfirms to expand their technological competencies.Yet recent graduates bring both enthusiasm and acritical approach that stimulates others and raisesstandards. Moreover, the skills acquired during edu-cation are often a necessary precursor to the develop-ment of more industry-specific skills and knowledge.

Since graduates provide a key mechanism for thebenefits of public funding to be transferred to indus-try, it is vital that government-funded basic researchand student training are conducted in the same insti-

tution. There are also some benefits to be derivedfrom training students in research institutes engaged

Ž .in industry-related research Gibbons et al., 1994 .Such students gain practical experience of firms’needs and competencies. At the same time, one hasto strike a balance between developing an under-standing of industrial practice and providing an edu-cation that equips students with more generic andlong-lasting skills.

5.3. New instrumentation and methodologies

Ž .Historical research e.g. Rosenberg, 1992 hasshown that the development of new instrumentationand methodologies is a key output of government-funded basic research. However, few attempts havebeen made to evaluate this form of benefit. In partic-ular, innovation surveys rarely consider instrumenta-tion because of the limited ability of industrial R&Dmanagers to recognise the contribution made byearlier government-funded research.

The challenges involved in basic research contin-ually force researchers to design new equipment,laboratory techniques and analytical methods totackle specific research problems. Some of thesemay eventually be adopted in industry. Examplesinclude electron diffraction, the scanning electronmicroscope, ion implantation, synchrotron radiationsources, phase-shifted lithography, and supercon-

Ž .ducting magnets OTA, 1995, p. 38 . As RosenbergŽ .1992 has noted, scientific instruments have becomealmost indistinguishable from industrial capital goodsin many industries such as semiconductors: AIndeedmuch, perhaps most, of the equipment that one seestoday in an up-to-date electronics manufacturing planthad its origin in the university research laboratoryBŽ .ibid., p. 384 .

There are strong feedbacks between instrumentsdeveloped in the course of research and those devel-

Ž .oped by firms Von Hippel, 1987; Schrader, 1991 .Not only is new instrumentation drawn upon byfirms as scientists take advantage of the new tools toexpand their research; often the introduction of newinstrumentation can then lead to the development ofa new field of research such as computational physics

Ž .or artificial intelligence. Rosenberg 1992 arguesthat such instrumentation would generally not have


been developed without government funding allow-ing researchers to probe fundamental questions.

Ž .Work by Nelson et al. 1996 on the licensing oftechnologies at Columbia University indicates thatfirms tend to license mainly research tools and tech-niques from the university — in other words, muchof the ‘technological output’ of the university system

Žapparently lies in instrumentation and techniques cf.. ŽRosenberg, 1992 . Likewise, the Pace Report Arun-.del et al., 1995 shows that firms rate instrumenta-

tion as the second most important output of publicresearch, particularly firms in the pharmaceuticals,glass, ceramics and cement, electrical and aerospaceindustries — see Table 5.

5.4. Networks and social interaction

Ž .De Solla Price 1984 was one of the earliest toidentify the role of government-funded research inproviding an entry point into networks of expertiseand practice. Government funding affords individu-als and organisations the means to participate in theworld-wide community of research and technologicaldevelopment. The economic benefits of such net-works are difficult to measure. Nevertheless, the

Ž .evidence e.g. in Table 4 suggests that firms do findinformal methods of interaction an important meansof learning about public research and technologicalactivity.

Networks have been the focus of much empiricalresearch. This work indicates that firms and indus-tries link with the publicly funded science base inmany different ways and these links are often infor-

Ž .mal. Faulkner and Senker 1995 studied public–private sector linkages in biotechnology, engineeringceramics and parallel computing. Good personal rela-

tionships between firms and public-sector scientistsemerged as the key to successful collaboration be-tween the public and private sectors. Personal rela-tions develop the understanding and trust on whichlong-term contractual relationships can then be built.Networks may also provide access to sophisticatedinstrumentation and equipment for small firms.

Ž .Rappa and Debackere 1992 suggested that thereare technological communities in which loosely cou-pled individuals converge toward a common goal.Using a survey of the neural network community,they showed that, although academics are more will-ing to publish their findings and communicate newdiscoveries in general, industrial actors also seek anexchange of ideas within their technological commu-

Ž .nities ibid. . In other words, industrial actors areembedded in strong and often informal technologicalcommunities drawing on network relations with aca-demic researchers as they develop new products andprocesses.

Ž .Callon 1994 has suggested that governmentfunding for basic research should be seen as a meansof establishing new networks. Funding stimulatesnew combinations of relations between organisationsand individuals, creating new forms of interaction.The market, by contrast, tends to Ause upB theexisting sources of variety, leading to convergenceand irreversibility and locking society into particulartechnological options. Government action is requiredto break this cycle, to create new options and therebyto counter these market tendencies to exhaust theexisting stock of ideas and relations. Through gov-ernment funding, it is possible to create novel ap-proaches to addressing and resolving technical prob-lems by increasing the variety of scientific optionsavailable to firms. In short, government funding

Table 5Importance to industry of different outputs of public research

Ž .Form of output % rating as important High scoring industries scores in percentages

Ž . Ž . Ž . Ž .Specialised knowledge 56 Pharmaceuticals 84 , utilities 64 , food 57 , aerospace 57Ž . Ž . Ž . Ž .Instrumentation 35 Pharmaceuticals 49 , GCC 45 , electrical 42 , aerospace 39Ž . Ž . Ž . Ž .General knowledge from basic research 32 Pharmaceuticals 76 , chemical 38 , computers 38 , instruments 36

Ž . Ž . Ž . Ž .Prototypes 19 Food 28 , pharmaceutical 27 , electrical 26 , basic metals 24

Ž .Source: Arundel et al. 1995 . Respondents rated the importance of each output on a 7-point scale. The figures indicate the percentage ofrespondents rating each output 5 or higher.


expands social interaction, creating new links be-tween social actors; those links then add to the poolof technological opportunities available to firmsŽ .ibid. .

Ž .Lundvall 1992 has also stressed the need forgovernment funding to generate new forms of inter-actions among actors in the national innovation sys-tem. He suggests that the interactive nature of thelearning process that characterises innovation re-quires support from institutions which facilitate con-tact between actors within the system. Increasingly,knowledge and intelligence are organised in socialways, rather than being accessed on an individualbasis. The capacity for networking is essential for

tapping into the intelligence of others. The networkmodel recognises the growing relevance of the tacitdimension of knowledge and the extent to which thisis often grounded in the informal sharing of knowl-edge and ideas among firms and other relevant insti-tutions such as universities. The very density of thesenetworks and the links that comprise them is perhapsan indicator of the vibrancy of a regional or national

Ž .economy Cooke and Morgan, 1993, p. 562 .

5.5. Technological problem-solÕing

Publicly funded research also contributes to theeconomy by helping industry and others to solve

Table 6Relevance of university research and knowledge of science to technology

Scientific field No. of industries rating university No. of industries rating knowledge ofresearch as important and high science as important and high scoringscoring industries industries

Biology 12 Animal feed, drugs, processed 14 Drugs, pesticides, meat products,food animal feed

Chemistry 19 Animal feed, meat products, 74 Pesticides, fertilisers, glass,drugs plastics

Geology 0 4 Fertilisers, pottery, non-ferrousmetals

Mathematics 5 Optical instruments 30 Optical instruments, machinetools, motor vehicles

Physics 4 Optical instruments, electron 44 Semiconductors, computers,tubes guided missiles

Agricultural science 17 Pesticides, animal feed, 16 Pesticides, animal feed, fertilisers,fertilisers, food prod. food prod.

Applied mathrO.R. 16 Meat products, loggingrsawmills 32 Guided missiles, aluminiumsmelting, motor vehicles

Computer science 34 Optical instruments, 79 Guided missiles, semiconductors,loggingrsawmills motor vehicles

Materials science 29 Synthetic rubber, non-ferrous 99 Primary metals, ball bearings,metals aircraft engines

Medical science 7 Surgicalrmedical instruments, 8 Asbestos, drugs, surgicalrdrugs, coffee medical instruments

Metallurgy 21 Non-ferrous metals, fabricated 60 Primary metals, aircraft engines,metal products ball bearings

Chemical engineering 19 Canned foods, fertilisers, malt NrAbeverages

Electrical engineering 22 Semiconductors, scientific NrAinstruments

Mechanical engineering 28 Hand tools, specialised NrAindustrial machinery

Ž .Source: Klevorick et al. 1995 . Respondents rated the importance of each scientific field on a 7-point scale. The figures indicate the numberŽ .of industries out of a total of 130 rating each field 5 or more on the scale.


complex problems. Many firms in technologicallydemanding industries need to combine a variety oftechnologies in complex ways, and publicly sup-ported research provides an extensive pool of re-

Žsources from which these firms may draw Patel and. Ž .Pavitt, 1995 . Vincenti’s 1990 analysis suggests

that advanced engineering benefits indirectly fromthe publicly supported research base through theprovision of trained problem-solvers as well as thebackground supply of knowledge. Pertinent here isthe Yale survey of 650 R&D directors in large USfirms that provides one of the most systematic analy-ses of the benefits of basic research. The authorsdistinguished between the role of science in provid-ing a general pool of knowledge and the role ofspecific university research for firms. The formerwas judged to be important to recent technologicaladvance by approximately three times as many re-spondents as the latter — see Table 6.

The discrepancy between the measured relevanceŽ .of generic science a pool of knowledge and that

Ž .of university science new results is greater forbasic than for applied research because researchin applied sciences and engineering disciplines isguided to a large extent by perceptions of practi-cal problems, and new findings often feed directly

w xinto their solutions . . . T his by no means impliesthat new findings in fundamental physics, forexample, are not relevant to industrial innovation.

Rather, we read our findings as indicating thatadvances in fundamental scientific knowledgehave their influence on industrial R&D largelythrough two routes. One . . . is through influencingthe general understandings and techniques thatindustrial scientists and engineers, particularlythose whose industrial training is recent, bring totheir jobs. The other is through their incorporationin the applied sciences and engineering disciplinesand their influence on research in those fields’Ž .Klevorick et al., 1995, pp. 196–97 .

The Pace survey of European companies confirmsthese conclusions although the links between indus-tries and sciences vary by sector and country. As inthe Yale study, respondents were asked to rate differ-ent fields of science in terms of their importance totheir firms’ technological base. Applied areas ofresearch received fairly high scores as did chemistry,while physics, biology and mathematics were judgedless important — see Table 7.

Ž .Nelson and Rosenberg 1994 argue that the lowscores accorded in innovation surveys to fundamen-tal sciences such as physics do not necessarily implythat these fields fail to provide benefits to industry.They suggest that insights from basic research oftentrickle down to industry via engineering schools,which draw upon fundamental sciences to developtechnical knowledge for engineering and design.There is a strong feedback between engineering

Table 7Importance to technology base of publicly funded research in past 10 years

Ž .Scientific field % of respondents High scoring industries scores in percentagesrating as important

Ž . Ž . Ž . Ž .Material sciences 47 Aerospace 77 , basic metals 76 , electrical 72 , GCC 63Ž . Ž . Ž . Ž .Computer sciences 34 Aerospace 60 , telecom 56 , automotive 47 , computers 47Ž . Ž . Ž . Ž .Mechanical engineering 34 Automotive 64 , aerospace 64 , utilities 53 , computers 47Ž . Ž . Ž . Ž .Electrical engineering 33 Computers 78 , aerospace 73 , telecom 70 , electrical 56

Ž . Ž . Ž . Ž .Chemistry 29 Pharmaceuticals 78 , petroleum 52 , chemical 46 , computers 33Ž . Ž . Ž . Ž .Chemical engineering 29 Petroleum 60 , pharmaceuticals 55 , chemical 46 , plastics 42Ž . Ž . Ž . Ž .Physics 19 Computers 64 , basic metals 33 , plastics 25 , GCC 25

Ž . Ž . Ž . Ž .Biology 18 Pharmaceuticals 71 , food 33 , petroleum 18 , chemical 17Ž . Ž . Ž . Ž .Medical 15 Pharmaceuticals 85 , instruments 29 , computers 27 , food 15

Ž . Ž . Ž . Ž .Mathematics 9 Computers 25 , aerospace 20 , automotive 20 , GCC 13

Ž .Source: Arundel et al. 1995 . Respondents in 16 industries rated the importance of publicly funded research on a 7-point scale. The figuresindicate the percentage of respondents rating publicly funded research 5 or higher.


knowledge and fundamental science; for example,knowledge about electrical engineering often de-pends on fundamental discoveries in physics or

Ž .mathematics ibid., p. 342 .

5.6. Creation of new firms

The creation of new firms is sometimes cited as abenefit of government-funded research. However,the evidence is mixed as to whether new firms havebeen established on a significant scale as result ofgovernment funding. There are certainly some spec-tacular examples of regional agglomerations of newfirms clustered around research-intensive universitiessuch as MIT and Stanford. Yet a review of severalstudies found little convincing evidence that signifi-cant investment in basic research generates spin-offcompanies; while the correlation between universityresearch and firm birth18 is positive and statisticallysignificant in the electronic equipment sector, in

Žother sectors it is statistically insignificant Bania et.al., 1993 . New firm growth is not, however, the

only issue. Often the companies created by spin-offsŽremain quite small and have a high failure rate for

.example, in the software sector . Such companiescan provide an important source of Schumpeterianclustering around a new technology, but simple headcounts of numbers of new firms can be misleading.Firm entry and exit rates vary considerably betweensectors and regions. Moreover, many of the firmswhich are ‘spun out’ of universities have low growth

Ž .rates Massey et al., 1992 . Studies of firms locatedin science parks indicate that a connection to auniversity can be advantageous for new small firms,

Žbut this advantage is often quite limited Storey and.Tether, 1998 . Successful and sustained innovation

involves more than the development of an idea. AsŽ .Stankiewicz 1994, p. 101 has noted, AAcademics

do not make good entrepreneurs and the effectiveexploitation of their technology usually requires thatthe ownership of the technology and the managerialcontrol are taken out of their hands at an earlystageB.

18 The study ignored firm deaths.

6. Conclusions

6.1. Findings

In this study, we have critically reviewed theliterature on the economic benefits of publicly fundedresearch. As we have seen, this literature falls intothree main categories. One consists of econometricstudies, where there have been numerous attempts to

Ž .estimate the impact of research in general on pro-ductivity. Virtually all have found a positive rate ofreturn, and in most cases the figure has been compar-atively high. However, these attempts have beenbeset with both measurement difficulties and concep-tual problems such as the assumption of a simpleproduction function model of the science system. Inparticular, they tend to assume that research is, firstand foremost, a source of useful information to bedrawn upon in the development of new technologies,products and processes. This ignores the other formsof economic benefit discussed in Section 5.

As regards the specific case of basic research,one can try to estimate the rate of return but only onthe basis of very questionable assumptions. Mans-field’s work suggests there is a very substantial rate

Ž .of return, but the precise figure he arrives at 28% isŽ .open to some doubt. Among the problems are i the

complementary linkages of basic research activitieswith much larger ‘downstream’ investments in de-velopment, production, marketing and diffusion; andŽ .ii the complex and often indirect contributions ofbasic research to technology, the balance of whichvaries greatly across scientific fields and industrial

Ž .sectors. Recent work by Narin et al. 1997 based onthe scientific papers cited in patents provides a toolfor mapping these linkages and suggests that theknowledge flow from US science to US industry issubstantial and growing rapidly, although this find-ing is again subject to certain methodological reser-vations.

The econometric literature on localisation effectsand spillovers emphasises that advanced industrialcountries need their own, well developed basic re-search capabilities in order to appropriate the knowl-edge generated by others and to sustain technologicaldevelopment. Personal links and mobility are vital inintegrating basic research with technological devel-opment. This, in turn, highlights the importance of


linking basic research to graduate training, a point towhich we return later.

Surveys and case studies of different forms ofeconomic benefit from basic research represent thetwo other types of literature reviewed here, and thesehave yielded a number of findings. One concerns thetraditional justification for the public funding ofbasic research which is based on the argument thatscience is a public good, with the emphasis being onthe role of basic research as a source of new usefulknowledge, especially in a codified form. However,numerous studies have shown that there are severalother forms of economic benefit from basic research,and that new useful knowledge is not necessarily theprincipal type of benefit. This review has proposed aclassification scheme based on six categories of ben-efit.19

The first category relates to basic research as asource of new useful knowledge, while the secondconsists of new instrumentation and methodologies.As we have seen, the transfer of a new instrumentfrom basic research to industry can open up newtechnological opportunities or dramatically alter thepace of technological advance. Thirdly, there are theskills developed by those involved in carrying outbasic research, especially graduate students, whichcan also lead to substantial economic benefits asindividuals move on from basic research, carryingwith them both codified and tacit knowledge. Thetacit knowledge and skills generated by basic re-search are especially important in newly emergingand fast-moving areas of science and technology. Afourth type of benefit stems from the fact that partic-ipation in basic research is essential if one is toobtain access to national and international networksof experts and information. Fifthly, basic researchmay be especially good at developing the ability totackle and solve complex problems — an ability thatoften proves of great benefit in firms and other

19 As one referee has pointed out, the first category of benefit isconcerned more with the production of knowledge, while theothers focus on different mechanisms for the transfer and appro-priation of knowledge. This brings us back to the thesis of CallonŽ .1994 that scientific knowledge is not a public good becausepublication alone is not sufficient for the transfer and appropria-tion of that knowledge.

organisations confronted with complex technologicalproblems. Lastly, basic research may lead to thecreation of ‘spin-off’ companies, where academicstransfer their skills, tacit knowledge, problem-solv-ing abilities and so on directly into a commercialenvironment. However, the available evidence is notso convincing as to the importance of this form ofbenefit compared with the others mentioned above.

The relative importance of the different forms ofeconomic benefit distinguished here varies with sci-entific field, technology and industrial sector — i.e.there is great heterogeneity in the relationship be-tween basic research and innovation. Consequently,no simple model of the nature of the economicbenefits from basic research is possible. In particu-lar, the traditional view of basic research as a sourcemerely of useful codified information is too simpleand misleading. It neglects the often benefits oftrained researchers, improved instrumentation andmethods, tacit knowledge, and membership of na-tional and international networks. It should not,therefore, be used on its own as the basis for policymeasures. In short, the overall conclusions emerging

Ž .from the surveys and case studies are that: i theeconomic benefits from basic research are both real

Ž .and substantial; ii they come in a variety of forms;Ž .and iii the key issue is not so much whether the

benefits are there but how best to organise thenational research and innovation system to make themost effective use of them.

This brings us back to the issue of the rationalefor public funding of basic research. Governmentsare under increasing pressure to justify public expen-diture on basic research and the traditional justifica-

Žtion for public funding of basic research as first set.out by Nelson and Arrow needs to be extended. Not

only is basic research a source of codified informa-tion but it also yields a variety of other forms ofeconomic benefit. A more effective rationale for thepublic support of basic research should take thesefully into account. Such a rationale has yet to beconstructed. However, this review of the literaturehas pointed to some of the likely constituent ele-ments. These include the argument of Klevorick et

Ž .al. 1995 and others that basic research createstechnological opportunities, and the view of CallonŽ .1994 that basic research provides a source of newinteractions, networks and technological options, thus


increasing technological diversity. Also relevant hereis the work of Rosenberg, Pavitt and others showing

Ž .the importance of basic research as a source of iŽthe skills particularly those based on tacit knowl-

.edge required to translate knowledge into practice,Ž .ii an enhanced ability to solve complex technologi-

Ž .cal problems, and iii the ‘entry ticket’ to the world’sstock of knowledge, providing the ability to partici-pate effectively in networks and to absorb and ex-ploit the resulting knowledge and skills.

This new rationale for public funding of basicresearch also needs to take into account the debateabout the changing nature of research. Gibbons et al.Ž .1994 have suggested that nature of knowledgeproduction is shifting toward ‘Mode 2’, with greatercollaboration and transdisciplinarity and with re-search being conducted ‘in the context of applica-tion’. They argue that, in this new mode20 of knowl-edge production, the distinction between what ispublic and private in knowledge production andtherefore in science has become blurred, if not irrele-

Ž .vant. David 1995 has disputed these claims, argu-ing that the institutions of public science such aspeer review and publication provide the standards oftrust and authority essential not only for the effectiveperformance of research but also for the formulationof public policy and the provision of information for

Ž . 21societal choices David, 1995 .In recent editions of Research Policy, there has

been a heated debate between Kealey and DavidŽabout the benefits of such public investment Kealey,

.1996, 1998; David, 1997 . This review finds con-vincing evidence to support the idea that there areconsiderable economic benefits to the public fundingof basic research. These benefits are often subtle,heterogeneous, difficult to track or measure, andmostly indirect. Publicly funded basic research shouldbe viewed as a source of new ideas, opportunities,methods and, most importantly, trained problem-solvers. Hence, support for basic research should be

20 Some have questioned whether this is a ‘new’ mode ormerely a return to a form of knowledge production more commonin the 19th and earlier part of the 20th Centuries.

21 Others might dispute David’s notion of the scientific processŽ .e.g. Fuller, 1997 .

seen as an investment in a society’s learning capa-bilities.

6.2. Policy implications

What are the policy implications that follow fromthis review of the literature? As we have seen,although its economic benefits are hard to quantify,basic research is crucial for the strategic position ofindustrialised nations in the world economy, and forremaining at the leading edge of technology. This

Žhas been true in the past especially in chemicals and.pharmaceuticals and will remain true in the future

as new technologies draw increasingly on the outputsof basic research, on leading-edge scientific prob-lem-solvers, and on the emerging fields based on acombination of scientific and technological know-how.

However, for various reasons emerging from thisreview, it is difficult to arrive at simple policyprescriptions. One reason relates to the variations inthe forms of interaction between basic research andinnovation, and in the relative importance of differ-ent forms of economic benefit with scientific field,technology and industrial sector.22 Secondly, there isthe dependence of new products and processes on arange of technologies, and the dependence of newtechnologies on a large number of scientific fields;another way of expressing this is in terms of growingtechnological complexity and the need to ‘fuse’ pre-viously separate streams of science or technology. Athird reason concerns the importance of ‘spillovers’,including both geographical effects and the interac-tions between one form of activity and another. Inshort, there can be no simple unified policy for basicresearch.

Nevertheless, several policy lessons emerge fromthis review. First, policies must ensure that basicresearch is closely integrated with the training ofgraduate students, with the latter carried out in or-ganisations at the forefront of their field. This hasimplications for the appropriate mix of public fund-ing of basic research in universities and in central

22 This has implications for national ‘Foresight’ exercises wherethe approach adopted in different fields and sectors needs to besufficiently flexible to take account of these differences.


government research institutes.23 Secondly, given thecontribution to innovation which can flow from newinstrumentation, research grants should include ade-quate resources for accessing the latest instrumenta-tion, for developing experimental facilities and newmethodologies, and for funding technicians to assistin these tasks. Thirdly, the evidence that skilledgraduates who enter industry are a major channelthrough which basic research is transformed intoeconomic benefit suggests that policies should bedirected towards increasing the industrial recruitmentof qualified scientists and engineers, particularly byfirms that currently lack this resource. Fourthly,since a single piece of basic research may contributeto many different technological and product develop-ments, and those developments may, conversely,draw upon a number of research fields, nations needa portfolio-based approach to the public funding ofbasic research — a portfolio both in terms of re-search fields and technologies but also in terms of afull range of mechanisms and institutions for ensur-ing that the potential benefits of publicly fundedresearch are transferred and exploited successfully.Fifthly, the return from research depends crucially onhaving access to the outputs of publicly funded basicresearch, whether skilled people, techniques, instru-mentation or other outputs. Without access to these,none of the downstream benefits are likely to becaptured.

A sixth and final policy conclusion is that nonation can ‘free-ride’ on the world scientific system.In order to participate in the system, a nation orindeed a region or firm needs the capability tounderstand the knowledge produced by others andthat understanding can only be developed throughperforming research. Investments in basic researchenable national actors to keep up with and, occasion-ally, to contribute to the world science system.24 Yet

23 We are grateful to the referee who made this point.24 Ž .Nelson and Wright 1994 have suggested that some form of

modified free-riding might have been possible in the 1970s and1980s as Japan and other Asian countries were able to exploitareas of Western research and to make more rapid advances intheir technological capabilities. Whatever the merits of this claim,these countries have substantially increased their investments in

Ž .basic research in the 1990s OECD, 1999, p. 33 , which suggeststhat free-riding, at least in its ‘pure’ form, is no longer feasible.

the research reviewed in this paper does not suggesthow much public support should be provided nor inwhat areas it should be invested. Currently, we donot have the robust and reliable methodological toolsneeded to state with any certainty what the benefitsof additional public support for science might be,other than suggesting that some support is necessaryto ensure that there is a ‘critical mass’ of researchactivities. The literature available has shown thatthere are considerable differences across areas ofresearch and across countries and that additionalresearch is needed to better define and understandthese differences. This limitation in current sciencepolicy research should not be seen as implying aneed for less government funding of science. Rather,it indicates that public funding for basic research is,

Žlike many areas of government spending e.g. de-.fence , not easy to justify solely in terms of measur-

able economic benefits.

Acknowledgements

The authors wish to acknowledge the pioneeringcontributions of the late Edwin Mansfield to thisfield. Mansfield developed innovative methods foranalysing the relationship between basic researchand industrial innovation and his work has inspired anew generation of research. We also thank DianaHicks, Mike Hobday, Richard Nelson, Keith Pavitt,Jacky Senker, Margaret Sharp, Ed Steinmueller, NickVon Tunzelmann, David Wolfe, Frieder Meyer-Krahmer and the two anonymous referees for com-ments on earlier drafts. The usual disclaimers apply.

References

Abramowitz, M., 1986. Catching up, forging ahead, and fallingŽ .behind. Journal of Economic History XLVI 2 , 385–406.

Acs, Z.J., Audretsch, D., Feldmann, M., 1991. Real effects ofacademic research: comment. American Economic Review 82,363–367.

Adams, J., 1990. Fundamental stocks of knowledge and productiv-ity growth. Journal of Political Economy 98, 673–702.

Aghion, P., Howitt, P., 1995. Research and development in theŽ .growth process. Journal of Economic Growth 1 1 , 49–73.

Anselin, L., Varga, A., Acs, Z., 1997. Local geographic spilloversbetween university research and technology innovations. Jour-

Ž .nal of Urban Economics 42 3 , 422–448.


Arrow, K., 1962. Economic welfare and the allocation of re-Ž .sources for invention. In: Nelson, R. Ed. , The Rate and

Direction of Inventive Activities. Princeton Univ. Press,Princeton, pp. 609–625.

Arundel, A., Van de Paal, G., Soete, L., 1995. PACE Report:Innovation Strategies of Europe’s Largest Firms: Results ofthe PACE Survey for Information Sources, Public Research,Protection of Innovations, and Government Programmes. FinalReport, MERIT, University of Limburg, Maastricht.

Baldwin, J., Da Pont, M., 1996. Innovation in Canadian Manufac-turing Enterprises: Survey of Innovation and Advanced Tech-nology 1993. Cat. No. 88-513-XPB, Statistics Canada, Ottawa.

Bania, N., Eberts, R., Fogarty, M., 1993. Universities and thestart-up of new companies. Review of Economics and Statis-

Ž .tics 75 4 , 761–766.Beise, M., Stahl, H., 1999. Public research and industrial innova-

tions in Germany. Research Policy 28, 397–422.Bergman, E.M., 1990. The economic impact of industry-funded

university R&D. Research Policy 19, 340–355.Callon, M., 1994. Is science a public good? Science, Technology

and Human Values 19, 345–424.Cohen, W., Levinthal, D., 1989. Innovation and learning: the two

faces of R&D. Economic Journal 99, 569–596.CBO, 1993. A Review of Edwin Mansfield’s Estimate of the Rate

of Return from Academic Research and its Relevance to theFederal Budget Process. CBO Staff Memorandum, April, pp.1-26, Congressional Budget Office, Washington, DC.

Cooke, P., Morgan, K., 1993. The network paradigm: new depar-tures in corporate and regional development. Environment andPlanning D: Society and Space 11, 543–564.

Cooke, P., Morgan, K., 1998. The Associational Economy: Firms,Regions, and Innovation. Oxford Univ. Press, Oxford.

Dasgupta, P., David, P., 1994. Towards a new economics ofscience. Research Policy 23, 487–521.

David, P., 1995. Post-Modern Visions of Research and the Curseof Success. MERIT Working Paper 2.96-002, University ofLimburg, Maastricht, December.

David, P., 1997. From market science to calypso science policy: areview of Terence Kealey’s the economic laws of scientificresearch. Research Policy 26, 229–255.

David, P., Foray, D., 1995. Accessing and expanding the scienceand technology base. STI Review. OECD, Paris, pp. 13–68.

David, P., Mowery, D., Steinmueller, W.E., 1992. Analysing theeconomic payoffs from basic research. Economics, Innovationand New Technology 2, 73–90.

Dennison, E., 1985. Accounting for Growth. Harvard Univ. Press,Cambridge.

De Solla Price, D., 1984. The sciencertechnology relationship,the craft of experimental science and policy for the improve-ment of high technology innovation. Research Policy 13,3–20.

Dosi, G., 1982. Technological paradigms and technological trajec-tories: a suggested interpretation of the determinants anddirections of technical change. Research Policy 11, 147–162.

Dosi, G., Pavitt, K., Freeman, C., Nelson, R., Soete, L., 1988.Technical Change and Economic Theory. Cambridge Univ.Press, Cambridge.

Ž .Edquist, C. Ed. 1997. Systems of Innovation: Technologies,Institutions and Organizations. Pinter, London.

Faulkner, W., Senker, J., 1995. Knowledge Frontiers: PublicSector Research and Industrial Innovation in Biotechnology,Engineering Ceramics, and Parallel Computing. ClarendonPress, Oxford.

Feldmann, M., Florida, R., 1994. The geographic sources ofinnovation: technological infrastructure and product innova-tion in the United States. Annals of the Association of Ameri-

Ž .can Geographers 84 2 , 210–229.Freeman, C., 1984. The Economics of Industrial Innovation.

Pinter, London.Fuller, S., 1997. The secularization of science and a new deal for

Ž .science policy. Futures 29 3 , 483–504.Gibbons, M., Johnston, R., 1974. The role of science in technolog-

ical innovation. Research Policy 3, 220–242.Gibbons, M., Limoges, C., Nowotony, H., Schwartzman, S.,

Scott, P., Trow, M., 1994. The New Production of Knowl-edge. Sage, London.

Godin, B., 1996. Research and the practice of publication inindustries. Research Policy 25, 587–606.

Griliches, Z., 1995. R&D and productivity. In: Stoneman, P.Ž .Ed. , Handbook of Industrial Innovation. Blackwell, London,pp. 52–89.

Grossman, G.M., Helpman, E., 1991. Innovation and Growth inthe Global Economy. MIT Press, Cambridge.

Grossman, G.M., Helpman, E., 1994. Endogenous innovation inŽ .the theory of growth. Journal of Economic Perspectives 8 1 ,

23–44.Hall, B.H., 1993. Industrial research during the 1980s: did the rate

of return fall? Brookings Papers: Microeconomics 2, 289–343.Hicks, D., 1995. Published papers, tacit competencies and corpo-

rate management of the publicrprivate character of knowl-edge. Industrial and Corporate Change 4, 401–424.

Hicks, D.M., Isard, P.A., Martin, B.R., 1996. A morphology ofJapanese and European corporate research networks. ResearchPolicy 25, 359–378.

Hicks, D., Katz, J.S., 1997. The British Industrial Research Sys-tem. SPRU Working Paper, University of Sussex, Brighton.

Hicks, D., Olivastro, D., 1998. Are There Strong In-state LinksBetween Technology and Scientific Research. Issue Brief,Division of Science Resources Studies, CHI Research, CherryHill.

Jaffe, A., 1989. Real effects of academic research. AmericanEconomic Review 79, 957–970.

Katz, J.S., 1994. Geographical proximity and scientific collabora-Ž .tion. Scientometrics 31 1 , 31–43.

Kealey, T., 1996. The Economic Laws of Scientific Research.MacMillan, London.

Kealey, T., 1998. Why science is endogenous: a debate with PaulDavid. Research Policy 26, 897–923.

Klevorick, A.K., Levin, R., Nelson, R., Winter, S., 1995. On thesources and significance of inter-industry differences in tech-nological opportunities. Research Policy 24, 185–205.

Link, A., 1982. An analysis of the composition of R&D spending.Southern Economic Journal 49, 342–349.

Los, B., Verspagen, B., 1996. R&D spillovers and productivity:


evidence from US manufacturing microdata. Paper presentedat the 6th Conference of the Joseph A. Schumpeter Society,2–5 June, Stockholm, Sweden.

Lucas, R.E., 1988. On the mechanics of economic development.Ž .Journal of Monetary Economics 22 1 , 3–42.

Lundvall, B.A., 1988. Innovation as an interactive process: fromuser–producer interaction to the national system of innovation.

Ž .In: Dosi, G. Ed. , Technical Change and Economic Theory.Frances Pinter, London, pp. 349–369.

Ž .Lundvall, B.A. Ed. 1992. National Systems of Innovation: To-wards a Theory of Innovation and Interactive Learning. FrancesPinter, London.

Mansfield, E. et al., 1991. Academic research and industrialinnovation. Research Policy 20, 1–12.

Mansfield, E., 1998. Academic research and industrial innovation:an update of empirical findings. Research Policy 26, 773–776.

Mansfield, E., Lee, J.Y., 1996. The modern university: contributorto industrial innovation and recipient of industrial R&D sup-port. Research Policy 25, 1047–1058.

Marsili, O., 1999. The anatomy and evolution of industries:technical change and industrial dynamics. Doctoral thesis,SPRU, University of Sussex, Brighton.

Martin, B.R., Irvine, J., 1981. Spin-off from basic science: thecase of radio astronomy. Physics in Technology 12, 204–212.

Martin, B., Salter, A., Hicks, D., Pavitt, K., Senker, J., Sharp, M.,Von Tunzelmann, N., 1996. The Relationship Between Pub-licly Funded Basic Research and Economic Performance: ASPRU Review. HM Treasury, London.

Martin, F., 1998. The economic impact of Canadian universityR&D. Research Policy 27, 677–687.

Massey, D., Quintas, P., Wield, D., 1992. High Tech Fantasies:Science Parks in Society, Science and Space. Routledge,London.

Metcalfe, J.S., 1995. The economic foundations of technologypolicy: equilibrium and evolutionary perspective. In: Stone-

Ž .man, P. Ed. , Handbook of Industrial Innovation. Blackwell,London.

Meyer-Krahmer, F., Schmoch, U., 1998. Science-based technolo-gies: university–industry interactions in four fields. ResearchPolicy 27, 835–851.

Mowery, D., 1995. The practice of technology policy. In: Stone-Ž .man, P. Ed. , Handbook of Industrial Innovation. Blackwell,

London, pp. 513–557.Narin, F., Hamilton, K., Olivastro, D., 1997. The linkages be-

tween US technology and public science. Research Policy 26,317–330.

Nelson, R.R., 1982. The role of knowledge in R&D efficiency.Quarterly Journal of Economics 97, 453–470.

Nelson, R.R., 1987. Understanding Technical Change as an Evo-lutionary Process. North-Holland, Amsterdam.

Nelson, R.R., 1993. National Innovation Systems: A ComparativeAnalysis. Oxford Univ. Press, New York.

Nelson, R.R., 1995. Why should managers be thinking aboutŽ .technology policy. Strategic Management Journal 16 8 , 581–

588.Nelson, R.R., 1998. The agenda for growth theory: a different

Ž .point of view. Cambridge Journal of Economics 22 4 , 497–520.

Nelson, R., Rosenberg, N., 1994. American universities and tech-Ž .nical advance. Research Policy 23 3 , 323–348.

Nelson, R.R., Wright, G., 1994. The erosion of U.S. technologicalleadership as a factor in postwar economic convergence. In:

Ž .Baulmol, W.J., Nelson, R.R., Wolff, E.N. Eds. , Convergenceof Productivity: Cross-national Studies and Historical Evi-dence. Oxford Univ. Press, Oxford.

Nelson, R.R., Gelijis, A., Raider, H., Sampat, B., 1996. A prelimi-nary report on Columbia inventing. Mimeo, September 25,1996, University of Columbia, New York.

Nightingale, P., 1997. Knowledge in the process of technologicalinnovation: a study of the UK pharmaceutical, electronic andaerospace industries. Doctoral dissertation, SPRU, Universityof Sussex, Brighton.

Nonaka, I., Takeuchi, H., 1995. The Knowledge-Creating Com-pany. Oxford Univ. Press, Oxford.

NSF, 1969. Technology in Retrospect and Critical Events inŽ .Science TRACES . National Science Foundation, Washing-

ton, DC.OECD, 1996. Science, Technology and Industry Outlook. Organi-

sation for Economic Co-operation and Development, Paris.OECD, 1999. Science, Technology and Industry Scoreboard 1999:

Benchmarking Knowledge-based Economies. Organisation forEconomic Co-operation and Development, Paris.

OTA, 1986. Research funding as an investment: can we measurethe returns? A technical memorandum, Office of TechnologyAssessment, US Government Printing Office, Washington,DC.

OTA, 1995. Innovation and Commercialization of EmergingTechnology. Office of Technology Assessment, US Govern-ment Printing Office, Washington, DC.

Patel, P., Pavitt, K., 1995. The nature and economic importance ofnational innovation systems. STI Review. OECD, Paris, pp.9–32.

Pavitt, K., 1991. What makes basic research economically useful?Research Policy 20, 109–119.

Pavitt, K., 1998. The social shaping of the national science base.Research Policy 27, 793–805.

Polanyi, M., 1962. Personal Knowledge: Towards a Post-CriticalPhilosophy. University of Chicago Press, Chicago.

Rappa, M., Debackere, K., 1992. Technological communities andŽ .the diffusion of knowledge. R&D Management 22 3 , 209–

222.Romer, P.M., 1990. Endogenous technological change. Journal of

Ž .Political Economy 98 5 , S71–S102.Romer, P.M., 1994. The origins of endogenous growth. Journal of

Ž .Economic Perspectives 8 1 , 3–22.Rosenberg, N., 1982. Inside the Black Box: Technology and

Economics. Cambridge Univ. Press, Cambridge.ŽRosenberg, N., 1990. Why do firms do basic research with their

.own money ? Research Policy 19, 165–174.Rosenberg, N., 1992. Scientific instrumentation and university

research. Research Policy 21, 381–390.Saxenian, A., 1994. Regional Advantage: Industrial Adaptation in


Silicon Valley and Route 128. Harvard Univ. Press, Cam-bridge.

Schrader, S., 1991. Informal technology transfer between firms:co-operation through information trading. Research Policy 20,153–170.

Scherer, F.M., 1982. Inter-industry technology flows in the UnitedStates. Research Policy 11, 227–246.

Schmoch, U., 1997. Indicators of the relations between scienceŽ .and technology. Scientometrics 38 1 , 103–116.

Senker, J., 1995. Tacit knowledge and models of innovation.Ž .Industrial and Corporate Change 4 2 , 425–447.

Smith, K., 1991. Economic returns to R&D: method, results, andchallenges. Science Policy Support Group Review Paper No.3, London.

Solow, R.M., 1957. Technical change and the aggregate produc-tion function. Review of Economics and Statistics 39, 312–320.

Stankiewicz, R., 1994. Spin-off companies from universities. Sci-Ž .ence and Public Policy 21 2 , 99–108.

Steinmueller, E., 1994. Basic research and industrial innovation.Ž .In: Dodgson, M., Rothwell, R. Eds. , The Handbook of

Industrial Innovation. Edward Elgar, Aldershot, pp. 54–66.Storey, D., Tether, B., 1998. New technology-based firms in the

European union: an introduction. Research Policy 26, 933–946.

Storper, M., 1995. The resurgence of regional economics, 10years later: the region as a Nexus of untraded interdependen-

Ž .cies. European Urban and Regional Studies 2 3 , 191–221.Storper, M., 1997. The Regional World: Territorial Development

in a Global Economy. Guilford Press, New York.Teece, D., 1986. Profiting from technological innovation: implica-

tions for integration, collaboration, licensing and public policy.Research Policy 15, 285–305.

Vincenti, W., 1990. What Engineers Know and How They KnowIt. John Hopkins Press, Baltimore.

Verspagen, B., 1993. Uneven Growth Between InterdependentEconomies: an Evolutionary View on Technology Gaps, Tradeand Growth. Avebury, Aldershot.

Von Hippel, E., 1987. Co-operation between rivals: informalknow-how trading. Research Policy 16, 291–302.

Wolfe, D., 1996. The emergence of the region state. In: Courch-Ž .ene, T.J. Ed. , The Nation State in a GlobalrInformation Era:

Policy Challenges, John Deutsch Institute for the Study ofEconomic Policy, Queen’s University, Kingston, Ontario, pp.205–240.

Wolfe, D., Salter, A., 1997. The Socio-Economic Importance ofScientific Research to Canada. A Discussion Paper preparedfor the Partnership Group for Science and Engineering, Ot-tawa, October.

Documents

The economic benefits of publicly funded basic …Research Policy 30 2001 509–532Ž. The economic benefits of publicly funded basic research: a critical review Ammon J. Salter),