Research Policy 37 (2008) 1854–1864

Contents lists available at ScienceDirect

Research Policy

journa l homepage: www.e lsev ier .com/ locate / respol

Close enough but not too far: Assessing the effects ofuniversity–industry research relationships and the riseof academic capitalism

Rick Welsha,∗, Leland Glennab,1, William Lacyc,2, Dina Biscottid,3

a Department of Humanities and Social Sciences, Clarkson University, P.O. Box 5750, Potsdam, NY 13699, United Statesb Department of Agricultural Economics and Rural Sociology, The Pennsylvania State University, University Park, PA 16802, United Statesc Outreach Department, University Outreach and International Programs, University of California, Davis, 220 Mrak Hall, Davis, CA 95616, United Statesd Department of Sociology, University of California, Davis, 1 Shields Avenue, Davis, CA 95616, United States

a r t i c l e i n f o

Article history:Received 1 March 2007Received in revised form 7 July 2008Accepted 17 July 2008Available online 30 August 2008

Keywords:Agricultural biotechnology

a b s t r a c t

Analysts assessing the impact of university–industry research relations (UIRRs) and increas-ing proprietary behavior on the part of universities often focus on single-indicators or adoptpromotional or critical stances. However, assessing impacts of shifts toward a more propri-etary university is inherently complex because of potential countervailing or mediatingfactors within working relationships. From interviews with 84 biological scientists at nineuniversities we find scientists view UIRRS and university intellectual property (IP) policiesin complex and often conflicting ways. For example, university scientists believe UIRRs are

Intellectual propertyTechnology transferUniversitiesResearch

valuable for increasing contact with scientists, but are problematic because working withindustry can restrict communication among scientists. Also scientists believe universityIP policies should shield their work from opportunistic behavior and at the same time bedesigned to attract industry partners. In addition scientists believe universities use their IPpolicies primarily as revenue raising vehicles and secondarily to address public good issues

transf

such as technology

1. Introduction

According to Nelsen (1991, p. 39), “The Universitywas the birthplace of the biotechnology industry, and

it continues to be the source of most of the basicnew technology that fuels the industry.” He argues thatuniversity–industry research relationships (UIRRs) are the“lifeblood” of the biotechnology industry because univer-

∗ Corresponding author. Tel.: +1 315 268 3988; fax: +1 315 268 3983.E-mail addresses: welshjr@clarkson.edu (R. Welsh), llg13@psu.edu

(L. Glenna), wblacy@ucdavis.edu (W. Lacy), dbiscotti@ucdavis.edu(D. Biscotti).

1 Tel.: +1 814 863 8636; fax: +1 814 865 3746.2 Tel.: +1 530 752 6376.3 Tel.: +1 530 752 5811.

0048-7333/$ – see front matter © 2008 Elsevier B.V. All rights reserved.doi:10.1016/j.respol.2008.07.010

er.© 2008 Elsevier B.V. All rights reserved.

sity biological research is by its nature basic and embryonic(Nelsen, 1991, p. 39). To secure access to the intellectualproperty associated with new discoveries and technologies,private biotechnology firms seek to develop and strengthentheir relationships with universities. Such relationshipshave been important for agri-biotechnology applicationssince much of the research and technology transfer thathas yielded transgenic crops has come from industry rela-tionships with U.S. public land-grant universities (Busch etal., 1991; Charles, 2001).

Rudy et al. (2007) argue that the historical develop-ment of land-grant universities has been driven by two

sometimes competing missions. One, a populist mission,is to serve the public good through the production anddissemination of scientific and technical knowledge thatis widely available and useful to a broad range of ruralresidents and entrepreneurs. The other is to enhance

h Policy

tatcpbglohTmB(e

dgunaeoitrDovtL2

gtniiptianot1cdfipgfir

tUlbufL

R. Welsh et al. / Researc

he productivity of agriculture through the developmentnd transfer of technologies and dissemination of scien-ific farming techniques. When these two missions haveonflicted, administrators have had to negotiate and com-romise on their institution’s commitment to one oroth (Rudy et al., 2007). The relationship between landrants and the agri-biotechnology industry, combined withegislative and policy changes that promote a more market-riented research and development policy environment,as highlighted the tension between these two missions.he most prominent agricultural example of this tensionay be the controversy surrounding the now defunct UC

erkeley-Novartis agri-biotechnology research agreementsee Busch et al., 2004; Kirp and Popp-Berman, 2003; Rudyt al., 2007).

Three interrelated developments lie at the center of theebates concerning the appropriate links between land-rant universities and private firms and their effects on theniversity’s societal role: (1) a series of legislation (begin-ing with the Bayh-Dole Act of 1980), executive ordersnd court decisions served to facilitate patenting of fed-rally funded research and drove increases in the numberf universities actively engaged in patenting and licens-ng technologies and discoveries (Mowery et al., 2004); (2)he decrease in state and federal support for agriculturalesearch relative to private sector investment (Caswell anday-Rubenstein, 2006); and (3) the increasing emphasisn university biotechnology research as an engine of inno-ation that promises to lead to economic growth throughhe commercialization of new technologies (Slaughter andeslie, 1997; Mowery et al., 2004; Rudy et al., 2007; Shane,004).

In fact, many of the scholars investigating the emer-ence and expansion of the biotechnology industry viewhese three developments as interconnected with eco-omic cycles. These scholars view the biotechnology

ndustry as a response to the realization of the lim-ts of nationalist-centered, mass-production developmentrojects that were dominant in the second half of thewentieth century (McMichael, 2000). As the ability to real-ze sustained economic growth and profits through suchpproaches peaked in the late 1960s and early 1970s, theew technologies that comprised the “information econ-my” provided the opportunity for companies to add valueo the production process (Kloppenburg, 2004; Kenney,986). They also facilitated a shift away from nationalist-entered development strategies toward a more globalivision of labor by increasing the mobility of capital andrms (McMichael, 2000). These developments in turn putressure on the fiscal situation of national and sub-nationalovernments, making it more difficult to raise funds fornancing public investments in such things as agriculturalesearch (Slaughter and Leslie, 1997).

In general prior to the 1980s, it was widely perceivedhat efforts to initiate biotechnology developments in thenited States had been hindered by the fact that the

abor and expertise that were to spur the commercialiology phenomenon had been located in U.S. researchniversities, which had, for the most part, been shelteredrom labor and commercial market forces (Slaughter andeslie, 1997; Kenney, 1986; Sampat, 2006). In the case

37 (2008) 1854–1864 1855

of agri-biotechnologies, a second obstacle to developingcommercial applications emerged as public funding forresearch universities declined (Buttel et al., 1984; Klein-man and Vallas, 2001; Caswell and Day-Rubenstein, 2006).A strategy for overcoming these obstacles has been to struc-ture research performed at U.S. universities to be morereadily utilized by U.S. firms (Buttel et al., 1984; Mow-ery et al., 2004). Through legislative changes such as the1980 Bayh-Dole Act and the 1986 Federal Technology Trans-fer Act that promoted university–industry collaborations,and through funding from the federal and state govern-ments and direct investments from biotechnology firms,research universities became the primary source of newbiotechnology techniques and products (Busch et al., 1991).Described as the emergence of “academic entrepreneur-ship” and “academic capitalism,” universities have createdand expanded technology transfer and university–industryrelation offices to promote collaborations and the market-ing of research discoveries, created or revised policies toenable university professors to form start-up companies,and generally pursued a shift in emphasis to producingresearch for commercial applications (Slaughter and Leslie,1997; Shane, 2004). These changes have led to a burgeoningliterature (descriptive, promotional and critical) that ana-lyzes and assesses their impact on the functioning of theuniversity and the net benefits to society.

In this paper we take a brief historical look at the link-ages between universities and industry and the proprietaryactivity of universities, and then review the descriptive,promotional and critical literatures of the impacts on theuniversity and society from the university’s shift towarda more proprietary orientation. After this we provide theoutline of a framework for assessing the outcomes linkedto academic entrepreneurialism based on our observationthat it is often the beneficial aspects of university–industryrelations that can also be the most problematic, and viceversa. We support this paradoxical assertion through theuse of interview data collected from 84 agricultural sci-entists with industry connections at nine universities (themajority of the interviews come from six major land-grant universities). We conclude by making the case thatuniversity-level policies, especially university IP policies,are critical vehicles for managing the relationships betweenindustry and universities in order to enhance the long-termeconomic and social welfare of universities, their scientists,and private sector firms.

2. Assessing the effects of academic capitalism

In a political–economic analysis of the Bayh-Dole Act,Mowery et al. (2004) find that patenting activities at uni-versities and by university scientists has a long history inthe U.S. In fact, in 1907 the UC Berkeley scientist Freder-ick Gardner Cottrell received the first of several patentson an air pollution control mitigation technology. How-ever, Cottrell declined to have UC administer the patents

because he believed that “. . .the involvement of universityadministrators in licensing management could have detri-mental consequences for the culture of scientific researchat the university (Mowery et al., 2004, p. 59).” Mowery et al.(2004) go on to document selective participation in patent-

h Policy

1856 R. Welsh et al. / Researc

ing and licensing activity on the part of universities fromthe mid-1920s until the 1960s and 1970s, when the trendticks upward. They conclude that Bayh-Dole did not causeincreases in patenting activity as the trend was underwayalready. Rather Bayh-Dole provided uniform federal poli-cies regarding the patenting of federally funded research,thus replacing agency specific policies negotiated on indi-vidual bases with universities and other actors. This shiftattracted a greater number of universities into proprietaryactivity through the establishment of offices of technologytransfer (see also, Sampat, 2006; Shane, 2004).

Despite the long history and centrality to the U.S.economy of working relationships between industry anduniversities and evolving university IP policies, and thelong acknowledged trade-offs and need for care in devel-oping and managing UIRR policies, efforts to analyze theeffects or impacts of the relatively recent increase in pro-prietary activity by universities and the related UIRRs haveoften taken a normative stance, and tend to appear in twobasic forms. The first comes from the economics of inno-vation and management literatures and operates withinan analytically descriptive to promotional framework, byexamining the motivations of, and payoffs to, individualscientists, managers, and administrators in industries anduniversities and to society as a whole. This view focuseson industries gaining new technologies and expertise fromthe universities, and universities increasing their relevanceand enhancing their reputation with economic growth andwealth creation (see Kunhardt, 2004). For example, Meyers(2006) finds that academic researchers who patent theirfindings (inventor–authors) generally publish at aboveaverage rates, but are not the most published universityscientists. And Breschi et al. (2005) find that patenting hasa more durable and beneficial impact on publication ratesfor scientists with more than one patent. While, Hicks andHamilton (1999) have discovered that university–industryco-authored papers are cited more frequently than single-university papers, indicating that university researchersmay enhance their scientific impact by collaborating withindustry partners. Franzoni (2006) argues that scientistsboost their productivity by working on industrial technolo-gies related to research applications.

Other studies have examined broad economic effects.For example, Cohen et al. (2002, p. 21) find amongother things that “. . .almost a third of industrial R&Dprojects for our sample firms-made use of research find-ings from public research.” Mueller (2006) determinedthat university–industry relations spur regional economicgrowth. And Mansfield (1998) calculated that 10% of infor-mation industry’s new products between 1986 and 1994could not have happened without academic research.Mansfield and Lee (1996, p. 1057) have found that “. . .firmslocated in the nation and area where academic researchoccurs are significantly more likely than distant firms tohave an opportunity to be among the first to apply thefindings of this research.” Therefore, tight geographic and

working links between industry and universities allowfirms to innovate and grow. Such findings led Zucker andDary (1998) and Zucker et al. (2002) to conclude that theaccess to academic research and star scientists is an impor-tant predictor of a biotechnology firm’s success.

37 (2008) 1854–1864

Regarding goal attainment on behalf of universities andindustry, Lee (2000) has discovered that generally univer-sity scientists have their primary interests met (such asadditional research funds and access to new materials andequipment). In contrast, industry scientists and managerstend not to see their primary interests met (easily marketedproduct). But they often meet their secondary interests(staying current on scientific expertise). Despite the failureto achieve their primary interests, the industry representa-tives tend to agree with their university counterparts that itis beneficial to expand or maintain collaborations. A num-ber of researchers emphasize that these collaborations areimportant because most university-produced technologiesneed significant additional development and investment tocommercialize them successfully. This fact has a profoundinfluence on a number of important factors including thetype of firm that licenses these technologies, the impor-tance of university patents relative to in-house industrypatents of private sector firms, and the critical need to tiethe university-based inventor’s income to the commercial-ization of the technology and involve the inventor in thecommercialization process (e.g., see Jensen and Thursby,2001; Thursby and Thursby, 2000, 2002, 2003).

Shane (2004) follows these arguments through his workon “academic entrepreneurship,” focusing on the increasein firms spun-off from universities by university scientiststo exploit intellectual property created at the university.Shane argues that such firms are important and benefi-cial to national, regional and local economies for a numberof reasons. In addition to the economic contribution, thefirms tend to be profitable, to locate close to their univer-sity progenitors, to provide high-wage jobs, and to providea vehicle for involving the inventor in the commercializa-tion process, which often enhances chance for successfultechnology commercialization. In addition, Shane (2004)explains why some universities produce relatively largenumbers of spin-offs while others are less prolific in thisarea: central factors in this regard include industry fundingof university research as well as the intellectual propertypolicies of universities (see also Landry et al., 2006).

A second major approach to analyzing UIRRs can befound in the higher education and the social studies ofscience literatures. It examines the broader institutionalconsequences of the relationships. This approach tends torest on assumptions that the industry and the universityplay overlapping but different roles in society and ques-tions whether those roles might be compromised whenthe organizations increase interactions in particular ways.For example, Hackett (2001) finds that increasingly societysends an ambivalent message to universities: universitiesshould perform their traditional (less business-oriented)role while also responding to national economic impera-tives. And, Krimsky (2003) believes that universities havefocused historically on four missions: knowledge as virtue,knowledge as productivity, knowledge in the service ofdefense, and knowledge in the service of the public inter-

est. He asserts that knowledge as productivity is subsumingknowledge in the public interest so that public interestis being redefined to mean technology commercializationand profit for private sector firms. Public-interest researchas something distinct from a market solution is being

h Policy

amsteram

sacartrT‘bnoc

asaesitat

rtHboiStgupaWsd

tdfatlbs

fou

R. Welsh et al. / Researc

bandoned. In addition, it is often argued that the rise ofarket-oriented solutions to technical problems has the

pillover effect of inhibiting research in particular scien-ific areas—an “anti-commons” effect. For example, in anmpirical examination of the effect of the granting of IPights on scientific citations, Murray and Stern (2007) findmodest negative “anti-commons” effect, that becomesore pronounced over time.In one of the more far-reaching and thorough critiques of

hifts toward a more market-oriented university, Slaughternd Rhoades (2004, 1996) and Slaughter and Leslie (1997)oin the term “academic capitalism.” Because much of thedvanced knowledge in the United States is contained inesearch universities, a central component of the construc-ion of the knowledge economy has been to integrate theesearch university into the intellectual property process.hey also document a bi-partisan political shift toward a

competitiveness agenda’ as the Cold War defense agendaegan to decline in political significance. The competitive-ess agenda entails the focus on universities as enginesf innovation and potential growth, and an emphasis onompetitive grants for allocating federal funds.

In addition, Slaughter and Leslie (1997) assert thats a professional class, academics have shielded them-elves from the vagaries of labor markets by maintainingmonopoly control over specific kinds of knowledge in

xchange for a tacit social contract: do research to benefitociety, not to maximize private gain. However, the pol-cy changes designed to generate economic growth duringhe 1970s and 1980s have led universities and professors todopt market-like behavior, using goods, services and laboro pursue profit.

These studies provide useful insights for assessingecent manifestations of the longer term working rela-ionships between universities and private sector firms.owever, it is our contention that an emphasis on theeneficial effects of academic entrepreneurism, or the sub-rdination of the public good to private interests, may resultn an understatement of the complexity of the situation.ome universities may use their relationships with indus-ry and their IP policies to undergird regional economicrowth and the enhancement of technology transfer. Otherniversities may seek to maximize licensing revenue fromatenting scientific knowledge or technology advances thatre not marketable without private sector collaboration.hile it is true that the legislative and policy context has

hifted such that liberalization is being encouraged, theifference can be viewed as in degree rather than kind.

In one study of how university scientists have respondedo these developments, Shin and Lamy (2006) use interviewata to determine that university scientist–entrepreneurs

all into one of three categories: academics, pioneers,nd Janus. Academics see strong interaction betweenheir commercial and academic activities, pioneers seeittle interaction, and Janus scientist–entrepreneurs crossetween the two worlds in a sequential fashion-operating

olely or primarily in one.

In a similar fashion, Owen-Smith and Powell (2001) haveound that some scientists tend toward an “old school”rientation and are skeptical of increasing ties betweenniversities and private sector firms. Other university sci-

37 (2008) 1854–1864 1857

entists are more “new school” and embrace the blurringof traditional lines between the university and the forprofit sector. Such “new schoolers” do commercially viableresearch in their labs and may “start-up companies” inwhich they play a major managerial role and have sig-nificant financial interest but also continue to work forthe university. They also describe two types of hybrid sci-entists: the “reluctant entrepreneur” and the “engagedtraditionalist.” The “reluctant entrepreneur” views the aca-demic and industrial worlds as distinct but engages inproprietary activity, patenting, while assigning ownershipto the university. To some extent the scientist considerspatenting necessary to protect academic freedom and theuniversity’s commercial or intellectual interest from com-mercial encroachment. The “engaged traditionalists” viewthe academic and industry worlds as distinct, but use theiracademic credentials for commercial gain through patent-ing and consulting work, outside of their university duties.

These studies are helpful and informative because theyillustrate how UIRRs and related university IP activities cantake different forms and outcomes based on the scien-tists’ value preferences in relation to policies. Both studiesdescribe a number of different perspectives on how onemight balance the trade-offs that scientists and their uni-versities face when considering their relationship withthe private sector. We hope to build on these contribu-tions by delineating scientist experiences of promisingdevelopments and concerns. Through in-depth interviewswith university scientists engaged in UIRRs, with a specialemphasis on agri-biotechnologies, we intend to measureand discern the micro-level interpretations and reactionsto macro-level changes as described above. We are espe-cially interested in the macro-level changes as mediatedthrough scientists involved in research relationships withprivate sector firms and the intellectual property (IP) poli-cies of the universities. By highlighting the promising andproblematic aspects of UIRRs, we intend to contribute tothe development of a model of assessment constructed tobalance the trade-offs inherent in the most recent mani-festations of the long-term interdependence between theuniversity and industry.

Since university scientists are significant stakeholdersin UIRRs, their perspectives on industry relationships andthe policies that govern those relationships are necessaryfor understanding the impacts of UIRRs and for developingeffective university and government policies for managingthem. In addition, focusing on agri-biotechnology scien-tists brings with it a number of advantages. Agriculture isan area that traditionally has been the recipient of sub-stantial public investment to support and attract privatesector investment (Mowery et al., 2004); and only in themid-1980s did private sector research investment in thisarea surpass public sector investment (Caswell and Day-Rubenstein, 2006). Moreover, biotechnology applicationsin agriculture are socially relevant because of their impor-tance to food production and the economy in general;

the fundamental role of university research in developingthese technologies; the fact that this technological focushas been a major source of interaction between industryand the university (Busch et al., 1991; Charles, 2001); and,that biotechnological techniques can be easier to patent

h Policy 37 (2008) 1854–1864

Table 1Rank order of university scientists views of characteristics of UIRRs (1 = NotCharacteristic–7 = Highly Characteristic)

Characteristic of UIR Mean Std. Dev.

Provides new research funds 6.0 0.9Increases access to new research tools 5.7 1.2Provides new support for graduate

students/postdocs5.6 1.2

Provides contact with a wider network ofscientists

5.5 1.5

Accelerates product development 5.4 1.2Increases access to new knowledge 5.0 1.7Increases potential for tension with

university conflict of interest policies4.9 1.6

Restricts scientific communication amonguniversity researchers

4.6 1.5

Inhibits materials transfer 4.5 1.8Increases access to industry’s intellectual

property4.5 1.7

Elevates university’s prestige 4.4 1.6De-emphasizes non-proprietary agendas 4.1 1.6Increases lawsuits over intellectual

property4.1 1.6

De-emphasizes basic science research 4.0 1.7Undermines the credibility of university

scientists3.9 2.0

Limits or restricts faculty members’ abilityto publish

3.7 1.7

Limits or restricts students’ ability to 3.6 1.8

1858 R. Welsh et al. / Researc

than traditional agricultural innovations (Day-Rubenstein,2003).

3. Data and methods

As part of a study of UIRRs in agricultural biotechnol-ogy, in-depth interviews were conducted with eighty-fouruniversity scientists at nine universities with research pro-grams related to agricultural biotechnology. The studypopulation was selected purposively rather than randomly,drawing on the concept of the information-rich case.Harper (2001, p. 27) argues that “a small number of well-informed informants are, in fact, a better sample thanmuch larger samples of minimally involved subjects.” And,according to Patton (1990, p. 169), “The logic and powerof purposeful sampling lies in selecting information-richcases for study in-depth. Information-rich cases are thosefrom which one can learn a great deal about issues of centralimportance to the purpose of the research, thus the termpurposeful sampling” (see also Charmaz, 2002).

We selected our LGU case study sites and intervieweesthrough a two-tiered purposeful sampling technique: onefor selecting the universities and a second for selectingthe scientists to be interviewed at each institution. Weselected five prominent LGUs that emphasize agriculturalbiotechnology research in the major U.S. regions. Prominentuniversities were defined according to size of sponsoredresearch budget, technological ratings, agricultural sci-ence citation ranking, and extensiveness of patenting andlicensing activities. Cornell University, North Carolina StateUniversity, Texas A&M University, University of Californiaat Davis, and University of Wisconsin became our majorcase studies, consisting of 54 interviews. For purposes ofcomparing insights from these prominent LGUs to otheruniversities, we also conducted several interviews at asmaller LGU, Oregon State University (15 interviews), twoprivate universities: Stanford University (eight interviews)and Duke University (three interviews), and a non-LGUpublic university, University of North Carolina (four inter-views).

The second-tier sampling was conducted by contactingactive researchers with industry contacts in agricul-tural biotechnology. We also contacted other researchersnot necessarily conducting agricultural biotechnologyresearch, but still involved with the private sector to someextent. The scientists came from a variety of backgrounds,but the overwhelming majority were biological scientistsconducting research directly related, or relevant, to agri-culture (see Glenna et al., 2007a,b). We focus on universityscientists with industry contacts in order to tap into theirworking knowledge and experiences regarding formal con-tractual arrangements and the important informal “culturaltraffic” identified by Vallas and Kleinman (2007) as criti-cal to understanding the nature and impacts of UIRRs andshifting societal roles of universities.

We informed respondents of the purpose of the case

studies by mail and contacted them by phone or e-mailto ask them to participate in an interview that wouldlast between 1 and 2 h. We assured interviewees that wewould maintain confidentiality. Interviews were conductedusing a semi-structured questionnaire and generally lasted

publishIncreases tensions between university

colleagues3.5 1.7

90 min. Scientists were asked a set of questions about theirbackgrounds and research interests including opinions oftheir university’s mission and the extent and nature oftheir contacts with private sector firms. Scientists were alsoqueried as to their views about the potential for conflict ofinterest and undue influence on the research agendas ofpublic universities from private sector research funds. Andthe scientists were asked about their opinions of the admin-istration of UIRRs and intellectual property and technologytransfer issues by their universities. We tape recorded andtranscribed the interviews for accuracy.

In addition to the open-ended questions, scientistswere provided several Likert-scale instruments designedto quantify their views on the characteristics of UIRRs andtheir views on the rationale and impacts of university intel-lectual property policies. Scientists were provided a listof descriptive statements about university–industry col-laborations and asked to rate the statements from NotCharacteristic to Highly Characteristic (1–7). Scientistswere also provided a list of several identified purposesor goals of intellectual property and patenting policies.They were asked to what extent they believed these goalswere important to their university by rating the statementsfrom Not Important to Very Important (1–7). Both listswere derived from a review of the literature on UIRRs withan emphasis on the agricultural biotechnology industry,and contained space for scientists to add their own items,

though none chose to do so (see Tables 1 and 2). Combiningthe answers to open-ended questions with the Likert-scaledata enables us to triangulate our interpretations of theinterviewees’ statements with the interviewees’ own quan-tified assertions.

R. Welsh et al. / Research Policy

Table 2Rank order of university scientists’ views of purpose of university IP poli-cies (1 = Not Important–7 = Very Important)

Purpose of IP policy Mean Std. Dev.

To provide licensing income for theuniversity

6.2 1.1

To encourage greater industry investmentin university research

5.6 1.5

To provide for applied research anddevelopment funding

5.4 1.5

To provide licensing income for the collegeand the department

5.4 1.6

To enhance university–industrypartnerships

5.3 1.5

To comply with federal and stategovernment policies

5.2 1.7

To ensure greater utilization of newknowledge

4.8 1.8

To avoid conflicts of interest 4.5 2To retain faculty 4.1 1.9To support regional development 4.1 1.7To ensure access to research results for the

general public4.1 1.8

To expand graduate student opportunities 4.0 1.8

I

trtutatsptK

LLoptprIua

TF

C

IPPAPIIR%

to work with the private sectornsure access to research results for

developing countries3.1 2.0

We focus on these aspects in order to gain insight intohe ongoing debates about the role of university scienceegarding societal welfare and economic development fromhe practitioner’s perspective. In this regard it is critical tonderstand the working perspective of scientists in rela-ion to their overall impression of UIRRs, but also how suchrrangements are promulgated and implemented throughhe intellectual property policies of the university. Univer-ity IP policies have been identified by both critics andromoters of UIRRs as a critical linchpin in establishinghe new academic capitalist/entrepreneurship model (e.g.,unhardt, 2004; McSherry, 2001).

In order to understand how scientists’ responses to theikert scales varied, we performed a factor analysis on theikert-scale items in Tables 1 and 2 that had average scoresf 4.6 or above. An average score of at least 4.6 (out of aossible 7.0) indicates a degree of consensus among scien-ists that the item was either characteristic of UIRRs or a

urpose of their university’s IP policies. The factor analysesesults for UIRR characteristics and Purpose of UniversityP Policies are presented in Tables 3 and 4, respectively. Wese the data collected from the open-ended interviews tossist us in interpreting our quantitative results.

able 3actor analysis of UIRR characteristics (Varimax Rotation)

haracteristic of UIRR Produc

ncreases access to new knowledge 0.814rovides new research funds −0.314rovides contact with a wider network of scientists 0.414ccelerates product development 0.455rovides new support for graduate students/postdocs 0.271ncreases access to new research tools 0.864ncreases potential for conflicts of interest 0.019estricts scientific communication among university researchers −0.497of variance 32

37 (2008) 1854–1864 1859

4. Results

4.1. Characteristics of UIRRs

To explore the extent to which these comments res-onate with our sample of interviewees, we conducted arank order analysis of scientists’ opinions on the char-acteristics of UIRRs with Likert-scale data (Table 1). Thestatements considered most characteristic of UIRRs (aver-age rating of 4.6 or above) were from highest to lowest:provides new research funds, increases access to newresearch tools, provides new support for graduate stu-dents/postdocs, provides contact with a wider networkof scientists, accelerates product development, increasesaccess to new knowledge, increases potential for tensionwith university conflict of interest policies, and restrictsscientific communication among university researchers. Ingeneral, university scientists involved with industry tend todescribe as characteristic of UIRRs the increases in accessto resources such as funding, research tools, more scientistsand new knowledge. Scientists also agree that UIRRs accel-erate product development. Less often mentioned werenegative characteristics of UIRRs such as conflict of interestsand communication restrictions.

The fact that university scientists engaged with industryview these relationships as primarily benign is not surpris-ing, and perhaps not that helpful to the development of auseful assessment framework. What may be more helpfulis what these scientists find useful about the relationshipsand what aspects appear problematic. For instance, accessto proprietary materials is nearly as valuable as funding.And the “cultural traffic” aspects such as increasing con-tact with scientists and access to new knowledge are alsoranked as very important. On the other side, publishingrestrictions or delays were not ranked above 4.5 (or even4.0), but scientists did believe that working with industryrestricted communication among scientists. This is partic-ularly noteworthy since an advantage of UIRRs is access toa wider community of scientists. Another potentially con-tradictory finding is that scientists value UIRRs because ofincreased access to industry IP and proprietary data andexpensive equipment, but also believe, to some degree, thatUIRRs are characterized by the inhibiting of material trans-

fer.

Table 3 presents the results of the factor analysisusing the higher ranked characteristics of UIRRs dis-cussed above. Three dominant factors emerged which welabel the Productivity Enhancing Factor, Scientific Interac-

tivity enhancing Scientific interaction Industry support

0.053 −0.185−0.265 0.736

0.607 0.282−0.396 0.132

0.064 0.835−0.016 0.158

0.843 −0.2800.659 0.017

19 13

1860 R. Welsh et al. / Research Policy 37 (2008) 1854–1864

Table 4Factor analysis of purposes of university IP policies (Varimax Rotation)

Purpose of policy Licensing income Industrial relations Governmental mandates

To comply with federal and government policies 0.152 −0.156 0.745To enhance university–industry partnerships −0.175 0.507 0.653To provide for applied research and development funding −0.155 0.815 −0.218To ensure greater utilization of new knowledge −0.587 −0.116 0.504

0.8780.7940.258

28

To provide licensing income for the universityTo provide licensing income for the college and the departmentTo encourage greater industry investment in university researchPercent of variance explained

tion Factor and the Industry Support Factor. Coefficientsin bold type entail primary components of the factors.The Productivity Enhancing Factor emphasizes increaseaccess to new types of new research tools and knowl-edge through university–industry interaction. Acceleratingproduct development is also a significant component,though less important. Restricting scientific communica-tion among university researchers is significant but loadsnegatively. Therefore, university scientists with high Pro-ductivity Enhancing factor scores see UIRRs positively. Thefollowing comments from scientists offer support for thisinterpretation:

• “I think, for us, it’s. . .access to techniques and knowledge[beyond just funding] that you might not otherwise haveaccess to. I know people who have worked with compa-nies, as those companies are developing a new technique,because. . .[the university scientists are] testing. . . [thefirm’s] experimental system which is being developed.At the end of the day, [the university scientists] mayget a tremendous amount of information, a lot of whichhas been funded by the industry. . .maybe not directly interms of a grant but in terms of the industry using yourmaterial in house and giving you those results. So you getaccess to results, access to data, access to techniques. . .isoften what people are interested in as well as the finan-cial. Sometimes it’s strictly money. Sometimes it’s morethan that.”

• “. . .availability of materials. Sometimes access to materi-als is a big deal, because they’re expensive. For instancesometimes you need a reagent for detecting some prod-uct and a company can send you some samples. . .”

• “. . .access to proprietary information and methods ormore advanced instrumentation than they can afford. Sooften you get access to really cool things from industry. . .I would say that’s the biggest advantage. . .and thenaccess to funding would be a second but equal advantageof collaborating with industry.”

• “The benefits are to use their technology to generate alarge amount of information that you wouldn’t be able togenerate otherwise.”

• “. . .it has made me aware of the economic considera-tions that go into the marketplace. . .I think working withthem [industry] has given me. . .a better realization of

how important team oriented research, multidisciplinaryresearch is.”

• “There is a lot of knowledge that has to reside in thosepractical people because you have to be doing it everydayto know that level of practical knowledge.

−0.109 0.0840.079 0.0260.823 0.111

24 18

• “There are a lot of great benefits beyond just the economicissues. They are two different cultures that can learn fromeach other. They have taught me – and they’ve taughteverybody the same thing, all the academic professors –how their work can have an impact, and how to makethat impact. [They have also taught us to have a] desireto make that impact more tangible. That’s what industryis doing for academics. They’re showing you how to takethe risks. Things are changing, but twenty years ago theacademic professor might have been a little bit risk averseand today they are much less.”

University scientists also believe that links with industrycan lead to the development and successful commercial-ization of a product. This is attractive because commercialsuccess is seen as having a social impact, beyond knowledgecreation and publication.

• “. . . getting information and intellectual property out intouse. I would see [that] as the primary one. Industry’s goingto do it or it’s not going to get done.”

• “I think people in the private sector and in industry havetheir finger on the pulse of what might be important tothe public, whether it might be a product that could be ofgreat benefit, that many times university researchers arenot aware of. [University researchers] don’t necessarilyrecognize a niche for a particular product.”

The Scientific Interaction Factor emphasizes contactswith a wider network of scientists, but at the same time theincreased potential for conflicts of interest and restrictionof scientific communication among university scientists.Therefore, university scientists that ranked wider scientificnetwork as characteristic also believed that this advan-tage brought with it some potential disadvantages, directlyrelated to it. This factor is a good example of how UIRRscan be advantageous and disadvantageous simultaneously;or how the beneficial aspects can also be the problematicaspects. These scientist comments are illustrative:

• “I started a company which I had for a couple years. I spenta lot of time looking at conflict of interest and trying tobe careful about that. Those statements could have been

clearer. If the faculty gets involved in industry, startingtheir own company, you have to figure out everythingyourself – about where your conflicts of interest are. It’sdoable, but I think I should have been able to get a littlemore help with that.”

h Policy

•

•

•

•

•

lva

•

•

ppfssv

R. Welsh et al. / Researc

“Well, I think you could have conflicts of interest if youdidn’t have very strongly articulated and defined poli-cies.”“. . .you can have conflicts of interest, but those can bemanaged. And this campus does a pretty good job of man-aging those.”“The classic characterization of the land-grant universityhas been a source of, quote, unbiased information. Myway of thinking, it’s those kinds of relationships [that]put a lot of pressure on that. I’ve seen it so many ways, somany subtle forms of it; it’s just very difficult to be reallyunbiased about things when you’ve got somebody that’sshoveling over a lot of dough for a particular reason.”“. . .there may be more constraints [when working withindustry partners] than what a university scientist is usedto; we’re used to open access, discussing your researchresults at meetings, publishing, talking with others aboutit. . .[depending on the research] a company can tell you,No, you can’t go to this meeting, you can’t disclose any ofthis information. . .”“Some of them are risks more than unavoidable problems,but they’re big risks. . .In some cases control over publi-cation and what gets said in it [is a risk]. I think mostuniversities try to [avoid this risk], there’s this we get our30 days or our 60 days or 90 days to review what we say,so we’re all prepared to rip and stretch. There are fairlyblatant, hopefully few, but blatant examples of informa-tion being silenced. So I think those are risks, they’re notalways outcomes but they’re very big risks.”

Several scientists indicated that communication can beimited as a result of UIRRs because industry tends to beery concerned with protecting its intellectual propertynd trade secrets.

“At the start, all seemed really good. As it develops, as theprocess goes on, industry got more protective of intel-lectual property rights. This slows down our path to theoriginal goals. There’s too much negative thinking on thecompany’s part. The scientist needs the confidence to stepup to the company and sell them on the longer term ben-efits of not being so protective. The board of investors[were afraid] to spend so much money and get no product.Everyone wants their own promoters. This is the biggestproblem on both sides, to lose sight of the goal, whichis to solve agricultural problems; they are distracted byintellectual property rights.”“. . .Particularly if you’re [working with] industries [with]intellectual property, they’re inherently secretive. Theydon’t want to share. They’re always afraid that someone’sgoing to come in and take their idea and get it down tothe patent office before they do.”

The final UIRR characteristic factor is the Industry Sup-ort Factor and is composed of two funding variables:

rovides new research funds and provides new supportor graduate students and postdocs. In general, universitycientists viewed industry as a source of critical fundingupport for their research, although the sizes of the indi-idual grants were often not very large. For example,

37 (2008) 1854–1864 1861

• “Occasionally, too, they [private sector firms] do provideresearch support which, in my case, is what happened.They provided a lot of research support to the lab andthat was a huge benefit to us. I had two collaborations.One was about $600,000 of research support. The otherwas about $300,000. It adds up.”

• And private sector firms were important for monetaryand intellectual support for graduate students and otherkey personnel.

• “. . .it’s a way—on the university side, to bring in somemore funding for a particular project and many times thatfunding is used to fund graduate students, so I think that’sa great thing because it allows students to come here andwork on fellowships or on stipends from the company,and otherwise maybe they couldn’t do their research herebecause the public funding wouldn’t have been here forthem to come here for graduate school. So in a way it helpsto promote the graduate education, and then many timesthat actually then opens up the opportunity for those stu-dents to go after they’re finished for jobs, so it’s a goodway for the placement of your students.”

• “I think one of the biggest benefits comes in the cross-educational process. The students see what’s going on inindustry, and industry can therefore recruit outstandingtalent.”

To discern if socio-demographic of the scientists inter-viewed are correlated with relative adherence to theUIRR Characteristic factors, we saved the factor scores asvariables and performed a correlation analysis betweenthe three sets of scores and the following variables: sex(1 = male), academic rank (full professor = 1 and all oth-ers = 0) and year earned Ph.D. The analysis resulted in nosignificant correlations (result not presented in tables).

4.2. Purpose of university IP policies

Turning to university scientists’ views of their univer-sities’ IP policies (Table 2), we find that the items rankedabove 4.6 are: provide licensing income for the univer-sity, encourage greater industry investment in universityresearch, provide for applied research and developmentfunding, provide licensing income for the college and thedepartment, enhance university–industry partnerships,comply with federal and state government policies, andensure greater utilization of new knowledge. Therefore,university scientists view IP policies primarily through amaterialist lens. The top four ranked items concern gener-ating revenue and investment for the university. Complyingwith federal and state policies and ensuring greater uti-lization of new knowledge were considered importantpurposes (5.2 and 4.8 scores, respectively), but less so thanthe revenue-oriented items. Public-goods goals such assupporting regional development and ensuring access toresearch results for the public were ranked relatively low bythe scientists. This is consistent with the findings of Thursby

et al. (2001) that income generated by university Technol-ogy Transfer Offices is the primary measure of their success.

In Table 4 a factor analysis of the higher ranked itemsrelating to the purposes of university IP policies resultedin three factors: Licensing Income Factor, Industrial Rela-

h Policy

1862 R. Welsh et al. / Researc

tions Factor and the Governmental Mandate Factor. Thefirst factor is composed of providing licensing income forthe university, the college, and the department. The secondfactor entails the enhancement of UIRRs, finding appliedresearch funding and encouraging industry investment.And the third factor primarily emphasizes complying withgovernmental policies and enhancing university–industrypartnerships, with lesser but still significant, emphasis onensuring greater utilization of new knowledge.

The interview data provides a variety of scientist opin-ions on the purposes of university IP policies. Often thoseopinions contain conflicting interpretations. Regarding theLicensing Income factor, the most often-mentioned moti-vation attributed to the administration of university IP isto generate revenue. However, this motivation is furtherdifferentiated by university scientists within a norma-tive context. At times, scientists use the term “cynical” todescribe the interests of the university administration. Theimplication is that the scientists think universities shouldhave higher principles, but in reality are primarily inter-ested in revenue generation.

• “Sort of obvious. They want to have ownership, so theycan launch this and make money. I don’t think they havealtruistic motives that I know of. Would there be any otherreason?”

• “Making money for the university. It should be limitingtechnology to the public sector so the public should useit.”

• “Money. I don’t mean to sound cynical and perhaps itcomes out that way. I think, again, the university. . .quitefrankly, it is a little cynical.”

Other scientists portray revenue generation as legit-imate, and claim that IP policies are a necessary wayof “protecting the public investment” and an importantattempt to partially recoup research costs.

• “I think in general the goals are to make sure that every-thing is done right, that the university is protected fortheir investment, because they are investing a lot withrespect to facilities, professors’ time and energy, and soon, so it’s only right”

• “The primary goal is to provide another source of fundingfor the university. It’s a valuable goal because we need it.”

When asked to consider the issues that comprise ourIndustrial Relations and Governmental Mandates factors,the university scientists were again conflicted as to how thevariables manifest themselves. For example, almost uni-versally, the scientists use the term “protect” to define thepurpose of IP policies resulting from Bayh-Dole and otherlegislation. And though protection always refers to estab-lishing clear university ownership of university generatedtechnologies or knowledge, the interpretation of why suchprotection is established by the university varies greatly

across scientists.

For example, a number of university scientists arguedthat establishing definitive ownership of a discovery orinvention in response to government policies enhancesuniversity–industry partnerships and ensures greater uti-

37 (2008) 1854–1864

lization of new knowledge because industrial interests willnow find that discovery or knowledge valuable. In this wayIP policies are developed to attract industry to the univer-sity in order to more quickly make products available in themarket, or provide increased applied research and develop-ment funding. Patenting enables licensing which providesprotection for private sector firms to invest in the furtherdevelopment and commercialization of technologies. Thisprocess brings goods to the public through the market andultimately benefits the public. As some put it:

• “[The purpose of university IP policy]. . .is to try and makeit attractive for the private sector to use things that aredeveloped in the university as opposed to sit on a shelfor sit in a research publication.”

• “I think that the primary goal is to do what I’ve been say-ing and that is to try to protect technologies that wouldbe useful to the public. When I say protect, I don’t meanprotect in the sense that the university is trying to keep allthe profits that come from that technology to themselves.When I say protect, I mean that that technology is recog-nized by the corporate world as being something that’sdevelopable. . .if that’s a word. Without that kind of intel-lectual property protection, corporate partners just aren’tinterested in looking at the technology because what’s tostop their competitor from coming in and doing the samething quicker and better than them and having wastedtheir time.”

On the other hand, university scientists also argued thatIP policies are needed to protect the academic freedom ofthe scientists. Patenting establishes clear ownership linessuch that universities and scientists can feel free to workwithout being sued by private sector firms for infringement.This includes “. . .freedom to publish, in regular outlets” asone scientist argues. That is, strong and effective univer-sity IP policies enhance university–industry partnershipsand ensure greater utilization of knowledge by protectingscientists from industry: For instance,

• “. . .to protect the breeder and the institution. It is, I don’twant to say a jungle out there, but it’s a new commercialworld that plant breeding is now in. If we’re going to dealwith Monsanto, we need somebody that knows the legalaspects of every word that’s going into the document.”

• “I think their goals; their obligation is to file the patents.And from my standpoint, what I’m looking for as a scien-tist is that can be done in a way that allows me to continueto do my research in a normal way and not be adverselyaffected.”

• “To protect me, number one, and number two to protectthe university. And by “protect” I don’t mean by patentprotect, I mean just protect me from being sued or some-thing, or protect me from not being able to publish, andso retain my academic freedom.”

Therefore, as with UIRRs in general, the same aspectsof IP policies, or outcomes related to their adoption, canbe viewed as both beneficial and problematic. IP poli-cies enable universities to raise revenue which can reflecteither a cynical shift toward attempts at accumulation more

h Policy

sttdttftpp

sesrt(Pacptmfo

5

uIefbtIbiucatattiftcueomvooat

cu

R. Welsh et al. / Researc

uited to private sector firms, or a legitimate vehicle for pro-ecting university investments and defraying costs. Doeshe establishment of clear ownership by the university overiscoveries and inventions and then licensing them, pro-ect the investments of private sector firms and thus attracthem and their dollars? Or does it protect the academicreedom of the university researcher who now must, morehan before, operate in “the jungle” of university–industryartnerships? The scientists seem to recognize that theolicies should do all of these things.

To discern if socio-demographic characteristics of thecientists interviewed are correlated with relative adher-nce to the Purpose of University IP Policy factors, weaved the factor scores as variables and performed a cor-elation analysis between the three sets of scores andhe following variables: sex (1 = male), academic rankfull professor = 1 and all others = 0) and year earnedh.D. Only sex had a modest positive correlation withdherence to the Governmental Mandates factor (Pearsonorrelation = 0.21; 2-tailed significance = 0.049—result notresented in tables). Male scientists are slightly more likelyo believe university IP policies are in place to meet govern-

ent mandates. In general, relative adherence to any of theactors is not related strongly to a scientist’s gender, rankr generation.

. Discussion and conclusions

From interviews with 84 biological scientists at nineniversities we find scientists view UIRRs and university

P policies in complex and often conflicting ways. Forxample, university scientists believe UIRRs are valuableor increasing contact with scientists, but are problematicecause working with industry can restrict communica-ion among scientists. Also scientists believe universityP policies should shield their work from opportunisticehavior and at the same time be designed to attract

ndustry partners. In addition scientists believe universitiesse their IP policies primarily as revenue raising vehi-les and secondarily to address public good issues suchs technology transfer. Therefore, attempting to untanglehe effects of enhanced university–industry relationshipsnd increased proprietary behavior on the part of universi-ies is complicated. The characteristics of UIRRs that makehem attractive are also a cause for concern. Contacts withndustry can benefit university science through boosts inunding and access to useful data, equipment and facili-ies. At the same time, this funding and access has limitedommunication and increased conflicts of interest amongniversity scientists. Likewise, university IP policies arextremely critical vehicles for the protection and durabilityf scientist academic freedom and technology develop-ent, but are also vehicles that can lead universities to

alue revenue gains over other more public goods orientedutcomes. Therefore, the increasing number and intensityf university–industry partnerships and the emergence of

cademic capitalism bring with them the paradox of oppor-unities and problems.

In addition, the different scientific behavioral outcomesatalogued and analyzed by Owen-Smith and Powell (2001)nder the academic capitalism rubric illustrate that the

37 (2008) 1854–1864 1863

effects of these changes in policies and practices is not eas-ily predicted. However, as we increase our understandingof the actors involved and the potential outcomes, it mightbe possible to craft policy that selects for the outcomes thepublic deems most attractive. For instance, academic capi-talism and UIRRs might prove more useful if policies are putin place to promote the “reluctant entrepreneur” scientist. Ifsuccessful, universities might embrace reluctantly the shifttoward academic capitalism, and use the tools that are asso-ciated with it to enhance their more traditional mission ofproducing public goods.

As Rudy et al. (2007) have pointed out land-grantuniversities have, since their inception, grappled with ten-sions between promoting public goods, and facilitatingprivate accumulation. The public policy changes that havepromoted UIRRs and the commercialization of universityresearch are only the latest, but arguably the most far-reaching, iteration in this relationship. The tension lies inthe fact that public-interest science is only valuable asa generator of economic development if it maintains adegree of autonomy from industrial interests (Glenna et al.,2007a,b; Swedberg, 1998).

Our research findings indicate that university scientistsbelieve university–industry partnerships are valuable inmany ways. However, they also recognize that the positiveoutcomes are most likely when strong conflict of inter-est policies are enforced, and when university IP policiesare constructed deliberately to establish clear universityownership over important discoveries. They also tend torecognize that the university should de-emphasize the rev-enue stream to be generated from ownership of a discoveryif it is to maintain its reputation as a reliable source ofinformation for both private firms and the general pub-lic. Rather, the focus should be on making the discoveryavailable for commercialization and protecting the abilityof university scientists to continue to refine and enhancetheir knowledge of the discovery, as well as understand-ing the functions and impacts of potential products derivedfrom it.

How universities navigate this terrain is important totheir success, but also to the economic growth and develop-ment in general within the new knowledge economy. In factDavid (1994, p. 66) argues that university science is to the21st Century knowledge economy what the machine toolsector was to the economies of the 19th and early 20th cen-turies. Maximizing the public benefits of this transforma-tion will require intelligent and creative policies to enhanceuniversity interactions with private sector firms whileprotecting the autonomy and freedom of operation of uni-versity scientists. Constructing effective policies at the uni-versity level is fundamental for realizing beneficial aspectsof UIRRs and for avoiding problems. At the same time, how-ever, developing policies on critical aspects of universityoperations such as IP protection is inherently complex.

Acknowledgements

This research was supported by the CooperativeState Research, Education and Extension Service, U.S.Department of Agriculture, under Agreement No. 2001-52100-11217. An earlier version of this paper was presented

h Policy

1864 R. Welsh et al. / Researc

at the Economic Sociology and Technology ConferenceSeptember 23–24, 2005 at Cornell University, Ithaca, NY.

References

Breschi S., Lissoni, F., Montobbio, F., 2005. From Publishing To Patenting: DoProductive Scientists Turn Into Academic Inventors? Revue d’EconomieIndustrielle, no 110, 2eme trimester.

Busch, L., Lacy, W.B., Burkhardt, J., Lacy, L.R., 1991. Plants, Power, and Profit:Social, Economic, and Ethical Consequences of the New Biotechnolo-gies. Blackwell Publishers, Cambridge, MA.

Busch, L., Allison, R., Harris, C., Rudy, A., Shaw, B.T., Eyck, T.T., Coppin, D.,Konefal, J., Oliver with, C., Fairweather, J., 2004. External Review ofthe Collaborative Research Agreement Between Novartis AgriculturalDiscovery Institute and The Regents of the University of California.Institutive for Food and Agricultural Standards, Michigan State Uni-versity, East Lansing, MI.

Buttel, F.H., Cowan, J.T., Kenney, M., Kloppenburg Jr., J., 1984. Biotechnologyin agriculture: the political economy of agribusiness reorganizationand industry–university relationships. In: Harry, K., Schwarzweller(Eds.), Research in Rural Sociology and Development. Jai Press Inc.,Greenwich, CT, pp. 315–348.

Caswell, M., Day-Rubenstein, K., 2006. Agricultural research and devel-opment. In: Wiebe, K., Gollehon, N. (Eds.), Agricultural Resources,Environmental Indicators. Economic Information Bulletin 16. Eco-nomic Research Service/U.S., Department of Agriculture, pp. 59–64.

Charles, D., 2001. Lord of the Harvest: Biotech, Big Money and the Futureof Food. Perseus Publishing, New York.

Charmaz, K., 2002. Qualitative interviewing and grounded theory anal-ysis. In: Gubrium’s, J.F., Holstein’s, J.A. (Eds.), Handbook of InterviewResearch: Context and Method. Sage Publications, pp. 675–694.

Cohen, W.M., Nelson, R.R., Walsh, J.P., 2002. Links and impacts: The influ-ence of public research on industrial R&D. Management Science 48,1–23.

David, P.L., 1994. In: Ove Grandstand (Ed.), Positive feedbacks and researchproductivity in science: Reopening another black box. Economics ofTechnology North Holland Press, pp. 65–89.

Day-Rubenstein, K., 2003. Transferring public research: the patent licens-ing mechanisms in agriculture. Journal of Technology Transfer 28,111–130.

Franzoni, C., 2006. Do Scientists Get Fundamental Research Ideas by Solv-ing Practical Problems?, SSRN working paper #935006.

Glenna, L.L., Lacy, W.B., Welsh, R., Biscotti, D., 2007a. University admin-istrators, agricultural biotechnology, and academic capitalism. TheSociological Quarterly 48, 141–163.

Glenna, L.L., Welsh, R., Lacy, W.B., Biscotti, D., 2007b. Industry perceptionsof university-industry relationships related to agricultural biotechnol-ogy research. Rural Sociology 72, 608–631.

Hackett, E.J., 2001. Science as a Vocation in the 1990s: the changing organi-zational cultures of academic science. In: Croissant, J., Restivo, S. (Eds.),Degrees of Compromise: Industrial Interests and Academic Values.State University of New York Press, Albany, NY, pp. 101–138.

Hicks, D., Hamilton, K., 1999. Does university–industry collaborationadversely affect university research? Issues in Science and Technology(Summer), 74–75.

Harper, D., 2001. Changing Works: Vision of a lost agriculture. Universityof Chicago Press.

Jensen, R., Thursby, M., 2001. Proofs and prototypes for sale: the licensingof university inventions. The American Economic Review 91, 24–59.

Kenney, M., 1986. Biotechnology: The University–Industrial Complex. YaleUniversity Press, New Haven, CT.

Kirp, D.L., Popp-Berman, E., 2003. A good deal of collaboration: the Uni-versity of California-Berkeley. In: David, L., Kirp Shakespeare (Eds.),Einstein and the Bottom Line: The Marketing of Higher Education.Harvard University Press, pp. 207–220.

Kleinman, D.L., Vallas, S.P., 2001. Science, capitalism, and the rise of the‘Knowledge Worker’: the changing structure of knowledge productionin the United States. Theory and Society 30, 451–492.

Kloppenburg, J., 2004. First the Seed: The Political Economy of PlantBiotechnology: 1942–2000. University of Wisconsin Press, Madison.

Krimsky, S., 2003. Science in the Private Interest: Has the Lure of ProfitsCorrupted Biomedical Research? Rowman and Littlefield Publishers.

Kunhardt, E.E., 2004. Necessity as the Mother of Tenure? Times,New York, www.nytimes.com/2004/12/14/opinion/14kunhardt.html(December).

Landry, R., Amara, N., Rherrad, I., 2006. Why are some universityresearchers more likely to create spin-offs than others? Evidence fromCanadian universities. Research Policy 35, 1599–1615.

37 (2008) 1854–1864

Lee, Y.S., 2000. The sustainability of university-industry research collab-oration: an empirical assessment. Journal of Technology Transfer 25,111–133.

Mansfield, E., 1998. Academic Research and Industrial Innovation: anupdate of empirical findings. Research Policy 26 (Fall), 773–776.

Mansfield, E., Lee, J., 1996. The modern university: contributor to industrialinnovation and recipient of industrial R&D support. Research Policy 25,1047–1058.

McMichael, P., 2000. Development and Social Change: A Global Perspec-tive, 2nd ed. Pine Forge Press.

McSherry, C., 2001. Who Owns Academic Work? Battling for Control ofIntellectual Property. Harvard University Press, Cambridge, MA.

Meyers, M., 2006. Are patenting scientists the better scholars? Anexploratory comparison of inventor–authors with their non-inventingpeers in nano-science and technology. Research Policy 35, 1646–1662.

Mowery, D.C., Nelson, R.R., Sampat, B.N., Ziedonis, A.A., 2004. Ivory Towerand Industrial Innovation: University-Industry Technology TransferBefore and After the Bayh-Dole Act. Stanford University Press.

Mueller, P., 2006. Exploring the knowledge filter: how entrepreneur-ship and university–industry relationships drive economic growth.Research Policy 35, 1499–1508.

Murray, F., Stern, S., 2007. Do formal intellectual property rights hinderthe free flow of scientific knowledge? An empirical test of the anti-commons hypothesis. Journal of Economic Behavior and Organization63, 648–687.

Nelsen, L.L., 1991. The lifeblood of biotechnology: university-industrytechnology transfer. In: Dana Ono, R. (Ed.), The Business of Biotechnol-ogy: From the Bench to the Street. Butterworth-Heinemann, Boston,pp. 39–75.

Owen-Smith, J., Powell, W.W., 2001. In: Steven Vallas (Ed.), Careers andContradictions: Faculty Responses to the Transformation of Knowl-edge and Its Uses in the Life Sciences, vol. 10. Research in the Sociologyof Work, pp. 109–140.

Patton, M.Q., 1990. Qualitative Evaluation and Research Methods. SagePublications.

Rudy, A., Coppin, D., Konefal, J., Shaw, B.T., Ten Eyck, T., Harris, C., Busch, L.,2007. Universities in the Age of Corporate Science. Temple UniversityPress, Philadelphia, PA.

Sampat, B.N., 2006. Patenting and U.S. academic research in the 20thcentury: the world before and after Bayh-Dole. Research Policy 35,772–789.

Shin, T., Lamy, E., 2006. Paths of commercial knowledge: Forms and con-sequences of university–enterprise synergy in scientist-sponsoredfirms. Research Policy 35, 1465–1476.

Slaughter, S., Leslie, L.L., 1997. Academic Capitalism: Politics, Policies, andthe Entrepreneurial University. The Johns Hopkins University Press,Baltimore and London.

Slaughter, S., Rhoades, G., 1996. The emergence of a competitivenessresearch and development policy coalition and the commercializationof academic science and technology. Science, Technology and HumanValues 21, 303–339.

Slaughter, S., Rhoades, G., 2004. Academic Capitalism and the New Econ-omy: Markets. The John Hopkins University Press, State and HigherEducation Baltimore, MD.

Shane, S., 2004. Academic Entrepreneurship: University Spin-offs andWealth Creation. Edward Elgar Publishers.

Swedberg, R., 1998. Max Weber and the Idea of Economic Sociology.Princeton University Press.

Thursby, J.G., Jensen, R., Thursby, M.C., 2001. Incentives, characteristics andoutcomes of university licensing: a survey of major U.S. universities.Journal of Technology Transfer 26, 59–72.

Thursby, J.G., Thursby, M.C., 2000. Industry perspectives on licensing uni-versity technologies: Sources and problems. Journal of the Associationof University Technology Managers 12, 9–21.

Thursby, J.G., Thursby, M.C., 2002. Who is selling the Ivory Tower? Sourcesof growth in university licensing. Management Science 48, 90–104.

Thursby, J.G., Thursby, M.C., 2003. Industry/University licensing: charac-teristics, concerns and uses from the perspective of the buyer. Journalof Technology Transfer 28, 207–213.

Vallas, S., Kleinman, D., 2007. Contradiction, convergence and theknowledge economy: the confluence of academic and commercial

biotechnology. Socio-Economic Review 6, 283–311.

Zucker, L.G., Dary, M.R., 1998. Entrepreneurs, star scientists, and biotech-nology. NBER Reporter Fall, pp. 7–11.

Zucker, L.G., Darby, M.R., Armstrong, J.S., 2002. Commercializing knowl-edge: university science, knowledge capture, and firm performancein biotechnology. Management Science 48, 138–153.