University of Oklahoma
This essay identifies three notable trends in recent science and technology policy research. By analyzing the keywords listed within published scholarship from 2010-2012, a predominant portion of articles focuses on universities, patenting, and innovation policy models. Scholars have gained some insight into these processes by focusing on how collaborations between the government, universities, and industry impact technological outcomes. However, data and measurement issues have limited research in this area.
The purpose of this review article is to provide a summary of recent trends in scholarship within the field of science and technology (S&T) policy. By providing a focus on recent literature, this essay gives scholars a quick overview and examination of current research directions within the subfield. I begin with a discussion of my methodology for identifying the boundaries of the field and my method for selecting articles to be included in this review. I examine articles published within the past few years (primarily 2010-2012) in the top S&T and public policy journals. Next,given the wide range of topics that fall within S&T policy, I begin by presenting a broad overview of recent research trends by analyzing the keywords listed within published articles. After providing an expansive perspective of the field, I then give a more detailed summary of published research that examines questions in the three most popular keyword-topic areas: universities, patents and innovation.
In order to organize and assess recent research in S&T policy, a more primary question must be addressed - is there a separate and definable subfield that can be called S&T policy? If so, what are the common theoretical or methodological questions that unite scholars within this area? Despite the recent emergence of several S&T policy review articles (Jasanoff 2010; Martin, Nightingale and Yegros-Yegros 2012; Martin 2012; Fagerberg, Fosaas, & Sapprasert 2012; Van Den Besselaar 2001), identifying the core research questions that shape the boundaries of the subfield of S&T policy has proven to be a difficult task. Scholars studying S&T policies come from a variety of disciplinary perspectives and utilize multiple methodological approaches. Unfortunately, the evolution of the field has resulted in vague terminology (see Bozeman 2001), making it sometimes unclear how various pieces of scholarship are interconnected. For example, some scholars use science policy, research policy, technology policy, and innovation policy interchangeably, yet others see these as quite distinctive concepts. This results in a field consisting of a wide variety of theoretical approaches and research orientations.
For example, within the literature, there is disagreement in regards to whether science and technology studies (STS), science and innovation studies, and science and technology policy research more generally, constitute as distinct disciplinary fields. STS is a subfield that includes scholars from multiple disciplines: anthropology, economics, history, law, philosophy, psychology, sociology and political science, and takes a more normative view of S&T policy issues. Jasanoff (2010) describes the field as a merger of two streams of literature: research that investigates the nature and practice of science, and research that investigates the impacts and control of S&T in regards to the potential risks of S&T to society. However, this merger did not include an integration of core assumptions or methods across S&T studies. Jasanoff argues that while the analysis of S&T policy has involved a wide range of questions from multiple disciplines, the central unifying feature has traditionally been the subject of study (for example, investments in S&T) and not specific theoretical commitments. STS includes, but is not limited to, questions such as: What are the social processes by which scientific and technical knowledge is created? How do people evaluate scientific and technical information? What is the role of science and technology in society? Finally, what are the normative implications of scientific and technological issues on society, culture and politics? Another important component within STS includes scientometric studies (see Martin et al. 2012). This is often referred to as the quantitative component of STS and includes the study of science indicators. Scientometric studies focus on methods for analyzing and measuring scientific research. Bibliometrics is the most commonly utilized methodological approach in scientometric studies and measures the impact of scientific publications.
A second subfield is science and innovation studies. Martin (2012) summarizes the evolution of science policy and innovation studies (SPIS), which he considers to be a separate field from STS. Martin focuses on policy, economic, management, organizational and some sociological research regarding national science policy.SPIS developed from a focus on the management and economics of science, technology and innovation. Over time, these concepts have fallen under the short title, “innovation studies.”SPIS includes, but is not limited to, questions such as: what is the role of innovation on the economy? How has technology and innovation impacted industrial development and economic growth? What contributes to new technological developments? What attributes of actors (in particular, firms)influence technological outcomes? Finally, how is basic research converted to usable knowledge? What are the economic, political and theoretical aspects of innovation?
A third subfield under the umbrella of “S&T policy” includes scholarship studying policy problems that span a broad spectrum of substantive issues (biotechnology, stem cell research, nanotechnology, etc.). Across these substantive issue areas, S&T scholars study a wide range of questions about regulatory processes, funding, expertise, scientists, public engagement, ethics, etc. Finally, it is imperative to note that some scholars also study the role of S&T in making policy decisions in other subfields, such as: environmental policy, energy and natural resource policy, defense and national security policy, health policy, etc. All of these alternative research projects have scientific or technological components, but are excluded in my discussion because explaining S&T policy is not their central focus.
I recognize that the distinctions between these subfields are a topic of major debate amongst some scholars within the discipline (see Leydesdorff 1989; Van Den Besselaar 2000, 2001). Although this debate is a critical issue to defining the boundaries of the field, this debate cannot be resolved within the limits of this essay. While there are some distinguishable publication patterns, wherein particular types of questions are published within specific journals (see Van Den Besselaar 2001), there are also areas of overlap (see my methodology section below). It is not clear that the research questions identified within specific focus areas are markedly different enough to warrant identification as entirely separate disciplinary fields- particularly given that in some recent publications, insights from STS are being incorporated into studies on the management of science policy (Garforth & Stockelova 2011; Garforth 2012; Parker & Crona 2012). While I agree that these areas of specialty utilize different theoretical or methodological approaches, I believe their common ground is found in their effort to contribute generally to questions in regard to national S&Tpolicies within the U.S. and abroad. Therefore, for the purposes of thisreview article, I do not treat these as separate subfields, but anchor my analysis on including all recent publications that explain S&T policies.
However, it is important to note that some scholars have conducted limited analyses to examine this question of a scholarly divide within the field and have argued that most of the apparent divisions within the S&T academic community are methodological in nature (Jasanoff 2010; Martin, Nightingale and Yegros-Yegros 2012; Martin 2012). One example of this division includes a divide between qualitative research on the production of scientific knowledge and quantitative, scientometric studies (Martin, Nightingale and Yegros-Yegros 2012; Besselaar 2000; Leydesdorff 1989; Besselaar 2001). Peter Van Den Besselaar (2000) found that while on the one hand there is a stronger segregation between qualitative and quantitative research within S&T studies, but on the other hand, there is a greater integration between scientometerics and S&T policy studies. However, this integration was mainly in regards to evaluation and performance studies. In addition, Van Den Besselaar (2001) also identified that while some scholars work as specialists in a subfield, some S&T policy authors (and their affiliated institutions) can be classified as generalists, because their published research cut across several of these categories.
In summary, although there have been a few recent attempts to review research in S&T policy, these reviews have been mainly expansive overviews that attempt to summarize the historical development of the field. These articles (Jasanoff 2010; Martin, Nightingale and Yegros-Yegros 2012; Martin 2012; Fagerberg, Fosaas, & Sapprasert 2012;Van Den Besselaar 2001) have identified the broad topics and important seminal works within the field, but do not identify the most recent research questions currently gaining scholars’ attention. In order to provide a systematic and objective overview of current research directions within the field, I collected and analyzed the keywords provided within selected articles published between 2010-2012 in 17 journals. I describe my methodology more fully in the next section.
Every review must bound its phenomena and reviewing S&T policy research presents a unique challenge, given that scholarship is published so widely across a variety of venues ranging from technology-specific journals, international technical journals, public policy journals, and various scientific journals. For example, Martin, Nighttingale, & Yegros-Yegros (2012) identified the top 155 core contributions of STS since the 1960s and found that they were cited in approximately 6000 journals covering a wide range of research areas. For the purposes of this research review, I employed multiple strategies that I believe provide a strong and semi-exhaustive list of the key publications in S&T policy. While recognizing that I may have excluded some articles published in specialty journals, I analyze articles published in 17 core S&T policy or general public policy journals. I have noted the journals’ 2012 impact factor (IF) in parentheses.1
After identifying these key journals, I only included articles in my review that were specifically about S&T policy and fall within one of the categories described in the previous section. I did not include book reviews, commentaries, opinion pieces, or editorial-introductory articles. For each article, I collected the keywords provided by the authors, as well as the authors’ geographic location. I examined a total of 668 articles, written by 1,423 authors.2 Ten articles did not have keywords and 15 articles did not have geographical affiliations of the author. I collected 3,282 keywords3, and used a word cloud of these keywords to identify trends within these publications.4 In the next section, I begin by providing a general overview of these trends across the entire collection of articles. After providing a wide snapshot of the field, I then give a more detailed summary of published research that examines questions in the three most popular keyword-topic areas: innovation; universities; and patents. However, before talking about these general trends, I quickly provide some basic geographic information about S&T policy scholars.
Because this review article is published within the Policy Studies Journal’s Public Policy Yearbook, I thought it would be interesting to present the geographic location and collaborative behavior of S&T policy scholars. Moreover, a growing body of literature within S&T examines whether scientific productivity is influenced by a scientist’s credentials, environment, or networking activities (see Hoekman, Frenken & Tijssen 2010; Caro, Catialdi & Schifanella 2012). I collected information on the geographic location of scholars’ listed affiliations in each of the published articles in my sample. Within the sample of 1,423 authors, S&T policy scholarship comes from institutions within 58 countries.5 The greatest proportion of scholarship comes from the USA (21%) and the United Kingdom (12%). Figure 1 shows the top fifteen countries that scholars’ identified as the location of their institutional affiliation. As the figure shows, 80% of scholars work at institutions within these 15 countries, while 20% work across the other 43 countries. Finally, in regards to the collaborative behavior of S&T policy scholars, I collected data on the number of authors that are identified on each of the 668 articles. Over half of the articles were sole-authored (235) or written by two authors (220). The remaining 32% of articles were written by the collaboration of 3 or more authors: 136 articles by 3 authors; 48 articles by 4 authors; 29 articles by 5 or more authors.
An examination of the keywords provided by scholars to classify the focus of their research articles offers us an interesting view into the current trends within S&T policy research. Figure 2 is a word cloud that was created in Tagxedo6 with 3,282 total keywords. The more prominent a word is in the word cloud (shown as a larger text size) means that the word or phrase appeared more frequently in the data’s text. Similar keywords were coded to be identical so that they weren’t misrepresented as different terms. For example, indicator and indicators were coded in its plural form. Genetically modified organisms and GMOs were coded as its acronym. Finally, in some cases, terms that were different words but referred to the same general concept were coded to be identical: for example, impact assessment and impact evaluation. The parameters within the Tagxedo program
were set to maintain an accurate portrayal of the relative proportion of the frequencies of each keyword. Furthermore, Tagxedo removes common words and stop words such as: “is”, “are”, “do”, etc. The graphic in Figure 2 lists the top 200 most frequent terms or phrases. I cross-referenced the word counts within the program with the dataset to ensure its reliability.
The benefit of using a word cloud is that we can graphically show numerical data in such a way to easily interpret or identify patterns within the tabular data. I recognize that frequency does not necessarily equal impact, but it does indicate a form of importance if scholars are repeatedly referencing a particular area of research. The word cloud shows that the top three keywords are innovation, universities, and patents. Looking a little broader, the top twenty unique keywords include (frequencies denoted in parentheses):
Several of the published articles in S&T, between 2010-2012, focus on important concepts in public policy scholarship such as: institutions and actors, including, universities and businesses. In addition, these articles are interested in topics such as governance, collaboration, expertise, gender and networks. Furthermore, several articles are focused on methodological topics including citation analysis or bibliometric analysis. Finally, the word cloud shows that the most popular substantive areas include biotechnology and nanotechnology. The next section describes in greater detail the research questions being investigated within articles containing the three most popular keywords: innovation, universities and patents. Given the connection and often overlap between university issues and patents, I begin with these two areas and finish with a discussion on the broader concept of innovation. It is important to note that while the goal of this review article is to focus on trends over the last few years, most of the current research is continuing to examine questions that have been asked by scholars for several decades. In order to situate the following discussion in its context, it is occasionally necessary to cite scholarship that precedes the 2010-2012 time frame. For each section, I begin with a very brief presentation of the context to which these research questions speak, and then present some of the main theoretical questions being addressed by scholars in the last few years.
The word cloud in Figure 3 shows 276 unique keywords that coincided with the keyword “university”. Unlike Figure 2, these keywords were not collapsed into similar terms, in order to preserve insight into the variety of areas being studied. This came from 53 articles that were written by 133 authors. Most of these authors are located within the United States (26), United Kingdom (14), Italy (13), Spain (11), Mexico (10) and China (9). Through a closer examination of this subset of articles, it is apparent that the most common coinciding keywords include: research, development, collaboration, industry relationships, industry interactions, commercialization and technology transfer offices. Using these terms as a guide, I briefly summarize the core developments within these articles below. Within S&T policy research, scholars have pointed to two primary policy changes (dating back to the 1980s) that have influenced university activities: performance based funding and patent law reforms (Fisher & Rubenson 2010; Doern & Stony 2009). Because the next section focuses in on patent laws, the following discussion will focus primarily on the effects of funding policies on university knowledge transfer.
Post-WWII research policy in the United States provided scientists at public universities or government laboratories with generous public funding and immense flexibility in the scientific endeavors that they pursued. In the postwar period, the traditional linear model, which describes a process for connecting scientific knowledge, technological development, and economic growth was the primary way for understanding how governmental support for science research should be structured. The historical underpinnings of the linear model, which are often traced back to arguments made by Vannevar Bush in Science: The Endless Frontier (1945)(see Godin 2006), not only advocated for an ideal of ‘pure’ or ‘basic’ science7
to precede ‘applied’ research, but also justified the necessity of major investments in ‘basic’ science research to be accomplished primarily through universities.8 Although S&T research takes place in both government labs and industrial organizations, U.S. policymakers still hold the view that the strength and success of the university system is the key to S&T success (USHSSTC 2012, 4). In 2009, academic institutions were responsible for carrying out 53% of all funded basic research within the United States (USHSSTC 2012).
However, over the past several decades, the relationship between the government, university and industry actors has changed (Leydesdorff & Etzkowitz 2001; Upham & Small 2010). This has resulted in part because of a growing body of literature that challenged the postwar framework. First, several scholars argue that the relationship between S&T is more dynamic than what the linear model described (i.e. Stokes 1997). Second, the proliferation of policy models such as the National Innovation Systems (NIS) model (Freeman1987; Lundvall 1992; Nelson 1993), laid the foundation for connecting economic performance with the technological capabilities of the institutions within a nation (Mokyr 2002; North 2005; Castellacci 2008; Fagerberg & Srhole 2008; Filippetti & Archibugi 2010; Zhao & Guan 2012). Theseideas had a strong influence on transformations in public funding mechanisms for science- which now require justifications for funding research to be connected with applied uses and/or economic benefits to society.
Therefore, science policies in the current period are less concerned with providing research autonomy and more concerned with public accountability, as evident by requirements that research programs show the value of their contributions in order to secure funding (Demeritt 2000). In fact, the concept of knowledge transfer, which recognizes scientific knowledge as a significant driver of economic growth, is now the central argument for contemporary university policies for research.9 These recent shifts in university research policies have been attributed to several structural changes in the funding application process. During the 1980s, the US experienced a proliferation of programs to build university research centers to bridge academic research, education and industrial innovation (Ponomariov & Boardman 2010). As another example, dating back to 1993, policy changes under the Government Performance and Results Act (GPRA) include requiring impact statements as a justification for research grants funded by the government, with similar adoptions occurring in the United Kingdom, Asia and the European Union (Demeritt 2010). Structural changes in universities are also evident by the massive growth in creating technology transfer offices. In 1980, there were about 25 technology transfer offices in the US, increasing to approximately 230 in 2004 (Thursby & Thursby 2011a). Finally, a continued movement toward performance based funding is evident in recent policy changes as well. One recent notable policy change that has been proposed with the intent to encourage use-inspired research is innovation inducement prizes (see Williams 2012). In 2010, the America COMPETES Reauthorization Act of 2010 was passed and provided all federal agencies with the authority to offer innovation inducement prizes. Agenciescan advertise specific problems and provide open calls for submissions for applied research that can serve as potential solutions to those problems.
These policy changes that emphasize applied research have generally resulted in increased collaboration between university and industry actors. Industrial partners are responsible for an increasing share of university funding, and some argue with concern, that basic science funding for universities is decreasing (see Gulbrandsen & Smeby 2005; Godin & Gingras 2000; Goldfarb 2008). It is important to note that this concern about the relationship of ‘basic’ versus ‘applied’ research on university activities has been going on for several decades (see; Pielke & Byerly 1998; Gibbons 1999; Geuna 1999; 2001; Ziman 1996; Martin 2003; Ranga, Debackere, & Von Tunzelmann 2003; Sarewitz & Pielke 2007). Furthermore, scholars continue to be divided over the true implications of this shift, given a general lack of empirical substantiation to these claims (see Van Looy et al 2004).
Recent work examining these policy changes is a continued attempt to understand two broad questions. First, what is the impact of the research funding process on university activities? Secondly, what are the effects of these policy changes on industry-university collaboration? In an effort to answer these broader questions, scholars focus on more specific questions across a wide variety of contexts. First, how have changes in funding affected the nature of research results (i.e. basic versus applied research)(Weingart 2010; Parker & Crona 2012; Smith 2010; Furman, Murray and Stern 2012; Auranen and Nieminen 2010; Hessels, Grin & Smits 2011)? Secondly, how have these changes affected the behavior of academic scientists (Lam 2010; Yang & Change 2010; Perkmann et al. 2013; Perkmann et al. 2011)? Third, what institutional factors promote or inhibit university-industry relationships (Bruneel, D’Este and Salter 2010; Shapiro 2012; Sá and Litwin 2011; Hessels, Grin and Smits 2011; Hewitt-Dundas 2012)? A selection of articles that highlight the major arguments within the literature is discussed below.
First, how has a shift in S&T policy that emphasizes applied research impacted contemporary university activity? In particular, has university knowledge production shifted away from traditional “basic” science, to more “applied” orientations? As the government adopts new processes that emphasize economic productivity, some scholars suggest that this has caused shifts in the types of scientific knowledge that are being produced (Boden & Epstein 2006, 2011; Gulbrandsen & Smeby 2005), and that these policies are potentially transforming the identity of many academic institutions (Abramo, Cicero & D’Angelo 2011; Weingart 2010; Parker & Crona 2012). These scholars are concerned that the push for applied research will not only impact funding for “basic” research, but that it will also threaten the survival of the university system as a whole, because it will shift the types of research scholars must engage in to secure funding. For example, Smith (2010) analyzed interviews with academic researchers in Scotland and England between 1997 and 2007, and found that the pressure to produce policy-relevant research is diminishing the capacity of academia to provide a space in which innovative and transformative ideas can be developed, and is instead promoting the construction of “institutionalized and vehicular (chameleon-like) ideas” (Smith, 176). In addition, some scholars contend that changes in university-industry relationships are transforming the external boundaries of the knowledge base of academic work as well as academic professional identities (Beck and Young, 2005). Since the 1980s, scholars have continued to examine this question and remain quite divided in finding evidence to support these claims.In order to move the literature forward, scholars have proposed changes in the types of variables under investigation.
First, scholars argue that research examining changes in knowledge production should focus on the different types of funding that are used and also the rate of change in funding policies. While some scholars have argued that these policy changes have negatively affected universities, others disagree. Several scholars argue that university-industry relationships can have positive impacts on university scientific productivity, especially when multiple funding sources are used (Manjarres-Henriquez, Gutierrez-Gracia, & Vega-Jurado 2008). Martin (2003) argues that the consequences of this shift for universities may not be too egregious if changes in funding policies are made incrementally. According to Martin, universities will evolve to match these changes and universities will continue to be important for national science policy, but perhaps in new ways. However, other scholars argue that federal policies are not the only explanation for changes in publication outputs. For example, Auranen and Nieminen (2010) analyze performance-based systems in eight countries, and analyze whether different funding environments impact publication performance. They found that while there are clear differences in the competitiveness of funding systems across countries, there was not strong evidence that these financial incentives were related to increases in publication productivity. Their results relate strongly to other previous studies examining funding and publication behavior (Albert 2003; Van Looy et al. 2004, and Behrens & Gray 2001). In addition, although Beaudry and Allaoui (2012) found a strong relationship between public funding and publication outputs in the case of Canadian nanobiotechnology research academies, there was no relationship between research funding and the citation count (typically used as a proxy for ‘quality’ or ‘prominence’) for academic articles. Furthermore, private contract grants (which are typically closely tied to applied research) did not have a negative affect on article publications. This provides some support that concern that use-inspired funding was diminishing basic research may be overstated. Instead, an additional factor that was connected with increased publication output was the presence of strong collaborative networks.
Research along these lines have led many scholars to examine whether changes in knowledge production are better explained by institutional factors. Hessels, Grin and Smits (2011) examined what institutional features can moderate how performance based funding influences academic practices in Dutch agricultural science. They found mixed evidence based on their use of bibliometric indicators. Factors such as collaboration with societal stakeholders and attributes of leading research questions within a field were found to be important predictors for article production. In some cases, researchers who engaged with local stakeholders were able to maintain a basic research agenda. In other contexts, when researchers worked under pressure to meet performance criteria, this tended to result in a shift of research that was more application-oriented. Similarly, in the case of South Korea, Park & Leydesdorff (2010) found that performance-based funding mechanisms that emphasized publication performance for tenure unintentionally discouraged collaborative behaviors. This suggests that an unintended consequence of these policies can result in a lack of networking between universities, industry and government; which subsequently diminishes potential technological outcomes.
Secondly, scholars argue in favor of changes in the types of dependent variables being examined in this literature. Some scholarship is emphasizing that academic publications are not the only valuable contribution made by universities. For example, Bramwell & Wolfe (2008) contend that universities are not only producing knowledge transfer, but also contribute to economic growth through generating local talent, technical support and industrial collaborations. Furthermore, Grimaldi et al. (2011) have argued that academics participate in several types of entrepreneurship that go beyond publishing research, including: social critic, patenting, licensing, academic spinoffs, collaborative research, contract research and consulting. In addition, Perkmann et al. (2013) provide a meta-analysis of articles about university-industry relationships and argue that universities engage in several additional avenues of knowledge dissemination beyond commercialization, such as academic entrepreneurship and creating intellectual property. They emphasize that university-industry partnerships extend beyond financial motives as academics and industry engage in collaborative research, contract research, consulting and informal relationships
For example, Ding and Choi (2011) examine the different ways that scientists may be involved with commercialization, focusing on two activities such as participating on a firms’ scientific advisory board or founding a company to commercialize a new discovery. In particular, they analyze when and why scientists choose to engage in one of these activities. In their analysis of 6100 scientists, they find many factors to be important. First, the timing during a scientist’s career cycle is an important predictor, with founding activity occurring much earlier than advising activities. The more productive a scientist is in their career (measured as publications), the more likely they are to found their own company, with no effect on advising. In addition, prior experience as an advisor to a firm decreases the likelihood that a scientist will engage in academic founding activities. In both activities, females lag behind male scientists but have a greater likelihood to participate in advising. These findings are similar to other research that finds that women are less likely to become a scientist entrepreneur (Allen et al. 2007; Aldridge & Audsretsch 2011). Finally, institutional factors such as coming from a top ranked university increases the propensity to engage in founding over advising by two-fold.
The foregoing discussion relates closely to a second body of questions within the S&T literature that examines how performance-funding systems have affected the behavior of scientists. Lam (2010) also argues that critics concerned that university-industry relationships are negatively impacting academic scientific practices have been too harsh. Whereas early characterizations of scientists describe them as ambivalent to external challenges, a growing literature has analyzed the active role of scientists in redefining the boundaries of their work in response to funding changes (Gieryn 1983; Lamont & Molnar 2002; Calvert 2006; Ashforth et al. 2000; Kreiner et al. 2006). In Lam’s analysis of scientists from five UK research universities, she tracks the behavior of scientists and identifies four dominant roles that characterize scientists as strategic actors. These roles vary from one extreme, a ‘traditional’ scientist, in which scientists are resistant to collaboration with industry and are motivated by the need to obtain funding, to the opposite extreme, a ‘entrepreneurial’ scientist, in which scientists believe in the importance of science-business collaboration and are motivated by the need to pursue the most important application of science and engage in networking that facilitates knowledge exchange. The implication of this study is that while some scholars have shown concern that the increasing collaboration between industry and universities is hurting academic practices, her findings show that these changes do not necessarily mean that the values associated with university research are eroding. Rather, academic scientists in her study were still able to maintain a sense of autonomy in defining their research goals. Furthermore, Yang and Change (2010) also found scientists that operated with an entrepreneurial commitment were more likely to move away from traditional methods for sharing knowledge and enhance the development of applied research.
Similarly, Furman, Murray and Stern (2012) found that in the case of the United States’ policy regarding public funding of human embryonic stem cell research, in the period between 2001 and 2003, changes in science policy resulted in scientists responding strategically to research funding restrictions. Despite limited funding targets within U.S. policies, scientists found alternative mechanisms to fund their research. Accordingly, the initiative of both U.S. university research programs and collaboration amongst scientists with global research partners drove an increase in the production of stem cell research in 2003.
Third, scholars in S&T are continuing to examine, what factors promote or inhibit university-industry collaboration? Perkmann et al. (2013) their meta-analysis highlights several individual characteristics that predict an increase in likelihood to collaborate with industry. Gender is an important predictor, as male academics were more likely to engage with industry. Age was insignificant, although seniority had a positive relationship with collaboration. Finally, more productive scientists were generally more likely to engage with industry (see Perkmann, King & Pavelin 2011). However, they note that their findings remain tenuous given the nature of their data, and that future research needs to provide more consistent measures and could benefit our understanding by providing longitudinal data.
Building on our understanding of what factors encourage collaboration, Bruneel, D’Este and Salter (2010) investigate the organizational barriers to university-industry collaboration by analyzing several UK firms that partnered with universities on publicly funded research projects (see also Leisyte 2011; Thune & Gulbrandsen 2011). Coming from the industry’s perspective, they found that industries with prior experience (reoccurrence) in collaborating with universities’ and industries with high levels of trust in university practices were better able to overcome organizational and contractual barriers to ensure productive partnerships. In addition, Sá and Litwin(2011) identified several policy tools employed by Canada to improve university-industry linkages. These policy tools include: information tools that foster communication; legal instruments that include tax incentives and credits for collaborating; and finally, financial instruments that include federal level institutions create specifically for distributing funding for collaborative efforts. These policy tools highlight the multifaceted ways that countries have restructured their innovation systems to encourage university-industry linkages.
In summary, scholars investigating the impact of funding policy changes on universities continue to be quite divided in regards to the extent of these impacts. While several scholars certainly agree that these policies have changed the nature of university activities, the findings in their analyses show that negative concerns may have been overstated. Most of this scholarship, however, has only grown incrementally over the past several decades, with the primary challenge being a methodological one (see Salter & Martin 2001). Insufficient longitudinal data and measurement issues make it extremely difficult to pinpoint the effects of these policies on knowledge production. For example, the literature commonly compares the effects of these performance-based systems across countries (for example, Auranen & Nieminen 2010; Hicks 2012). However, scholars are recognizing the difficulties with making generalizations across nations, given their different cultural, political and social contexts (see Jongbloed & Vossensteyn 2001;Perkmann et al. 2012). In addition,academic institutions utilize different metrics for evaluating performance (Abramo, D’Angelo & Solazzi 2010). Some examples include: peer review processes; teaching quality versus research output as a metric; different timescales; differences in journal quality across disciplines; etc. (Hicks 2012). Therefore, scholars are presented with major measurement challenges as they sort through and consolidate different indicators across both countries and institutions (Dilling & Lemos 2011). Finally, while shifts in research production are evident in some cases, several recent studies have identified ways in which universities and scientists adjust to these external constraints. One example of these adjustments includes responses to changes in patent laws. This is discussed more fully in the next section.
The word cloud in Figure 4 below was built from 268 keywords that coincided with the keyword “patent”. These terms came from 52 articles that were written by 111 authors. Most of these authors are located within the United States (34), Germany (12), Italy (10), Spain (9), United Kingdom (6), and Canada (6). Through a closer examination of this subset of articles, it is apparent that the most common coinciding keywords include:intellectual property rights, Bayh Dole Act, academic patenting, and innovation. Using these terms as a guide and building upon the overlap with research on university issues, I briefly summarize some of the key developments in regards to patents.
In addition to changes in funding processes for universities, some of the changes in partnerships between industry and universities have occurred in response to reforms in patent laws. Scholarship on patent production falls mostly along two lines. First, what factors promote the production of patenting? Secondly, what is the impact of changes in patent policies on university research activities? To provide the reader with context, The Bayh-Dole Act, The Stevenson-Wydler Technology Innovation Act, and Uniform Federal Patent Policy Acts of 1980 shifted intellectual property rights of publicly funded research away from the government to the institutions responsible for conducting the research (Demeritt 2000; Mowery et al. 2001). The Bayh-Dole Act gave universities full autonomy to patent and license intellectual property that was produced at their university, even if it was publicly funded. These policies were developed to encourage university scientists to engage in patenting activities, and to incentivize industrial partners to license these products. In the early 2000s, most S&T policy scholarship focused on
analyzing the quantity and quality of university patents (Mowery & Ziedonis 2000; Mowery et al. 2001; Mowery, Sampat, & Ziedonis 2002; Sampat 2006; Berman 2008). Several studies found that in the US and in Europe, the quantity of patents increased, although the quality has decreased (Mowery et al. 2001; Geuna & Nesta 2006; Geuna & Rossi 2011). In an effort to build upon our understanding of the impacts of patents, scholars are showing continued interest in several different questions. First, scholars treat patenting as the dependent variable and examine what factors influence the choice to patent (Huan, Feeney, & Welch 2011; Grimaldi et al 2011)? Furthermore, what explains the general decline in patents over the last decade (Leydesdorff & Meyer 2010)? Secondly, what is the impact of patenting on knowledge production (Evans 2010a; 2010b; Wong & Singh 2010; Thursby & Thursby 2011a, 2011b; Crespi et al 2011; Geuna & Rossi 2011; Parthasarathy 2011)?
First, recent research has continued to build upon the previous literature by investigating the factors that determine the propensity for a scientist to engage in patenting activities. First, several studies are pointing to institutional factors that encourage patenting. For example, Huan, Feeney, and Welch (2011) utilized a national survey of university scientists and engineers and found that both individual and organizational factors were correlated with individual patent production. Primarily, they found that universities that provided departments financial incentives were associated with higher quantities of patenting. Evolving institutional performance mechanisms can also be attributed to a possible decline in patenting. Leydesdorff and Meyer (2010) address concerns that since the early 2000s, university patenting has been on a general decline.They argue that data quality issues, including the new university ranking system, which do not include patents or spin-offs, have made it difficult to accurately draw conclusions about the cause of recent patent declines. However, Geuni & Rossi (2011) attribute a general decline in patenting in Europe to a shift from academic patent ownership to university patent ownership. Under academic patent ownership, an individual researcher would have the intellectual property rights to their research. Under, university patent ownership, the institution that supports the research maintains rights to the product. This research suggests that scientists have adjusted to these changes by choosing alternative commercialization practices.
Along similar lines, Link, Siegel and Fleet (2011) found that the Bayh-Dole and the Stevenson-Wydler Act did not have an impact on patenting at National Laboratories in the way that these policies have on universities. Rather, policies that focused on financial incentives, such as the Federal Technology Transfer Act, had a greater influence in prompting patenting.Furthermore, their results showed that senior scientists who demonstrated strong publication records and had open-science attitudes (measured as the extent to which the scientist favored open-science norms and values) were more likely to have multiple patents.
In addition, Wong and Singh (2010) found additional evidence that strong publication records are an important predictor of patent quantities. Their study showed a positive relationship between the patenting output of major universities around the world and the quantity and quality of their scientific productions. However, while production is high within North America, the degree of internationalization of faculty members is found to reduce patenting performance within North American universities. Furthermore, patents had less of an impact on the quality of productions for universities outside North America. In similar research, Haeussler and Colyvas (2011) find several individual characteristics of university scientists that predicting greater participation in academic entrepreneurship. Scientists with attributes such as professional security and productivity were more likely to value patenting and other commercial achievements.
Secondly, scholars are continuing to draw connections between the existence of patents and its impact on knowledge production. Thursby and Thursby (2011a) examined eight U.S. universities’ patenting activities between 1983-1999, to test a reoccurring hypothesis that patent practices under the Bayh-Dole Act have diverted faculty away from participating in traditional basic research (see also Sampat 2006). Their study found no evidence to support this view. In contrast, they found that both basic and applied research has increased since the passing of the Bayh-Dole Act.Although around 2002, basic research leveled off, with applied research continuing to increase.
Andersen and Rossi (2011) also found that patents had a positive impact on industrial knowledge production. In a study of 46 universities in the UK, patents were seen to result in the most effective knowledge dissemination in the economy. Furthermore, proprietary forms of IP led to increases in financial resources and knowledge transfer flows from university to industry, but non-proprietary forms of IP were primarily beneficial to the universities’ own innovation processes. Nonetheless, shifts in patent laws led several universities to engage in activities with industry, such as commercial ventures that might lead to profit.
Along similar lines, concerns about the impact of the Bayh-Dole Act have led several scholars to investigate the impact of patent reforms on knowledge sharing.Evans (2010a) examined the impact of university-industry relationships on knowledge dissemination and sharing by comparing interviews with academic scientists and industrial researchers. While members of academia are acculturated toward communicating and sharing scientific ideas, professional and career goals may incentivize scientists to withhold scientific information, at least until they can secure credit for their results. Contrastingly, Evans found that companies tend to manage their resources for long-term control. The results of his study found that industry sponsorship was associated with a reduced effort by scientists to share research results and methodology. However, this may be in part due to differences in the types of research these scientists undergo.Horta and Lacy (2011) found that information sharing is highest at large research institutions than small research institutions. In addition, academics at large research institutions were more likely to publish internationally. In a separate study, Evans (2010b) found that government funding tended towards confirmatory work with established methodology, but industrial funding was more likely to explore new scientific ideas with speculative techniques. In summary, perhaps due to commercialization incentives, industrial researchers are engaging in more innovative research, but they are also more cautious about sharing their research results.
In summary, scholars are finding continued support that the Bayh-Dole Act of 1980 successfully prompted universities to increase their patenting efforts. However, it was the National Technology Transfer Act that had a greater influence on national laboratory patenting. Secondly, whereas some scholars have argued that these reforms may be diminishing the production of basic science research, evidence by Thursby and Thursby (2011a) does not support this view. Instead, increases in patenting seem to encourage both basic and applied research to rise. Along similar lines to the university research discussed above, research on patent reform is limited primarily due to methodological concerns. Several studies noted poor data quality and measurement issues (Geuna & Nesta 2006; Crespi et al. 2011; Leydesdorff & Meyer 2010). For example, Aldridge and Audretsch (2011) discuss the limitations of data coming from technology transfer offices because they do not include robust measures of scientist entrepreneurship. Furthermore, similar to the university issues, scholarship on patents also highlights the limitations with comparative studies that examine changes across nations (Geuna & Rossi 2011). Most of the research discussed above about performance-funding mechanisms and about patent reform emphasized how these policies resulted in increased collaboration between university, industry and government organizations. These collaborations are the structural elements of several different innovation models that explain national technological outcomes. The next section describes in greater detail, research on innovation processes.
Figure 5 below is a word cloud that was built from 797 keywords that coincided with the keyword “innovation”. Innovation was the most frequent keyword to appear in recently published S&T articles (210). These terms came from 155 articles that were written by 348 authors. Most of these authors are located within the United Kingdom (64), United States (56), The Netherlands (31), Spain (19), Sweden (17), China (16), Denmark (12), Italy (12), and Germany (11). The word cloud reveals that the most common keywords to coincide with “innovation” include: technological innovations, innovation policy, national system of innovations or national innovation systems, patents and policy sciences. Several terms that appear less commonly but reflect these innovation policy systems included:, national innovation systems, triple helix model and mode 1 and mode 2 models of innovation. Using these terms as a guide, I briefly summarize some of the key developments in regards to some of these primary innovation models.
Starting around the 1980s, scholars began to challenge the foundation and utility of the postwar paradigm for science policy. First, scholars argued that the linear model was empirically and descriptively incorrect (Kline & Rosenberg 1986; Malecki 1997; Stokes 1997; Evangelista 2000; Tether 2005; Bodin 2006; see Blaconi, Brusoni, & Orsenigo 2010 for counter-argument). Secondly, concerns about the fallibility of scientific knowledge raised scholarly concerns, primarily on normative grounds, as to whether an expert-driven process was effective for guiding future directions in science policy. These concerns have led to many criticisms of postwar science policy as a useful guide for structuring R&D initiatives, and consequently resulted in the proposal of more dynamic models. While scholars agree that knowledge production is important for technology development, recent scholarship continues to examine in greater detail when, how and why knowledge translates into the successful production of technology.
Therefore scholars are concerned with understanding the conditions under which scientific knowledge translates into the successful development of technology, and ultimately results in economic growth. Within the field of S&T policy, the three most common models that developed subsequent to the linear model include: national innovation systems (Lundvall 1992; Nelson 1993); new conceptualizations of knowledge by mode 1 and mode 2 (Gibbons et al. 1994); and the triple helix model (Etzkowitz & Leydesdorff 1997). The following models argue, although to different degrees, that the optimal utilization of knowledge is found in the synergy between government, industry, universities, and society.
Starting in the late 1980s, Christopher Freeman (1987), Bengt-Ake Lundvall (1992) and Richard Nelson (1993) developed National Innovation Systems (NIS) model for understanding innovation processes. These authors were primarily interested in the relationship between economic growth and technological production. Scholars at this time argued that international competitiveness was linked with knowledge production. Earlier work in NIS focused on the role of firms in long-term economic change, but in the last several decades, NIS scholars have argued that economic benefits from knowledge production will vary depending on social, political, organizational and institutional factors at the national level (Edquist 2006; Gray 2011). While the growth of research in NIS has remained steady (Uriona-Maldonado, Santos, & Varvakis 2012), it is important to note that scholarship in NIS set the path for the emergence of more recent approaches to understanding innovation systems. These arguments were critical for understanding the transformation in science policy from focusing on the production of basic information to focusing on usable or applied information (Fagerberg & Sapprasert 2011). More importantly, NIS research laid the foundation for connecting economic performance with the technological capabilities of the institutions within a nation (Mokyr 2002; North 2005; Castellacci 2008; Fagerberg & Srhole 2008; Filippetti & Archibugi 2010; Zhao & Guan 2012).One important implication of NIS studies and previous research on innovation is that technological advancement is not directly related to investments in R&D, but rather learning and innovation occur between different users and producers of knowledge and technology (Brooks 1968, 1990; Freeman 1987; Nelson 1993; Fagerbert & Sapprasert 2011; Hung & Whittington 2011). Recent work in a scientometrics study by Zhao and Guan (2012) are continuing to reach these same conclusions, finding that that there was no significant connection between R&D expenditures in nanotechnology and the actual practices of research within universities. Rather, their study emphasized the importance of collaboration between university and industry actors in order to promote technological developments. The main theoretical premise behind NIS led to the proliferation of other science policy models that examine the processes that connect knowledge outputs with technological developments. There is a rapidly growing literature on two of these models: Mode 1 and Mode 2 forms of knowledge and the Triple Helix model.
This emphasis on the interactions between key actors such as government, universities and industry is central to another scholarly effort to model technological outcomes called mode 1 and mode 2 conceptualizations of knowledge. In 1994, in The New Production of Knowledge: The Dynamics of Science and Research in Contemporary Societies, Gibbons et al. (1994) contended that the research process had changed dramatically since the postwar period. First, these scholars argued that the relationship between S&T was much more interactive than the linear model proposed, and secondly, the process involved a broader set of participants in the decision-making process. Gibbons et al. call the old paradigm ‘Mode 1’ and argued that this paradigm had evolved into what they call ‘Mode 2,’ or a research process that had an applied focus, was dominated by a multi-disciplinary approach, and involved a broader set of participants in the decision-making process. The evidence they cite for this transformation included: ‘users’ or ‘consumer’ participants on peer review boards, detailed impact studies that acknowledge the applications that can be derived from funded research projects, and scientific knowledge being no longer considered a public good, but rather was considered intellectual property that is traded in a similar fashion to other goods and services. The transformation of knowledge into intellectual property was a result of globalization and the development of new research communities including knowledge organizations, think tanks, management consultants and activist groups (Nowotny, Scott & Gibbons 2003).
Mode 2 processes promote a more interactive and multi-disciplinary approach to problem identification and research practices than Mode 1. Recent scholarship in S&T investigates the extent to which research programs designed upon the premises of the Mode 2 approach to research management actually promote interaction amongst government, university, and industry. In the United Kingdom, for example, Mitev and Venters (2009) assessed a three-year research project on environmental sustainability in the construction industry, funded by the Engineering and Physical Sciences Research Council. Although a Mode 2 research approach promotes a trans-disciplinary approach to knowledge production, Mitev and Venters found a lack of consensus and unresolved tension amongst research collaborators, resulting in different agendas from academic and industrial partners. They also identified impacts from institutional incentives that are tied with university performance criteria. They found that academic researchers behaved differently in the research funding application phase and post-funding phases, resulting in poor multi-disciplinary collaboration (see also Swan et al. 2010; Worrall 2008; and Scarbrough 2007). In addition, because the issue of environmental sustainability tends to be politically charged, Mitev and Venters conclude that tensions between public and private agencies were likely due to different preferences over research practices and different sources of accountability, resulting in a lack of agreement over proposed solutions. In summary, their study challenges the Mode 2 research management approach for neglecting institutional and political contexts (see also Hansen 2009). Therefore, these scholars argued that the boundaries between Mode 1 and Mode 2 research are not clear, and therefore they suggest improving the management and implementation of the Mode 2 approach.
A mode-2 society is conceptualized such that the production of scientific research is to be socially embedded and emphasizes open discussion and interaction between scientific experts and the public (Nowotny, Scott & Gibbons 2003; Lengwiler 2008). However, in a study of public engagement projects in fourteen approaches to nanoscience and technology funding, Kurath (2009) found little evidence that the programs increased social robustness. For example, a few of the approaches scored negatively on the criteria of acceptability. The tension in her findings may also be due to unclear boundaries between mode 1 & mode 2 processes, or they may be a reflection that the conceptualization of mode 2 overstates its promise to produce socially robust knowledge. While more empirical work is needed to understand the implications of the Mode 2 paradigm, other scholars within S&T have focused on the relationship between government, university and industry through a triple-helix model.
The triple-helix model is another recent way that scholars describe innovation, or the translation of knowledge production into technological outcomes. This model also describes technological outcomes as a result of the interaction between universities, industry and government, but also emphasizes the transformations of these actors as result of these interactions (Etzkowitz & Leydesdorff 1997, 2000; Webster & Packer 1997; Etzkowitz 2003; Metcalfe 2010; Shin et al. 2012; Park, Hong & Leydesdorff 2005). Proponents of this model argue that this tri-fold relationship is the key to maximizing the conditions for innovation. The government utilizes contractual agreements that stabilize interactions, the university provides new knowledge, and finally industry provides the production of new technologies. Organizational arrangements within the government have changed such that it has resulted in many new developments, including hybrid academic research centers that collaborate with both industrial and governmental partners. Whereas former views of the role of universities was that universities limit itself to research and education, Henry Etzkowtiz (2003) argues that currently, entrepreneurial universities play a critical role in the production of innovation as these universities seek to influence the economy. Entrepreneurial universities arose as university policies began responding to competitive research funding systems. Similar shifts are occurring within industry and the government. The federal government is playing more of a steering role in the research-funding process and within industry we are seeing the rise of spin-off organizations that have academic characteristics. The triple-helix model argues that as universities, industry and government increase their interactions, new organizational forms will arise to promote innovation and enhance each actors’ general performance.
But are government, universities and industry the only important actors? Amy Metcalfe (2010) examines intermediating organizations that operate in the spaces between public, private and academic organizations. These organizations are typically nonprofit organizations that include: professional associations, foundations, consortia, independent research support organizations and special interest groups. Using case analysis of two intermediating organizations in Canada, Metcalfe found these organizations to have a major influence on the exchange of actors, resources and flow of commerce. Acting as liaisons, these organizations were important for the creation of strategic alliances and collaboration, as well as drawing attention to new products and services.
Initially, the Triple Helix Model underemphasized the role of the public in optimizing S&T programs. However, drawing upon the literature in Mode 1 and Mode 2 conceptualizations, some scholars have synthesized these frameworks into a new mode for future science policy. Carayannis and Campbell (2012)argued that innovation modeling should be expanded to a Mode 3 or a Quadruple Helix model that includes interaction between government, universities, industry, and civil society. Whereas Mode 1 was a focus on basic university research, and Mode 2 was centered on knowledge application and knowledge-based problem-solving, a Mode 3 approach focuses on a synthesis of top down government, university and industry policies with a bottom-up civil society initiatives. They argue in this framework that by utilizing civil society organizations as well as media-based and culture-based public organizations, science policies can gain valuable insight into socio-economic and socio-political factors that surround issues in S&T. By integrating all four actors (government, industry, academia, and civil society), the interaction between these groups will promote co-development and co-specialization that will facilitate the emergence of different knowledge modes. Taking their model further, Carayannis and Campbell also conceptualize the possibility of the Quintuple Helix that would incorporate the natural environment or societal context. Arguing for a systems theoretical approach to understanding these dynamics, they argue that by encompassing these additional factors, S&T policy can move toward a knowledge-based and innovation-based democracy that integrates the many factors that have been woven throughout S&T policy research.
Although the literature above has been the predominant way for scholars to model technological progress, these efforts have been criticized for being mostly a theoretical and needing greater empirical verification (Kleinman 2010). How accurately do these theoretical models describe the real-world policy making environment? Logar (2011) examined several of these models as well as Post-normal Science (Funtowicz and Ravetz 1993), Pasteur’s Quadrant (Stokes 1997), and Well-ordered Science (Kitcher 2001) in light of the policies adopted within the US Department of Agriculture, the Naval Research Laboratory, and the National Institute of Standards and Technology. In these institutions, he found that two primary factors discussed in these policy models were accounted for: inclusion of applied-oriented research and input into the decision-making processes. From a policy application perspective, Logar argued that attributes of the more complex models, such as Mode 2 science, are too vague and difficult to evaluate. This suggests that by improving the empirical examination of innovation systems, policy scholars may improve their ability to build models that are not only more descriptively accurate, but that are also able to provide policy makers with guidance.
As presented in this review, the subfield of S&T policy is ripe for further exploration and new avenues of research. Most of the research over the past few years is continuing to examine questions that have prevailed over the past several decades. It seems that due to data limitations and measurement issues, this research has grown only incrementally since the 1990s. On the basis of the evidence currently at hand, it is becoming widely recognized that financial investment in scientific knowledge alone will not necessarily generate economic growth. Furthermore, there is disagreement as to whether the Bayh-Dole Act is responsible for the rise in patenting after the 1980s (Berman 2008). Thus, scholars are interested in finding more rigorous ways to explain when, how and why these processes are effective. In both cases, scholars are pointing toward institutional factors. First, national S&T policy has direct implications for academics working in research universities as recent scholarship continues to examine how policy changes cause shifts in university outcomes. Although policy changes are affecting the nature of university activities, specifically in terms of the type of research scholars are engaged in and patenting activities, preliminary findings in scholarship are indicating the ways in which universities and scientists are adjusting to these constraints. Secondly, the complex aspects of many S&T issues has created a policy context that has called for restructuring the processes for managing science, technology and innovation programs. Scholars have gained some insight into these processes by understanding how collaborations between the government, universities, and industry impact technological outcomes. However, research in this area has been primarily focused on theorizing and needs more rigorous empirical substantiation.
In conclusion, one general development that is reflective in this review article is a shift in focus in S&T policy scholarship, away from the production (both in terms of quantity and quality) of new technologies to a focus on the relevance and acceptability of technological outputs. However, the main challenges that emerge from such an assessment of the field are a lack of clarity regarding important concepts and the difficulties associated with conceptualizing and operationalizing variables within these studies. Looking across the entire spectrum of scholarship in S&T policy, it seems that scholarly efforts have mostly been centered on theorizing. While normative and formal theorizing is an important precursor to the expansion of theoretical ideas, theory is advanced through empirical analyses. S&T policy scholarship is moving in this direction, as some scholars have recently made calls for improved methodological methods (see Lane and Black 2012; Fealing et al. 2011). By focusing on methodological advancements, this will provide stronger methods for hypothesis testing and theory development. In addition, better metrics will improve our ability to reliably measure the success or failure of scientific investments. As policymakers make decisions about future investments in R&D, further academic contributions in understanding the impact of S&T policy are vital for informing decision-making at the federal level. Although recent scholarship provides preliminary insight into these questions, there is still need for more rigorous research in these areas.
1 The Impact Factor is a measure reflecting the average number of citations to recent articles published in the journal, and is typically indexed for the two previous years.
2 This does not reference unique individuals. If a person authored more than one article, then they may have been included multiple times. Because these indicators are about influences on the nature of published research, it seemed relevant to include each author as many times as they appeared.
3 An individual “keyword” may be a single word or a phrase made up of multiple words.
4 By using the keywords provided by authors, it provides us with an interesting view into what they deem as the most important concepts within their article. Any given article can be about a range of important topics: methodological, substantive topical, theoretical, or all of the above. I use the provided keywords as a starting point to organize the structure of this analysis, and then provide a more in-depth narrative analysis that summarizes and discusses the most common topics. This provides us with an objective selection of a large number of articles covering a wide-variety of topics.
5 These countries include: Argentina, Australia, Austria, Belgium, Brazil, Brussels, Bulgaria, Canada, Chile, China, Colombia, Croatia, Cyprus, Czech Republic, Denmark, Ethiopia, Finland, France, Germany, Ghana, Greece, Hungary, Iceland, India, Iran, Ireland, Italy, Japan, Kenya, South Korea, Latvia, Macau, Malaysia, Malta, Mexico, Netherlands, Norway, Pakistan, Poland, Portugal, Russia, Saudi Arabia, Serbia, Singapore, Slovenia, South Africa, Spain, Sweden, Switzerland, Taiwan, Thailand, Netherlands, Turkey, United Arab Emirates, United Kingdom, United States of America, Venezuela and Vietnam.
6 Tagxedo is a software program that takes text and utilizes the frequencies of words to create a word cloud. It is available online at: www.tagxedo.com.
7 The defining characteristic of ‘basic’ science is experimental or theoretical work toward the purpose of a broad understanding of particular phenomena within a field
8 During that time, proponents of the scientific enterprise placed great value on the production of “basic” science, and national science policy placed a clear boundary between funded research and commercial applications. Vannevar Bush and his supporters were concerned that constraining the pursuit of general knowledge to ties with practical use could potentially stunt the creativity of scientific research and the country’s competitive edge. The idea of the “linear model” grew out of this insight and becharacteristic of the National Science FouSF) concept of “technology transfer” (Stok1997).
9 In the postwar period, the traditional linear model, which describes a process for connecting scientific knowledge, technological development and economic growth, has been the predominant paradigm for understanding how governmental support for science research should be structured. The linear model describes a process that clearly delineates amongst the following four stages: scientific information begins as basic research (acquiring information about a subject without specific applications), is then incorporated into applied research (acquiring information for specific objectives), and finally, that information is utilized in the development, as well as the production of new technology. This model has been the adopted approach for the past several decades and provided the structural basis for most research and development programs that were financed within the Department of Defense (Stokes 1997). The historical underpinnings of the linear model, not only advocated for an ideal of “pure” science to precede “applied” research, but also justified the necessity of major investments in basic science research to be accomplished primarily through universities. However, several scholars have criticized the use of the linear approach and have highlighted its insufficiencies for guiding future investments in science policy. More recently, scholars are analyzing alternative models that shift the making of national science policy away from the linear model to more interactive models, particularly give the evolving relationship between public funding, universities and industry.
Sarah Trousset is a doctoral candidate in Political Science at the University of Oklahoma. She is also a research assistant at the Center for Energy, Security and Society and Center for Risk and Crisis Management.