ABSTRACT. Academic promotion is both a selection mechanism, where highly-skilled individuals are selected to carry out certain tasks, and an incentive mechanism that is intended to provide motivation to researchers in order to increase performance and achieve excellence. In this regard, systems of promotion are expected to be designed in a way that contributes to increase the quantity, quality and impact of academic and scientific outputs. Among European countries there is a large variation regarding the academic hiring and promotion systems. Variation goes from strong centralization with national exams to decentralization to the single universities; somehow in between some countries have established accreditation systems within a two-step model. In Spain, the first university reforms, in 1983, combined access to the professional group of university professors (national civil servant status) with a strong decentralization that provided the departments and universities with strong power to influence the selection process. In the following years complains about the inbreeding and cronyism in Spanish universities started to emerge and created a favourable environment to the 2001 University reform: A national habilitation system, based on formal public exams and tournaments, was established and this highly centralized system was in place until end of 2007, when a new university reform replaced it with an accreditation model. The accreditation system, nationally organised and managed by a single agency (ANECA) was a first step, while the formal hiring or promotion was eventually determined by each individual university (and department) among the accredited professors.. The three models that have organized academic promotion in Spain over the years have raised political battles and policy debates regarding their costs and benefits, and relevant research questions to understand whether and to what extent different promotion systems lead to distinct academic pathways associated to different organizational performance measures. Our research is based on data coming from a survey carried out among Spanish science and engineering academics in a representative sample of universities across the national territory. We will empirically address some general questions such as: under what conditions centralized, decentralized or accreditation academic promotion systems could guarantee the best potential results in terms of the merits of the candidates; what is the relationship of the different promotion and selection systems with the type of academics who are selected, for example in terms of mobility trajectories? At the university strategy university level, the comparison among the consequences of the different systems could help us to better understand the connections of the implementation of the promotion systems at local level (under a model of governance of the universities connected to democratic elections) and the performance of the universities. At the micro level, and controlling for some relevant variables such as gender, age or field, this research will provide novel evidence on the differences of academic careers between individuals that were promoted through the three different systems. We focus on specific issues such as: national and international mobility, participation and coordination of research projects, supervision of PhD students, participation in committees, etc. Besides, we also analyse academics’ perceptions on the key factors that departments of Spanish universities take into account when promoting candidates (e.g. productivity vs. institutional loyalty, internal status vs. external status). Perceptions are relevant because they might be drivers of individuals’ strategies, for instance, those oriented to achieve further promotion.

ABSTRACT. Psychology research identifies social belonging as a basic human need and driving motivation (Walton & Cohen, 2007; Baumeister & Leary, 1995). By contrast, job satisfaction is typically characterized as deriving from factors such as financial compensation, opportunities for advancement, autonomy and independence, intellectual challenge, flexible hours, or job security. These measures of job satisfaction provide little insight into possible interpersonal determinants of job satisfaction, including whether an individual feels they "belong" with their co-workers.

In this paper, I combine data from multiple recent nationally-representative surveys to examine how science-related attitudes and perceptions, personal values, and the relative importance of various job attributes differ among U.S. resident men and women working in STEM occupations, versus those who considered careers in STEM but are working in other occupations, and versus those workers who say they never considered a STEM career. Using multivariate binomial and ordered logistic regression models, I investigate three related hypotheses. First, I hypothesize that individuals who share values and attitudes prevalent among STEM workers are more likely to have considered a career in STEM themselves. Second, I hypothesize that individuals in STEM (or non-STEM) occupations whose job-related priorities are consistent with prevailing priorities and attitudes among other STEM (or non-STEM) employees in their same sector (e.g., academia, industry, government, etc.) will will express higher levels of job satisfaction than those with more dissimilar attitudes. Finally, I hypothesize that gendered differences in values, attitudes, and job-related priorities may help to explain women's lower rate of participation and higher rate of departure from STEM careers.

Data for this paper include the 2012 General Social Survey (GSS), the 2010 Survey of Doctorate Recipients (SDR), and the 2010 National Survey of College Graduates (NSCG). The 2012 GSS includes a module capturing whether the respondent considered a career in science or in engineering (e.g., "Have you ever considered working in a science-related career?"), as well as their level of agreement or disagreement with several statements invoking positive or negative perceptions of scientists and engineers (e.g., "Scientists don’t get as much fun out of life as other people do"). Another GSS module then asks, for a series of personal value statements, the extent to which the individual identifies with each one. Combining these individual-level responses with other demographic and occupational characteristics, we do find significant and gendered differences in personal values held by individuals who considered and pursued STEM careers. For example, about half of those working in STEM occupations said it was important to them to show their abilities and be admired for them. Scientists and engineers who did not share this value were significantly less likely to say they were "very satisfied" with their jobs, and notably, only 1 in 3 women in STEM occupations identified with this value.

The 2010 SDR & NSCG survey instruments each included a series of questions asking the relative importance of various job attributes, and the extent to which the respondent was satisfied with those aspects of their current job. The 2010 NSCG sample includes both respondents from the 2008 NSCG, as well as a sample drawn from respondents with bachelor's and higher degrees in the 2009 American Community Survey, to provide a nationally-representative sample of the U.S. college-educated workforce. Preliminary analyses with the SDR data reveal that male and female STEM PhDs are equally likely to change jobs for an increase in pay or opportunities for advancement. But, female STEM PhDs who change jobs are more likely than their male counterparts to say they are doing so to improve their working conditions, or due to a change in career or professional interests. Additional analyses will examine correspondence between satisfaction with various job attributes by occupational category and employment sector, and the relative value or importance individuals place on each of these.

ABSTRACT. These researches about science and technology policy based on curriculum vitae (CV) of S&T researchers are originally a program named Research Value Mapping house at Georgia Tech. There is a major object to find the new method or technology for S &T policy research in that program. And the CV analysis method has been preset in this satiation and started to be systematically applied in the research now. For the academic researcher, the CV represents, all at the same time, a record of scientific accomplishment, a brief history of the professional life course including much information, for example, education, academic, career, publications, funds, awards and so on, an obligation to administrative superiors, and a job search resource. The past studies based on CV had included the international migrate of researchers and the evaluation of performance (KOEN JONKERS, ROBERT TIJSSEN, 2008; Benedetto, Lepori, 2009; Ulf Sandstrom, 2009), the career trajectory of scientist and engineer and so on (Dietz et al., 2000；Elizabeth Corley et al., 2003) In this study, we selected two fields to conduct case study that took Chinese researchers as sample. One is Immunology and the other is Management Science and Engineering. We collected the CV data of samples through the internet and set up the experiment group and control group by the different aim of study. Then we evaluate and analyses the career development and performance of researchers in the different field based on the Survival Analyses. In the first step, we focused on the influence of Postdoctoral positions to the career development of researchers. The past studies based on the CV analysis found that the postdoctoral positions has influence to the career development of researchers but is difference in the different countries: postdoctoral positions in France delay or deter academic careers, but have no impact on entry in the US (Monica Ganghan, Stephane Robin, 2004). Other study found that the postdoctoral experiencing has a negative effect on being assistant professor and professor combined with the method of survival hazard analysis (Tian Ruiqiang, et al., 2013). However, there is a problem in these studies that have not made a distinction between different research field and we think there have different impact of postdoctoral positions in the different research field. In the second step, we focused on the key influence factors on the performance of researchers, such as graduate school, institution prestige, academic experience or frequency of flow. The results of Survival Analyses are as follows: Kaplan-Meier analyses indicated that researchers who have the postdoctoral experience have higher risk on getting the position of associate professor in the Immunology discipline, indicating a shorter time from assistant professor to associate professor. However, opposite results are discovered in Management Science and Engineering discipline. In additional, COX proportional hazard analyses indicated that these factors such as gender, institution, and country of doctor degree are significance to the career development in the Immunology discipline. In the Management Science and Engineer discipline, other factors such as the reputation of institution and mobility are significance to the career development. Lastly, we analyzed and discussed probable reasons for such phenomena.

ABSTRACT. This paper addresses the differential effect of researcher evaluation systems on individuals’ career trajectories. The National Research Council of Argentina (“Consejo Nacional de Investigaciones Científicas y Técnicas” or CONICET) is the main research funding agency of the country and also the entity that employs most of research staff in public research institutions and labs. Researchers are said to be members of the scientific and technological research career within the agency when their research employment is affiliated with it. This paper analyzes the features in the research background and career trajectory of researchers in the agency that have an effect on the chances of entering and being promoted up its career hierarchy. The analysis focuses on 599 researchers who in 2012 were members of one of the initial levels of the agency’s research career (“assistant researcher” or “adjunct researcher”) in two evaluation committees, namely, the disciplinary committee on biochemistry and molecular biology and the committee on engineering and biotechnology processes. These two evaluation committees have different members and are assumed to use different evaluation criteria for promotion. This would lead to the inference that the career trajectories of researchers evaluated in different committees would have different features and that their scientific and technological reputation would be based on relatively different research achievements. The evaluation mechanisms in these committees are highly institutionalized and respond to the priorities of two different communities that provide “peers” for the evaluation system. We seek to establish whether these inferences are justified and differences are detectable in the career patterns of researchers promoted through the two different committees. The data are derived from the CVs of researchers entered into the CVar system of reporting of the Argentine Ministry of Science, Technology and Productive Innovation. Some missing items in the reporting system were gathered from researchers’ ordinary CVs available in CONICET records. Publication and citation data were drawn from SCOPUS and the SCImago Journal Rank indicators. From these data, a logit regression model and a classification tree were computed to estimate the probability of promotion in 2013 or 2014 in the CONICET career system with a set of relevant predictors to characterize researchers’ background and trajectories. Among these predictors, we include the participation in postdoctoral research abroad, position in a typology of research focus (created for this purpose and including categories such as “classical academic”, “academic with R&D service orientation”, “academic with R&D management profile”, “academic with a S&T dissemination focus” and “academic with multiple foci”), timing of promotion to current level (i.e. early promotion compared to researcher’s cohort), length of period in current level (from the minimum period allowed and longer), various measures of scientific and technological productivity in the last 5 years, mentoring of PhD students, prizes and awards, research grant awards outside CONICET and what evaluation committee the researcher belongs to, among others. We find that the profiles of researchers leading to increased probability of promotion differ between the evaluation committees but some results are surprising. For example, the greater technological focus of the industry and biotechnology committee is not reflected in what is actually rewarded with promotion and R&D management is rewarded more highly than scientific productivity in the later stages of a researcher’s career. These and other results are discussed in the paper showing that the institutionalization of evaluation has an effect on the career patterns of researchers.

ABSTRACT. I seek to answer the question of how compulsory licensing policies affect the inventiveness of both inventors who are compelled to license their patents and inventors who receive compulsory licenses to such patents. Compulsory licensing policies require patent-owners to allow others to make use of those patents in exchange for a compensation level determined by a third party, usually a government authority. Compulsory licensing stands in contrast to traditional patent licensing, in which the patent-owner can demand any level of compensation. Government authorities have invoked compulsory licensing policies in cases where widespread social benefits are obstructed by patent-owners. For example, some nations have used compulsory licensing policies to allow the manufacturing and distribution of medicines whose patent-owners demand prohibitively high levels of compensation.

Scholarly disposition on the impacts of compulsory licensing has been mixed. Some scholars have argued that patent-holders will likely receive a lesser reward for the expense endured during research and development if their licensing compensation is dictated by governing bodies. As a result, researchers may be less likely to pursue patents as a means of exploiting their inventions. Less patenting could slow innovation overall by restricting the public disclosure of discoveries provided by the patent system. Conversely, other scholars argue that reduced compensation for licensing increases the likelihood that a patent will actually be licensed and worked into a commercial product that could deliver social benefits. Patent-owners could still earn high returns on low-compensation licenses by investing in patent licensees, and accelerated commercialization of patents could even stimulate innovation by creating greater technological competition. As such, compulsory licensing could maintain or even accelerate the pace of innovation while accelerating societal returns from research and development efforts.

The US Department of Energy (USDOE)’s solid-state lighting (SSL) research program serves as our case of interest. SSL technologies produce light via electrification of novel materials structures and are valued for their high energy efficiency; the most common example of a SSL technology is the Light-Emitting Diode (LED). The US DOE expanded its SSL research portfolio in 2007 with the creation of a separate program under the Building Technologies Office; As part of the creation of this new separate program, the US DOE acquired an exemption to the Bayh-Dole act and instituted a compulsory licensing policy. Under this policy, any holders of patents based upon research funded by the new SSL program were required to engage in non-exclusive licensing negotiations arbitrated by the US DOE. The prospective licensees in these negotiations were a number of lead firms in the lighting industry who had been selected into a group called the Next-Generation Lighting Alliance (NGLIA) and who served as industry advisors to the US DOE’s program. I seek to reveal the effects of the US DOE’s compulsory licensing provision on both the researchers funded by the SSL program and the members of NGLIA. This case provides novelty to the compulsory licensing literature, since most cases of compulsory licensing studied revolve around international trade governed by the TRIPS provision within the WTO agreements.

To reveal the impact of compulsory licensing upon the researchers funded by the US DOE SSL program and the NGLIA firms, I analyze multiple data sources with a difference-in-difference methodology coupled with statistical matching techniques. Patenting inventors are the unit of analysis, and these can be either individuals or firms. I hypothesize that the compulsory licensing program caused affected inventors to produce more patents than comparable, unaffected inventors. The researchers funded by the US DOE’s SSL program and the NGLIA firms are both identified via public documents on the US DOE’s website. I use these documents to attribute the key independent variable, whether a researcher or firm was affected by the US DOE’s compulsory licensing policy, to each unit of analysis. I also gather key independent variables from EuroMonitor and Bloomberg business intelligence databases, as well as analyses of public web records on individual inventors. I use patents as the dependent variable, and I use the Derwent Innovations Index patent database to gather all patents in the solid-state-lighting technology area. The “inventor” and “assignee” fields allow attribution of patents to inventors in my dataset. I then use propensity-score matching to identify comparable inventors in the dataset unaffected by the compulsory licensing policy, and compare the patenting behavior between the affected inventors and the unaffected inventors via difference-in-differences methodology. I check for robustness using alternate statistical matching techniques, such as kernel regression.

ABSTRACT. This article aims to assess whether patents issued from a patent-paper pair or owned by a university are of a higher quality than other patents to which academic inventors have contributed and whether patents owned by universities. This research implicitly analyzes the impact of university-industry linkages on the quality of biotechnology and nanotechnology patents in Quebec. This article measures “quality” first as the number of forward citations, second as the number of claims and third as an originality index built from the principal component analysis of the number of forward citations, the number of claims and one minus the Herfindahl index of the number of backward citations by IPC class.

Finding potential patent-paper pairs requires integrating data from two sources: the United States Patent and Trademark Office (USPTO) for patents and Elsevier’s Scopus for papers. We first identify the potential patent-paper pairs by selecting the patents developed by authors-inventors. With the simple constraint that at least one of the inventors published one or many articles in the period [t-2, t+2], where t corresponds to the patent application date, we find 22,688 potential biotechnology patent-paper pairs and 20,003 potential nanotechnology patent-paper pairs in Canada. We then used the text mining software Rapidminer to calculate the cosine similarity between the paired patent and paper documents. The similarity measures range, in our sample, from 0 to 0.78 for biotechnology, and from 0 to 0.53 for nanotechnology (theoretically similarity measures range from 0 to 1). Similarly to Lissoni and Montobbio (2006), we selected the patent-paper pairs from the top 10 percentile of the similarity measures. This roughly corresponds to a similarity measure of 0.30 for both biotechnology and nanotechnology samples. Using this 0.30 threshold for the similarity measure, in Canada, we find 249 actual nanotechnology patent-paper pairs and 376 biotechnology patent-paper pairs.

Funding may play an important role in the quality of patents to which academics avec contributed. The third source of information necessary for this study is the Quebec University Research Information System (Système d’Information sur la Recherche Universitaire – SIRU) provided by the Quebec Ministry of Education. This database provides information about the yearly amounts of contracts and grants obtained by Quebec academics. Because no such database exists for Canada, we thus restrict our sample to the Quebec patent-paper pairs. We calculate yearly public and private funding, grants and contracts, for each individual Quebec university scientist.

Because the number of forward citations and the number of claims are count measures and by using Poisson regressions our models suffer from over dispersion, the preferred method of analysis chosen is the negative binomial regression (nbreg) that relaxes the assumption of the mean variable being equal to the variance. For the originality index models, a simple ordinary least squares (OLS) regression model was used. In addition, our models potentially suffer from endogeneity problems due to unobserved heterogeneity, omitted variables and the facts that patent quality may depend on the total average amount of contracts received by academic-inventors which are also related to the amount of grants raised (an explanatory variable). We therefore treat the contracts as an endogenous variable and correct for potential endogeneity by employing the Two-Stage Residual Inclusion (2SRI) method suggested by Terza et al. (2008) and use Two-stage least squares for the originality index and for the natural logarithm of the number of claims and the number of citations.

Our results find a negative impact of patent-paper pairs on the patent quality including number of forward citation, number of claims and an originality index. Prior studies have concluded that there is no significant impact of patents that have paper counterparts on patent citation flows. Our results corroborate these findings and further suggest that patents owned by universities and where the academic inventors are highly cited in the science world, yield a smaller number of citations in the technology world (lesser number of patent citations). The proximity between science and technology however crucial for knowledge transfer in general does not seem to have an equal importance when it comes to having an impact in the technology world. This is further exemplified by the lack of significance of academic assignees, but more importantly by the negative impact of a greater proportion of academics amongst the inventors. We however found mitigating factors, such academic inventors holding a prestigious chair, or having published highly cited articles, which contribute positively to the “quality” of the patents to which they contribute.

ABSTRACT. The number of international patent filings in many countries has been decreasing during (or even before) the recent economic crisis (Neuhäusler et al. 2013). The impacts on patenting are particularly evident in countries that are highly active in the USA, including the United States itself. In countries like Germany or Japan, the crisis seems to have affected patenting less severely, but a reduction in the international patent activity is still observable.

The drivers behind this decrease, however, are still unclear, or in other words, the question of which mechanisms within the innovation processes of firms, or on aggregate, whole innovation systems, are responsible for the decrease in patent filings during the crisis remains unanswered. On the one hand, it has been argued that a decrease in R&D expenditures, including delayed starts of R&D projects or prolonged innovation processes, indirectly affecting the output side of the innovation process in terms of patentable inventions, is responsible for the decline in patent filings (Rammer et al. 2012). In some countries, however, only a slight decrease in R&D expenditures during the recent crisis can be observed, which is particularly true for Germany (Schasse et al. 2012). Consequently, a more direct effect on patenting activity emanating from the economic crisis, namely adopting a cost-saving (international) patent strategy, can be assumed.

Based on these considerations, it can be assumed, that changes in the input side, i.e. a decrease in R&D expenditures, has lead to a decrease the firms' patent activities. However, it can further be hypothesized that firms have adopted a cost-saving patent strategy during the years of the crisis. This strategy might include a) not to file (mostly domestic) national filings also at an international level. Since keeping patents within the system up to a possible grant (and beyond) is also associated with a given amount of costs (examination fees, processing fees etc.), it can further be hypothesized that firms b) withdraw already filed patents or c) lapse already granted patents more often than in the years before the crisis. Finally, firms might simply choose not to file patents for already generated inventions at all, delay the filing to a later point in time or to use less expensive ways of intellectual property protection, e.g. secrecy (Cohen et al. 2000, Blind et al. 2006).

Both mechanisms, input (indirect) as well as output (direct) related, are analyzed separately within this study. The data for the analysis are based on a matched panel dataset on R&D investments and patent activity at the firm level for all companies that have performed R&D in Germany in 2007 and 2009. Furthermore, a patent dataset based on the PATSTAT database, including various indicators on the dimensions of changes in patent strategy for ten countries (CH, DE, FI, FR, GB, IT, JP, NL, SE, US), was applied. Besides descriptive statistics, multivariate models on the changes of the relation between R&D expenditures and patent filings as well as the changes in the patent strategy (share of transnational patent filings in patent families, share of withdrawn patents in the EPO patent stock, share of lapsed patent filings in the stock of granted EPO patents) were estimated.

Preliminary results show that patent processes have uncoupled from the R&D process during the crisis. Although the effect is comparably small (at least for Germany), it is notable that it can mostly be attributed to a decrease in the correlation between extramural R&D expenditures and patent filings. The correlation between extramural R&D and received patent citations, on the other hand, has been increasing. This implies that firms have cut back external R&D projects and only the most technologically promising external R&D projects have still received funds. Intramural projects, on the other hand, seem to have received a rather constant financing. With regard to the changes in patent strategy, it can be observed that especially withdrawing and lapsing patents are prominent cost-saving strategies. The strategy of filing patents at the international level less frequently, however, seems to play a minor role for most patenting companies.

These results have two kinds of implications. First, measuring innovative activities via patent indicators could lead to a distorted picture of the innovative output of companies especially during times of recessions. The second implication is that firms are well able to develop differentiated cost-saving patent strategies, which increase their room to maneuver also when costs have to be cut in order to stay competitive, at least for a certain amount of time. Since a given level of innovative potential can thus still be obtained, a potential innovation-hindering effect due to a reduction of R&D might, at least in the short-term, not be that severe.

ABSTRACT. Studies have depicted that the rate of unused patents comprises a high portion of patents in North America (35% Non-use on average), Europe (37% Non-use on average) and Japan (64% Non-use on average). The importance of the issue of nonuse is also highlighted within the literature on strategic patenting, IPR policy and innovation economics and in this regards, the literature has paid particular attention to blocking patents. Moreover, the current literature has emphasized on the role of patent pools in dealing with potential issues such as excessive transaction cost caused by patent thickets and blocking patents (overlapping IPRs) that might hamper the use of patents in the market for technology. In fact, patent pools have emerged as policy tools facilitating technology commercialization and alleviating patent litigations among rivals holding overlapping IPRs. Accordingly, patent pools may favor the use of the pooled patents through decreasing licensing transaction cost and providing equal and non-discriminatory access of all the members and potential licensees to the pool’s technology. This might be seen by companies involved in technology markets with excessive transaction cost and high IPR fragmentations as an opportunity to exploit their patents through participating in patent pools. In fact companies willing to license their patents through patent pools are taking advantage of faster, easier, broader and less costly access to the pool licensees (technology implementers). Nevertheless, in this study we argue that the willingness to use patents through pool participation by a pool member involved in high fragmented IP markets is not limited to the use of those patents that it includes in the patent pool which will be automatically licensed out after inclusion. Becoming a member of a patent pool may also favor the use of blocking and non-blocking patents held by a pool member outside the pool (non-pooled patents) by providing the opportunity to use the patent internally or externally in collaboration with other members. Such an effect can be the result of mechanisms provided by the patent pool at three different levels; Patent level mechanisms comprising greater access to the pool’s complementary technology for members, Firm level mechanisms including formal collaboration inside the pool and informal collaboration outside it, changes in members strategies as a result of pool membership, enhanced information sharing and increased technological spillover , partnership opportunities and greater access to other parties’ complementary assets and Market level mechanism including broader and less costly access and connections to the technology implementers (licensees). Considering such potential mechanisms favoring the use of non-pooled patents held by a pool member, in this study we show that pool members participating more intensively in patent pools are more likely to be willing to use their non-pooled patents through pool participation. Furthermore, we show that pool licensors are more likely to be willing to use their non-pooled patents by participating in patent pools with higher level of technological complementarity to their own technology. This study contributes to the current discussion on social, economic and technological benefits of patent pools. We show that companies involved in highly fragmented IP markets, do not only see participation in patent pools as an opportunity to facilitate their patent (technology) commercialization process by introducing their patents to the pool, as in the absence of a pool they can’t use their technology without infringing on their competitors’ patents, but they also see pool participations a way to facilitate the use of their unused patents which have not been included in the pool. Moreover, in order to contribute to the better understanding of how pool participation explains the willingness to use non-pooled patents we investigated those characteristics of the patent pools which are associated to willingness to use these patents through participating in patent pools. The results of this study have practical implications for strategic decision-making and for policy makers dealing with the issue of patent nonuse, overlapping IPRs and cumulative innovations.

ABSTRACT. The locus of wind power innovation has shifted from government laboratories, to private companies, to innovations in electricity grid and market rules and the development control systems provides an important window into how new technologies, supporting policies and institutions must co-evolve to support large scale implementation. It has involved parallel technological advancements in materials, aerodynamics, structural engineering, and control systems taking place in an international context of technology innovation and supported by both public policies and private investments.

With over 310,000 megawatts (MW) of wind built globally in 2014, integrating wind power into electric power grid operations is critical area for technology innovation and policy research. In the United States, wind power has evolved from a novel experimental technology, to a major part of the electricity system, with over 60,000 MW of wind power on the electric grid. In some states, like Iowa and South Dakota, wind-power generated electricity provides over 25 percent of electricity. But the variable nature of wind power has also required innovation in integration and wind power has changed how energy systems and markets operate {Wiser, 2013;Stephens et al., 2015}.

In this paper, we focus on the critical interplay between policy, technology innovation and integration of wind into electricity systems, with a concentration on the co-evolution of technological and institutional changes. To do this we focus on the challenges of wind integration in practice. Because wind resources vary over time and space and because the technologies and institutions of the legacy electricity system also differ, studies examining the development of large-scale wind power must be grounded in specific energy system contexts and focus on the multiple decisions shaping implementation in practice {Sandfort and Moulton, 2015}. While in theory there may be no difference between theory and practice, in the practice of wind development the contextual differences and decisions shaping siting, supporting transmission infrastructure development, and changes in electric market and power grid operational rules become critical. These differences shape the operation, value, and viability of wind technologies. We compare how the development of large-scale wind has led to different challenges to policy, institutions and technology innovation. We examine the creation of the Dispatchable Intermittent Resources (DIR) program in the Midcontinent Independent System Operator (MISO) region and the development of Energy Imbalance Markets within the Western Region of the United States.

MISO: Wind Power Technology and Policy Innovation for Deployment MISO created the DIR policies to solve problems of integrating variable wind resources into the electric power grid. Within the MISO region, wind generated electricity is often highest during winter nights when electricity demand is lowest. As wind power in the MISO system increased to the 13,000 MW today, the mismatch between the wind resource and system demand began to cause problems for grid operators. When wind power was a small resource, it was considered “self-scheduled” (MISO 2011) and as the least resource expensive on the system, wind turbines supplied power to the grid when the wind blew. But high wind production distorted electricity markets and forced system operators to manually curtail excess wind production and the system needed to innovate to accommodate wind power. We trace how technological innovations in wind power control systems and wind prediction models allowed for the new policy development of the DIR program and examine the institutional changes needed to develop the DIR. The DIR allows wind power to participate in day ahead electricity markets and automatically curtails wind plants for economic or reliability reasons, yet developing the DIR required several important policy and technology innovations.

Creating the Energy Imbalance Market: Wind Power Innovation in the Western U.S. Unlike the MISO region, the Western United States does not have coordinated electricity markets and integration of wind power faces a different policy and institutional challenges. The expansion of the Energy Imbalance Market (EIM) from the California Independent System Operator to other electricity balancing authorities in the Western Interconnection highlights the political and institutional challenges for large scale wind development which require new levels of policy innovation among regional energy stakeholders.

Together, these cases highlight the ongoing and interlinked roles of technology innovation, energy and science policy, and the evolution of stakeholder and research needs in practice. This research helps to critically explore the shifts in technology innovation in the context of large-scale implementation of renewable energy in practice.

ABSTRACT. Developments in the fields of nanotechnology, biotechnology, ICT and Cognitive science - or NBIC - can lead to disruptive innovation that deregulates existing practices. Prevalent reactions by Dutch policy makers and politicians tend to fall into one of two extremes. Either to cut regulation in an attempt to stimulate innovation, or to employ regulation as a brake on risky developments. Both approaches are however inadequate and create complications of their own. This project uses case studies to show how developing NBIC-technologies and regulatory regimes shape each other. The four case authors discuss: Liability for authorship on internet (Maurice Schellekens); Synthetic biology and different IP-regimes (Henk van de Belt); Innovation paths for automated driving in the Netherlands (Bonno Pel); and the Regulatory arrangements for European nano-medicine (Bärbel Dorbeck-Jung). The cases provide interesting lessons on how regulatory arrangements can be designed to deal with disruptive innovation in a more productive manner. NBIC-technologies produce two types of uncertainty. On the one hand, it is unknown whether innovations are permissible, acceptable, responsible and/or desirable. On the other hand, it is not clear whether current regulation is effective, legitimate, practical and/or appropriate. Both types of uncertainties need to be reduced, in order for innovations to become socially embedded. The question is how and to what extent this is possible. First, risks are unavoidable. Rather than using regulation to avoid or marginalize risks, the challenge is to deal with them, that is: acknowledging risk in a productive manner. This implies that the regulatory process requires the capacity for knowledge building, stakeholder engagement and responsiveness. Second, the regulation of (potentially) disruptive innovations tolerates no delay. On the one hand, because there are always regulatory regimes in effect that influence innovation - the better these interactions are understood, the better the capacity to respond. On the other hand, because the stage when technologies consist mainly promises and fears, is precisely when attention should go towards the question of what is acceptable. Also, addressing issues in a constructive manner (the process) is much more important in these early stages than defining correct measures (the content). Third, ’regulators’ and 'innovators' depend on each other for reducing uncertainties: regulation requires cooperation. The challenge is that different parties have different interests and conflict - as risk - is unavoidable. The objective should be to maintain tensions at a bearable level, rather than to eliminate them. Cooperative arrangements are most effective when responsibilities are clearly invested, when the roles assigned are well aligned with a party’s interests and abilities, and when other stakeholders are also included in some way. Finally, regulation requires a wide approach. ”Hard law”, "soft law” and self-governance tools should be combined in a manner that fits the specific situation. Flexible tools help allow for experimentation, knowledge building and learning. Yet ‘harder’ legal instruments are also always required. While the slow pace of law may constitute an obstacle for highly variable innovation paths, it also provides reliability and informed decision. Governments can be involved both as a regulator and as an innovator, in which case the interest is to serve policy goals. Other stakeholders - be they businesses, academic researchers or NGO’s - may find themselves in different roles as well. A suitable role for government - as follows from its duty to ensure an effective legal system - would be to orchestrate the required adaptations that a regulatory regime requires over time, as power and interests keep shifting with developments. But this demands a very careful management of possibly conflicting roles. As it turns out, the basic condition that any regime ought to fulfill in order to regulate disruptive innovation in a constructive manner, is to maintain trust among stakeholders.

ABSTRACT. Nanotechnology is considered the technology of this century because their high potential impact in the different economic activities (Roco etl, 2007; Shapira et al 2010; Hyun et al 2014), in this sense, new materials, products and services have been and will be developed in a wide range of applications (Wiek et al (2007). The development of nanotechnology depends of different factors, one of these critical elements are the networks (Bozeman, et al 2007, Wick et al, 2008; Malerba, Vonortas, 2009; Roco et al 2011), because by their nature, the collaborative and transdisciplinary work is fundamental. Networks are increasily recognized as an important mechanism of public policy and relevant instrument of one National Innovation System (Ludvall, 1992; Malerba, Vonortas, 2009), because it facilitate the access to resources, skills, global knowledge, the coordination of different actors, commercialization of complex technologies (like the case of nanotechnology), etc. (Mitchele and Sinlge, 1996; Swamina Thais, et al, 2001, Malerba, Vonortas, 2009, Friedler, 2010, Hyun et al, 2014). Etzkowitz (2008) has the vision that innovation networks needs to be fortified with partnership private and public in a context of a triple helix. In this innovation model, Government support with innovation polices and funds, firms with the knowledge of markets and funds and universities developing their third mission: the entrepreneurship one ( with patents, spin off, incubators, etc).

Otherwise, countries around the world have been making strong investments in Nanotechnology and specially deploying funds for implementing national networks(Maclurkan, 2005, Youtie, Shapira, 2008; Wick et al, 2008; Shapira et al , 2010, Rafols, 2010; Hyun et al 2014), Brazil is not exception. In front of this, each more practitioners and policy makers are interested in the management and specially in one topic, in their governance and impact of the performance of this kind of networks. Unfortunately the research about this issue is scarce. Exist in the literature some exceptions, but they are fragmented visions and not necessarily with focus to Nanotechnology. Inclusively exist some approaches against the network governance, because networks are not legal entities (Ness, et al, 2005; Provan et al 2007; Youtie, Shapira, 2008; Moller et al, 2005, Wick et al, 2008; Friedler, 2010, Rafols, 2010; Roco et al 2011). Managerial Toolkits than can support the network governance are also scarce. But according (Provan et al , 2007) some form of governance is necessary to ensure that participants engage in collective and mutually supportive action, that conflict is addressed, and that network resources are acquired and utilized efficiently and effectively. The focus on governance involves the use of institutions and structures of authority and collaboration to allocate resources and to coordinate and control joint action across the network as a whole. From a policy perspective, it should be clear that selection of governance and of and specific form, whether through mandate or funding incentives could have critical implications for overall network effectiveness. (Provan, Kenis, 2007, Arras et al, 2007; Nassimbeni, 1998, Malerba, Vonortas, 2009; Rafols, 2010; Hyun et al, 2014). It is important to mention that the development of emerging technologies is by definition uncertain comprising opportunities and risks, effective regulations and sustainable governance are also critical for the development of this potential industry (Wiek et al, 2007)_.

Thence, with the aim to contributing somehow with this research area, this on going paper pretended to develop an analytical framework which one could be an useful tool to analyze the governance in innovation networks. The nanotechnology area was chosen as was mention above, because their relevance as strategic area for the development of different countries and industries. Some important aspects must be considered when the framework is used, for instance the type of network (Arraz, 2007; Bessant, 2008; Malerba, Vonortas, 2009), the industrial dynamic (Rafols et al, 2010), because they have impact in the innovation network governance. Additional at the proposal some managerial toolkits were developed.

The research was exploratory and qualitative (Vergara, 2012; Yin, 2006). The qualitative approach, was supported in an intensive literature review and complemented with a semi structure interview applied to the main key actors of some relevant segments of the Brazilian Nanotechnology innovation networks.

It is expected that this work will be useful for the different actors related with this network, specially as a guide to improve the network governance and the formulation of public innovation policies. As futures works, the current proposal will be validated in others segments of the Brazilian nanotechnology network

ABSTRACT. The plug-in electric vehicle (PEV), as an innovative technology, its market has seen strong growth in recent years, both in purchase volume and charging infrastructure. Annual PEV sales are projected to reach 360,000 vehicles by 2017, with a large portion of these sales occurring in California, New York, and Florida (Pike Research, 2011). This market includes both plug-in electric hybrid vehicles (PHEVs), and full electric vehicles (EVs). While private industry has helped spur market growth through cheaper technology, public free-to-use charging stations, and wider electric vehicle offerings; government incentives have also played a key role in market development and reducing PEV usage barriers (Egbue, 2012). However, very little is known about the effects that these government incentives and infrastructure development have on the technology adoption of PEVs. This paper seeks to answer this question using statistical models with empirical data on government incentives, charging station infrastructures and PEV sales. With empirical data of electric vehicle sales for California cities in 2012, we estimate a series of regression models to test the impacts of government incentives and charging infrastructures on PEV technology adoption. Preliminary results show that public charging stations are found to be a very powerful government incentive, and should be a continued initiative for local and state governments. Results also indicate a stronger relationship of incentives to PEV adoption rates. The strongest incentive programs were those most directly related to PEV at-home or at-work charging, specifically those that offer a tax credit or deduction from energy utility bills. These programs promoted the most PEV adoption and appeared to resonate with consumers above other available incentives. This ultimately shows that PEV incentive programs do encourage adoption rates.

ABSTRACT. Background Presidential policy is formally conveyed through multiple mechanisms including, but not limited to, Presidential Proclamations, Presidential Memoranda, Statements of Administration Policy, Presidential Study Directives, Presidential Policy Directives, and Executive Orders (EOs) [1 – 3]. EOs are the most well-known among these types of documents. This is perhaps due to their long history of use (dating back to George Washington), requirement to be published in the Federal Register, and role as a medium through which many of the more influential presidential decisions are made available to the public. EOs cover a broad range of topics such as establishment of advisory committees, imposition of sanctions on foreign governments, adjustment of Federal employee salaries, and creation of specific initiatives.

Many EOs have a significant science and technology (S&T) component. Recent examples include Combating Antibiotic Resistant Bacteria (13676), Preparing the United States for the Impacts of Climate Change (13653), and Accelerating Investment in Industrial Energy Efficiency (13624). This non-negligible occurrence of S&T-related topics makes EOs an important corpus to consider when evaluating presidential interest in S&T issues over time. In addition, discussion and analysis of EOs has often focused simply on total counts of executive orders per president [e.g., 4]. This motivated us to perform a more sophisticated study directed toward a particular subset of EOs. We are not aware of other studies that have taken this comprehensive a look at S&T-relevant EOs.

Analysis After assembling a corpus consisting of the full text of EOs from 1945 (the beginning of Harry Truman's presidency) to the present, we performed various topical, temporal, and network analyses to discern features and trends in the appearance of S&T-related topics within EOs. An important temporal aspect of this EO corpus is the change in administration. The corpus spans 12 presidencies, six Democratic and six Republican. Whenever relevant, we assessed similarities and differences among administrations and political parties.

We identified S&T-related EOs through a variety of means, including mentions of agencies with a significant S&T mission (e.g., White House Office of Science and Technology Policy, National Science Foundation, etc.), topical queries, and manual tagging. We then developed a schema with which to categorize each S&T-related EO based on its topic and purpose to provide a more uniform means of comparing EOs in subsequent network and temporal analyses. We decided not to use the National Archives and Records Administration Executive Order Disposition Table Subject Indexes [5] as a topical schema because these subjects are too broad for our purposes and only were applied starting in 1993. In addition to charting the prevalence of S&T topics over time using our manually-assigned categories, we applied various text clustering methods to assess textual similarity between EOs and compared the results with the manual categorization. We then ascertained the extent to which automated clustering methods can be used to reliably identify related EOs. Having a manually categorized S&T corpus permitted us to estimate the accuracy of these methods.

One characteristic of EOs is that presidents frequently refer to other EOs within the text of any given EO. This is usually done to mention related efforts established by other EOs or to amend or revoke certain EOs (often those of a previous administration). This makes the EO corpus amenable to social network analysis techniques. After determining the linkages among all EOs in the corpus, we computed various standard network metrics to characterize how S&T-related EOs are distributed within the overall network, paying particular attention to the nature of linkages between and within administrations.

Conclusion This paper presents a novel approach to both the analysis of Executive Orders and the analysis of presidential interest in science and technology issues over time.

References [1] Relyea, Harold. Presidential Directives: Background and Overview. Congressional Research Service. November 26, 2008. Report 98-611. [2] Chu, Vivian S. and Todd Garvey. Executive Orders: Issuance, Modification, and Revocation. Congressional Research Service. April 16, 2014. Report RS20846. [3] Peters, Gerhard and John T. Woolley. “The American Presidency Project.” http://www.presidency.ucsb.edu/ [4] Mehta, Dhrumil. “Every President’s Executive Orders in One Chart,” FiveThirtyEight. November 20, 2014. http://fivethirtyeight.com/datalab/every-presidents-executive-actions-in-one-chart/ [5] National Archives and Records Administration. “Executive Orders Disposition Tables Index.” http://www.archives.gov/federal-register/executive-orders/disposition.html

ABSTRACT. Despite the unquestioned necessity to protect citizens from possible abuses in human subjects research and despite the good intentions of Institutional Review Board and, usually, researchers reviewed by these Boards, seemingly remarkable lapses occur to often, damaging not only the research subjects but also the public image of scientists and scientific research. In the face of well codified and extensive rules and generally well meaning and hardworking individuals, how does this happen? While there are several reasons why IRBs sometimes fail in their mission, the argument advanced here is that one of the most important why IRBs are not as effective as one might hope pertains to the interaction of their bureaucratic structures with the rules and rule systems to which they are subject. This paper interprets IRB from the standpoint of red tape theory (Bozeman, 1993; Bozeman, 2000; Bozeman and Feeney, 2011), arguing that if one were to set out to design an organization for the express purpose of ensuring red tape that IRBs would provide an excellent exemplar of design. Red tape theory suggest that problems inexorably occur when organizations are unstable and inconsistent in their members, when they are subject to extensive externally imposed rules that must be applied and interpreted locally and with considerable discretion, when there is little organizational learning or institutional memory and when formalism trumps judgment- all rampant with IRBs. The paper concludes with ideas for organizational reform.

ABSTRACT. This paper explores exchanges between research universities and corporations via individuals who represent firms while serving as university trustees. We argue that trustees of elite US research universities serve as channels for exchanges of ideas, human capital, and resources. This proposition poses important implications for the study of innovation. Numerous studies emphasize channels between university and industry such as the scientist-entrepreneur interface, technology transfer structures (e.g., Bercovitz & Feldman, 2011; Colyvas, 2007; Welsh et al., 2008), consultancies and contract research (e.g.,D’Este & Patel, 2007). Mathies and Slaughter (2013) conceptualized trustees as channels for innovation. We build upon their work by documenting the exchanges that trustees facilitate between universities and firms. We ask three questions about these exchanges: (1) Do trustees and trustees’ corporations enter into exchanges with the universities? (2) If so, what are they? (3) Do they vary across public and private universities and/or are they based on the extent to which universities are connected to firms?

MEASURING EXCHANGES We drew on one year of data (2010) from a NSF funded study of the network of interlocks to firms, nonprofits and government organizations (see Barringer & Slaughter (in press) for details). To emphasize variation across sectors and connectivity, we used the public and private universities that were most (MIT and the University of Pittsburgh (Pitt)) and least (Harvard and University of Minnesota (Minn)) connected to firms. We collected data on exchanges using Internet searches of university and firm websites for mentions of specific firms, trustees, or universities that were linked in the network. Searches, which are ongoing, followed a strict data collection protocol, and were rigorously cross-checked. Examples of exchanges that were related to innovation included: joint research, training, and education; partnerships; donations; and co-developed patents or publications.

SCOPE OF EXCHANGES MIT’s 79 trustees tied the institution to 92 firms. Currently we have 1,065 exchanges to date. Generally, these trustees do not seem to be “corporatizing” the university; rather, they seem an integral part of it, participating in the intellectual life and offering opportunities to the MIT community. In turn, members of the MIT community seem to regard themselves as part of a business cycle that underwrites their research, provides opportunities for start-ups, and returns funds to MIT in myriad ways. In short, these actors behave like colleagues engaged in shared knowledge/learning endeavors.

Pitt had 54 trustees through which it was tied to 71 firms. We have documented 1,182 exchanges to date. These exchanges were less science/technology focused than MIT, and dealt more with university units concerned with finance, e.g., the business school, and were more local/regional. Generally the trustees were primarily involved in university management rather than research. However, like MIT, the university provided infrastructure, a variety of resources, human capital, expertise and prestige for entities represented by the trustees. The organizations represented by the trustees did the same for the university, but on a much more limited scale.

Harvard was tied to 22 firms and nonprofits through 37 trustees. There are 708 exchanges so far, which differ substantially from the other universities. Like the other schools there are donors, alumni, administrator, faculty and lecturer ties. However these tend to be in the arts, social sciences and humanities. In terms of innovation the locus is on education and policy innovations rather than entrepreneurial scientific or financial.

Minn had 13 trustees through which it was tied to only 2 firms and which provided only 122 exchanges in total. These included extensive instances of trustees serving in various roles within the university, as well as some ties to government (mostly state rather than federal).

CONCLUSIONS Scholars have highlighted many different channels that connect universities to firms. We analyze the tangible forms that these relationships assume demonstrating that these exchanges are substantial, but that they differ notably by university. While we cannot make any generalizations to the population, our findings are suggestive and invite further empirical refinement. Results from MIT, for example, imply that universities with a technical emphasis, such as Cal Tech and Rochester, may prove particularly closely linked to the innovation economy. The regional and financial connections of Pitt suggest that some public universities may mirror private universities but within narrower bands of industry and/or geography. Analysis of UNC or Georgia Tech would allow for further assessment. Finally, the low connectivity of Minn suggests that some public universities may contribute to innovation and economic growth through other, more traditional means.

ABSTRACT. This study examines the presence and influence of women in U.S. science and technology policy at the federal level. We explore the proposition that who makes decisions about science and technology policy matters. The study examines the experience of women and men in leadership positions in six science-intensive agencies under two Presidential administrations, those of William Clinton and George Bush, both of whom appointed significant numbers of women to leadership positions in their administrations.

Through several different empirical approaches, we address three questions: (1) Does the presence and proportion of women in science and technology policymaking roles vary by type of organization, discipline, and appointment structure? (2) To what extent, and under what circumstances, do women in science and technology policy-making positions change the agenda, moving it toward a wider range of issues or toward different styles of policy production? (3) How does the (a) number and (b) proportion of women in the agency or on the decision-making body affect the influence women actors have on outcomes and processes? Are the effects different in science and technology policy leadership from those observed in other organizations? If so, why?

Quantitative analysis in our team (Smith-Doerr et al. 2014) has focused on occupational segregation and pay gaps for women, controlling for individual characteristics and grouping the agencies as “masculine” or “gender neutral” and “multidisciplinary” or not. The team hypothesized that masculine based science agencies (those with higher concentrations of physical scientists) would have larger pay gaps overall, but that women there would have to be exceptional to succeed and therefore might achieve pay equity (the “Marie Curie” hypothesis). Supporting both hypotheses, the results show that the gap was largest for women in female dominated occupations at masculine based science agencies and for women in male dominated occupations at gender neutral science agencies. Organizational context does matter.

The paper for the Atlanta Conference will report on qualitative data collected for the project. For this analysis, we began with a population of 1661 individuals identified in the Plum Book [a] as holding politically appointed positions [b] at senior levels in the agencies and administrations of interest (1202 men and 459 women). Preliminary observations from three focus groups held in Washington, D.C., contributed to the development of an interview protocol. The protocol asks about their experience in the agencies of interest, with particular attention to roles, agendas, decision making structures, and organizational cultures. The goal is to interview three men and three women from each of six agencies in each of the two administrations. Finding contact information and securing participation has been both time-consuming and challenging, and snowball sampling has been used to supplement the Plum Book search. Women have been much more likely to agree to an interview than men. At the time we are preparing this abstract, we have conducted 24 interviews, including five with men. Not many of these were agency heads themselves, but many worked for women who headed agencies.

Very preliminary observations suggest that the women at this top level of leadership were extraordinarily effective, as seen by those around them. With a few exceptions, they did not pursue women’s issues in particular but rather focused the attention of their agencies on increasing effectiveness with regard to the goals of the administration they were in. The women we talked to considered themselves effective as change-makers as well, although they did not see all women in government that way. They valued teamwork and often saw themselves and the women they worked for as particularly effective at building relationships. Many in the Clinton administration made broadening participation an important objective. So far, we have not found anyone who thought women systematically pursued different kinds of agendas from men; instead, their experiences were shaped more by the political visibility of these appointed jobs and complex relationships with a wide variety of external groups. We look forward to presenting a more sophisticated, NVivo-based analysis of a larger range of the interview material at the conference.

Reference:

Smith-Doerr, Laurel, Sharla Alegria, Kaye Husbands Fealing, Debra Fitzpatrick, and Donald Tomaskovic-Devey. “The Value of Women’s Work: Gender Inequality in Federal Science Policy Agencies.” Presented at the Association of Public Policy and Management, November 4, 2014.

[a] and [b] omitted to conserve characters in abstract submission, but will be provided in a final version.

The National Science Foundation supported this research through grants 1152800, 1152861, and 1152980. The views expressed do not necessarily represent the views of the National Science Foundation.

ABSTRACT. CONTEXT: The rapid increase in computing power allows researchers to manipulate and analyze increasingly large data sets, enabling a new form of data-intensive research that some have named the "Fourth Paradigm" of scientific discovery (Hey et al., 2009). Because data-intensive research requires costly computing infrastructure, significant investments by US funding agencies have been provided to support locally available high performance computing (HPC) resources. In previous research (e.g. Apon et al., 2015) we have found that the average technical efficiency in producing research outputs of academic departments that have access to HPC resources is higher than of departments without local HPC access. This raises the question whether funding for HPC has also transformed the nature of research and types of collaborations as suggested by the Fourth Paradigm. Can we quantify the change that HPC has made to the paradigms of scientific discovery, and can we observe acceleration of this change by funding for local HPC resources?

CONTRIBUTION: This paper advances previous research in two ways. First, building on punctuated equilibrium theory in public policy making (Baumgartner & Jones, 2009) we argue that access to local HPC resources has the potential to accelerate the shift towards the Fourth Paradigm of scientific discovery and to lead to a new equilibrium of research production. This punctuated equilibrium is characterized by (a) an increase in research based on computational methods, and (b) a shift towards "Mode 2" (Gibbons et al., 1994) knowledge production in which increasingly interdisciplinary teams of scholars collaborate to produce results. Institutions without locally available HPC resources, in contrast, will continue to produce the same type of research or only gradually move towards a new data-driven research equilibrium that is more difficult to sustain over a long period of time.

Second, we advance existing work by developing new empirical measures of research output. Research in higher education has traditionally relied on bibliometric indices to measure research efficiency, such as the number of publications or citations. A shortcoming of these indices is that they only provide a broad characterization of research output, but say little about the nature of the research that is done and the technology used in producing results (Toutkoushian & Webber, 2011). This is particularly problematic in the evaluation of funding for HPC or other technologies, where the goal of investment is not primarily to increase absolute numbers of research output, but to support capabilities in data-intensive science and engineering, and to enable researchers to produce output more efficiently.

METHODS: Testing our punctuated equilibrium theory of investment in HPC requires data across many institutions and over a long time period. To this end, we are using a novel data set that combines publication resources from an industrial partner with publicly available data sets. These data include information from about 200 US High and Very High research active universities from 1996 forward.

We develop innovative applications of content analysis to measure changes in the nature of research and its degree of interdisciplinarity. We do this by applying dynamic Latent Dirichlet Allocation (LDA) to our large database of scientific article abstracts that includes all published work in Chemistry, Economics, and Computer Science from 1996 to 2013. Analyzing articles in each discipline over a long period allows us to track changes in the development of research topics over time, and to identify whether there are significant differences in the proportions of topics related to computational methods before and after an institution invested in local HPC resources.

To measure the degree of interdisciplinary research, we test alternative bibliometric indicators based on citation analysis (e.g. Wagner et al., 2011). For example, counting the number of different academic fields cited in a paper or the number of citations outside the field in which a paper was published allows us to capture its wider degree of knowledge integration (Rafols & Meyer, 2010).

RESULTS: Our evaluation of a punctuated equilibrium theory of HPC investment will provide new insights into the short-term and long-term effects of technology-related funding. Our results will show whether there is empirical evidence for a shift towards more computationally driven research, whether locally available HPC resources accelerates this shift, and whether there are alternative pathways that require less funding, but perhaps proceed more gradually and at a slower pace. Because we analyze changes in the nature of research over a significantly long time period after local HPC resources become accessible, we are furthermore able to test whether investments lead to a new stable equilibrium of knowledge production.

ABSTRACT. In public policies supporting business R&D, companies are not only selected for treatment, but also awarded a different amount of support. The continuous treatment frame allows to investigate the form of the function of average treatment effects over all possible values of the treatment levels. Recent literature on public program evaluation has found out a convergent result of decreasing relation between the level of subsidy and the outcome (R&D additionality), as well as a notable substitution between private and public funds occurring for high intensity of public subsidy. However, such empirical evidence didn't find yet a sound theoretical explanation and this study aims at filling this gap. The innovative content of the paper rests on combining three streams of economic literature: (i) the policy ex-post evaluation literature, i.e. the R&D response function to different levels of subsidies; (ii) firms investment models introducing financial constraints; (iii) firm investment models including (convex and non convex) costs of adjustment. We refer to a vast set of studies to explain an heterogeneous sensitivity of firm investment to cash-flow injections: (i) models of investment including financial constraints (Fazzari et al., 1988; Kaplan and Zingales, 1997; Alti, 2003; Coad, 2010; for a survey Carreira and Silva, 2010); (ii) investment/capital adjustment cost models (Eisner and Strotz, 1963; Lucas, 1967; Gould, 1968; Treadway, 1971; Caballero and Engel, 1999; Abel and Eberly, 1998). The literature on R&D policy evaluation recognised that financial constraints have a negative impact on R&D investments, both in the meaning of “external” financial constraints, due to information asymmetry and agency problem and of ”internal” financial constraints, deriving from the availability of cash flow, i.e. from firm past net-profit. In this framework, subsidies are interpreted as an external injection (or shock) of cash-flow, which could also impact as positive signal towards external source of funds, allowing higher additional R&D investments. In the last 20 years a breakthrough in the literature on investment determinants concerned the introduction of a combination of financial and real frictions and the relevance of non linearities in the responsiveness of investment to fundamentals. We consider this a route to explore also within the policy evaluation studies. The positive or negative elasticity of R&D investment to different dose of subsidy depends on the interaction among many factors, which we can regroup in financial condition (liquidity constraints) and sunk costs (different types of adjustment costs), conditioned on firm’s structural characters. We estimated the effect of public research and development (R&D) support on company R&D performance by an application of a dose-response econometric model (continuous treatment model) proposed by Cerulli (2014) to real data. We employed a cross-section combining the 8th, 9th and 10th Unicredit surveys collecting a large body of information on various characteristics and activities of a sample of Italian companies. All surveys are built by stratifying on sector, size and location, thus being representative of Italian manufacturing companies with more than 10 employees. The final dataset has merged with companies’ balance sheet data coming from the AIDA archive. The specification of the output equation comes from the widely accepted view of the main drivers of company R&D performance in the R&D policy evaluation literature (David et al., 2000; Cerulli and Potì, 2012). We make a detailed exploration of the support-intensity response functions, by looking both at the pattern of the average treatment effects conditional on the covariates and at the factors which characterize and discriminate the two sub-groups of beneficiary firm through a multivariate probit analysis. The application of a dose-response model resulted in a positive effect of the R&D subsidy policy driven by companies getting a comparatively lower share of R&D covered by public support (no more than around 20% or 15% for gross and net R&D respectively). The public investment dose reaches a threshold after which there is no further advantage to invest public resources. This is a result common to other empirical studies using a dose–response model. These results ask for a better understanding of the reasons and factors impacting on this counterintuitive conclusion. Three clear results can be drawn from our study: (i) liquidity constraints or funding structure discriminate between different firm response to the public subsidy; (ii) the size of the R&D project, used as a proxy of fixed adjusting costs, discourages firm additionality behavior; (iii) there is a positive association between firm size and R&D additionality: larger firms are more responsive to public subsidy than smaller ones, a result that would ask for a rethinking of R&D policy design, especially those formally directed to SMEs.

ABSTRACT. The theoretical rationale for public R&D is well established; however, empirical studies clarifying the most effective means of investment are lacking. This paper exploits policy discontinuities at state borders to examine the differential effect of R&D policy on innovative projects. The Small Business Innovation Research (SBIR) State Match program provides a useful context to examine this variation by offering a noncompetitive match to successful SBIR Phase I projects that are then eligible to compete for the larger SBIR Phase II award.

This paper includes the population of Phase I awards for two states with the State Match – Kentucky and North Carolina – and four contiguous states without the State Match – Arkansas, Missouri, South Carolina and Virginia. Importantly, the two treated states were selected given that they adopted the program in 2006 and 2005, respectively, and offer varying sizes of the match. This allows for a differential and marginal examination of the policy. Placed within the context of the federal SBIR program, between state comparisons are feasible given that the population of Phase I projects achieved a minimum level of competency through the federal SBIR review. Moreover, this approach offers a strong design not only to treat the noncompetitive State Match as an exogenous increase to the R&D investment, but also to examine innovation policy directed toward projects rather than firms.

Results point to a heterogeneous treatment effect. Most notably, in contrast to projects without the State Match, the results indicate that additional funding improved the competitiveness of Phase I recipients competing for the Phase II awards among projects with a directed focus in the basic sciences – NSF and DHHS funded – and with less prior success with the SBIR program. Results from a comparison of means test indicate that these firms in particular are smaller and younger than the average SBIR recipients. Results from the marginal analysis also found positive effects of the program; for this set of models, the results indicate that more money is worth more when invested in projects with less experience with the program. Program success was defined as securing the Phase II award; however, employment change was also considered. The results for these models were not robust; this is likely due to measurement error with this metric.

The policy implications extend to both state and federal policymakers and firms invested in innovative research. Rather than offering a blanket policy of a standard State Match – as is the standard practice – policymakers may want to tailor the award. State policymakers may view these results as evidence that marginal investments can leverage significant follow on funding when invested in specific types of innovative endeavors. Federal policymakers, on the other hand, may consider adjusting the structure of the SBIR Phase I award for certain programs.

ABSTRACT. How can health innovation contribute to match social goals and market functioning? This is one of the main health innovation challenges. Any way to overcome it should include a policy mix that articulate policies and encourage market functioning. The aim of this paper is to analyze whether the policy reform in Uruguay of the last decade created a policy mix for health innovation. Policy mix refers to the combination, balance and interaction between different polices and policy instruments to achieve certain outcomes. Innovation policy embraces all public measures that boost or block innovation activities in a specific area. The policy mix for health innovation includes instruments elaborated by several public offices. It should articulate, at least, three policy arenas: STI policy, health policy and industrial policy. Innovation policy faces a twofold coordination challenge. First, a coordination among public offices, which usually do not act as a singular actor perfectly coordinated. Second, innovation policy should provide incentives and regulations that articulate policy objectives with market agent interests. That is particularly hard in a low innovative context, where knowledge demand from productive agents experiments chronic weakness. The analysis of the Uruguayan experience offers an interesting case study of a recent policy reform that has, at least at a rhetorical level, a systemic approach. Since 2005 the government has promoted a reform of both health care area and STI policies, which included institutional changes and budget increases. Nevertheless, even though STI policies proposed a systemic articulation for health area, those reforms have not been properly articulated. STI national policy is based on horizontal instruments oriented to the promotion of basic research and innovation activities. Meanwhile, national health reform was concentrated on health care prevention and health care access extension. But in the technological field, it only proposed the rationalization of the technological health costs through the central coordination of high tech medical services. Besides, since 2008, a new generation of industrial policies was implemented. These policies were oriented by a sectoral logic and included STI dimensions, but there was almost no connection with other policies. Based on a qualitative methodology –documentary revision and interviews with public officials and sectoral experts- we analyze evidence about the evolution of the policy mix for health innovation, organizing the study through three dimensions: i) The changing role of the State. The convergence –even in a non-coordinated way- of different public policies oriented to the health system, shows the new role that the Uruguayan State assumed since 2005. ii) The articulation of the different policies and instruments that compose the policy mix for health innovation. We analyze the systemic characteristics of the policy mix by considering: the policy agenda and rationale; the governance structure; the specific objectives of policies; and its instruments and implementation in the three public arenas specified above. We also build a health innovation policy typology based on the classic distinctions between vertical and horizontal policies; demand led and supply policies; and implicit and explicit policies. iii)The incidence of innovation policies on market functioning. One of the main objectives of innovation policies in developing countries is to boost demand for innovation and to create market relationship for innovation process. Based on evidence from previous works we analyze the challenges of policies regarding health market barriers and the innovative weakness of productive agents. Our results show that public policy changes implemented during the last decade implied the creation of the first STI explicit policies in Uruguay, while the creation of sectoral, health care and industrial policies, defines the strategic objectives in healthcare delivery and industrial development. However the Uruguayan policy reform did not build a policy mix for innovation health with a systemic rationale. Even if the role of the State was strengthened in both regulation and health care provision, the health reform was concentrated on the expansion of the health care system. Innovation policies were mainly devoted to horizontal, supply-side instruments, and there were almost no objectives dealing with innovation in the health policy agenda. Besides the interest on the specific policy reform, the analysis of the Uruguayan case offers interesting lessons for innovation policy in developing countries. A policy mix for health innovation should be strengthened by articulating the health innovation demand emergent from health care system with innovation and industrial agendas. It would require a revision of innovation policy instruments, mainly vertical instruments, to promote the innovation demand in health.

ABSTRACT. Central arguments behind the concentration of research funding in centers or larger projects is that it creates a critical mass (Johnston 1994, Bloch & Sørensen 2014) and that concentrating funding among top researchers increases production of high quality research and enhances conditions for the generation of path-breaking research results (Hicks and Katz 2011). Cumulative effects may also play an important role for research groups or units. The “best” units are argued to benefit from self-enforcing processes, confirming and strengthening their status, yet there may be limits to cumulative advantages (Merton 1968, DiPetre and Eirich, 2006; Fortin & Currie, 2013). The purpose of this paper is to examine the relation between size and performance for research grants, with particular focus on Centers of Excellence (CoE). Are publication and citation performance increasing in the size of grants? Can we identify and characterize accumulative processes over grant periods and what is the impact of past performance and other factors such as collaboration? Size may have a positive impact on performance if it facilitates research or research collaborations that would not have been fully possible otherwise, or at least not to the same degree. Conversely, large centers may create administrative burdens and be less efficient than smaller projects. There may also be a dynamic element through the accumulation of knowledge and recognition through grants. More specifically, the paper examines the relation between grant sizes for CoEs, funded by the Danish National Research Foundation (DNRF), and research performance. In order to examine performance for a wider range of grants in terms of size, we also include a set of research grants from the Danish Council for Independent Research (DFF), which are typically much smaller than CoE grants from the DNRF. The analysis is based on an extensive dataset that includes both grant amounts and lists of journal publications for each grant. The publication data has then been used to construct a number of indicators regarding citation impact and collaboration. The sample includes 57 CoE grants from the DNRF and 129 grants from the DFF. We examine four different measures of research performance: mean normalized citation score (MNCS); the proportion of highly cited articles (top 10%); overall productivity and productivity of highly cited articles (top 10%) per million euros. Two analyses are conducted in the paper. The first and main part is a dynamic panel data analysis using Generalised Method of Moments (GMM) that examines both the role of time/accumulation and of size for DNRF CoEs over the course of grant periods. Size is measured by the annual grant amount and by the annual number of articles. To allow for the possibility of decreasing returns to size, the square of annual grant amounts is also included. Accumulation of knowledge and expertise through earlier performance is captured through the lags of publication and citation activity and also through the degree of collaboration both among grant participants. Finally, the grant year and its square are included to capture trends in performance over the grant period. Secondly, and in order to examine the role of size across a wider range grant amounts, a cross sectional analysis examines the role of size for overall research performance for both DNRF CoEs and the smaller DFF grants. The results of our analysis will be compared with that of earlier work on the relation between size, CoEs and research performance (such as von Tunzelmann et al. 2003, Ida & Fukazawa 2013, Rogers et al. 2012). References Bloch, C. & Sørensen, M.P. (2015) The size of research funding: trends and implications. Science and Public Policy 42(1): 30-43. DiPetre, T.A. & Eirich, G.M. (2006) Cumulative Advantage as a Mechanism for Inequality: A Review of Theoretical and Empirical Developments. Annual Review of Sociology 32: 271-297. Fortin, J. M., & Currie, D. J. (2013). Big Science vs. Little Science: How Scientific Impact Scales with Funding. PLoS ONE, 8(6). Hicks, D., & Katz, J. (2011). Equity and Excellence in Research Funding. Minerva, 49(2), 137-151. Ida, T. and Fukuzawa, N. (2013) Effects of large-scale research funding programs: a Japanese case study, Scientometrics 94: 1253-1273. Johnston, R. (1994) Effects of resource concentration on research performance, Higher Education 28: 25-37. Merton, R.K. (1968) ‘The Matthew Effect in science’, Science 159(3810): 56-63. Rogers, J., Youtie, J. and Luciano, K. (2012) ‘Program-level assessment of research centers: Contribution of Nanoscale Science and Engineering Centers to US NNI goals’, Research Evaluation, 21, 368-380. Seglen, P.O. and Aksnes, D.W. (2000) ‘Scientific productivity and group size: A bibliometric analysis of Norwegian microbiological research’, Scientometrics 49(1): 125-143. von Tunzelmann, N., Ranga, M., Martin, B. and Geuna, A. (2003) The Effects of Size on Research Performance: A SPRU Review.

ABSTRACT. Mexican science and technology policy (S&TP) has been formally operating since 1970, when the National Council for Science and Technology (CONACYT) was founded, however the introduction of innovation objectives and goals, as well as, strategic planning approach and methods were incipiently introduced in the early 1990s and more systematically until 2011. Long term strategical planning and foresight methods are critical for S&TP and its potential impacts in creating national capabilities in scientific, technological and innovation knowledge. Nonetheless in the Mexican case, far from being institutionalized as a regular practice, these timid efforts have been struggling to become formalized and methodologically deepened as permanent activities within the STI policy cycle, which represent one of the main weaknesses of the Mexican STI policy initiatives.

This paper is intended to analyze institutional hindrances and challenges for the introduction and implementation of strategic methods and contents for designing national innovation policies in Mexico during the period 1970-2011. We are interested in identifying and charactering the different approaches and methodologies that have been implemented at federal level for building diagnosis, selecting priorities and goal setting. In this sense, we are also concerned with the analysis of the role of long term planning and technology foresight initiatives as an expected component of innovation policies. Finally, this paper searches into the main institutional failures which may explain why these approaches have not been formally and systematically implemented by the Mexican government.

This study is based on a historical policy analysis perspective and draws its main inputs from a couple of projects. The first one was an academic endeavor coordinated by the Scientific and Technological Consultative Forum (FCCyT)(Casas et al. 2013). The second project is a current research endeavor aimed at understanding the institutional challenges that technology foresight has been facing to become a recognized approach to deal with consensus building, priority identification, collaboration and long term planning in Mexico.

Our paper will be structured into five sections. Firstly, we provide a conceptual framework to observe and select relevant categories for the rational design of STI policies. This framework will contextualize our definition of strategic STI innovation policies in which the development and selection of long term scenarios has been a useful component. Additionally, we rely on some institutionalist concepts to understand the interaction of formal and informal rules and routines within the policy arena. Particularly, we benefit from the ideational and discursive institutionalism approach to observe and analyze the role of different actors in the designing of policy measures. (Surel, Y., 2000; Genieys, W., and M. Smyrl, 2008; y Schmidt, V.A., 2008)

Secondly, we review the historical background in the designing of national innovative policies from 1970 to 2011. In this way we try to observe to what extent, during this period, national STI policy has benefited or not from the learning of its practitioners and scholars; whether it has incorporated strategic elements for the conduction of strategic polices and programs; we observe its time span and to what extent it has been fragmented and locked into particular trajectories.

Thirdly, we deepen our analysis by focusing on a particular initiative as a case study. We review the context, development, characteristics and results of the FCCyT’s initiative entitled “National Strategic Objectives for the Special Program for Science, Technology and Innovation 2012-2037: Principles, criteria and methodology”. We analyze this experience in the context of recent reforms in the STI law (2011) and within the transition process between two political administrations, both in CONACYT as in the FCCyT.

Fourthly, in a next section we analyze the institutional challenges for the incorporation of long term planning and the conduction of technology foresight exercises. For this purpose we characterize the development of different policy approaches, the role of incumbent actors and the interaction of foresight specialists with policy makers.

Finally, we reflect on the main institutional challenges for the implementation of theoretically sound methods for the designing of strategic STI Policies, as well as for the introduction of long term planning and technology foresight. We link the institutional analysis at macro level in the STI policy arena with the meso level while analyzing organizational dynamics in CONACYT, FCCYT and science academies among other relevant actors.

Complete references emailed to D. Hicks, including them exceeded the abstract length limit.

ABSTRACT. With the rise of cooperation of universities with industry and government in the last 30 years, more graduate students are exposed to collaborations across sectors, disciplines and organization boundaries. Science and Engineering (S&E) disciplines, particularly, have shown to benefit from these collaborations as they often work with problems that require complex solutions where collaboration and exchange of different expertise are almost inevitable. One of the establishments that enable this cooperation is Cooperative Research Centers (CRCs). Previous evaluation of the CRCs offers some information about students’ experiences. However, little is known about professional social capital students acquire during their training and carry on to their professional careers. According to Bozeman’s Science and Technology Human Capital theory (Bozeman, Dietz & Gaughan, 2001), professional social capital is an important, but often ignored component of evaluation of the scientific outcomes. Social capital as manifested in professional connections, shared norms and trust are as valuable component of research as technical skills and knowledge. Bozeman claims that evaluation of the scientific outcomes and policies should involve human capital component that captures not only long-term outcomes, such as intellectual property and patents, but also the processes behind their creation. This study looks at the professional social capital, perceived career preparedness and satisfaction with graduate training and whether these outcomes differ by the type of graduate training: traditional training confined within university departments and training within NSF Industry/University Cooperative Research Centers (I/UCRC), one type of CRCs. The study also looks at the international students who comprise more than 50% of the graduate STEM population in the U.S. universities. The study also investigates whether type of training and other variables affect international students’ plans and intentions to remain in US and/or become US citizen after graduation. Professional social capital is captured at the time of graduate training of the future scientists and engineers. Similarly to any type of social capital, it contains two parts: bridging and bonding social capital (Patulny & Svendsen, 2007). Bridging represents available social networks, particularly, size of a person’s network and strength of the network connections. Bonding represents mechanisms that enable one’s social network. They represent trust and availability of positive norms and values about collaboration. In the context of the S&E graduate students, bridging social capital represents students’ social connections to and intended use of different kinds of professionals (US academics, non-US academics and non-academics). Bonding social capital represents the extent to which students view different kinds of collaborations. The study uses quasi-experimental research design. Graduate students trained in the team-based and sector-crossing I/UCRCs were compared to students who experienced a more traditional disciplinary training guided by an advisory committee. The contact information for the study participants was collected in two different ways. Since the authors of this study are part of the I/UCRC evaluation project’s team, we were able to receive contact information for center graduate students from the I/UCRC directors with assistance from the NSF. An electronic confidential survey was used as a tool to collect information from respondents. The traditional training control group was created by obtaining contact information for graduate students at matched universities and disciplines from public sources. Two groups were given the same survey instrument with some modification in wording and reduction of questions that are related specifically to the I/UCRC group. The final sample represents completed responses from 173 center students and 87 non-center students. Descriptive statistics of the sample demonstrate that two groups are compatible in: graduate standing (100% PhDs), age (M=26 y.o.), gender (23% female), immigration status (66% international) and time spent in graduate school (M=2 years). Preliminary findings of the study demonstrate that center students have higher scores on some measures of professional social capital but not others. Analyses also reveal that international students do not always share these benefits. Center students also have higher level of satisfaction and perceived career preparedness. Preliminary analyses also shed some light on the mechanisms involved in these impacts. Implications for policy and practice will be discussed.

Bozeman, B., Dietz, J. S., & Gaughan, M. (2001). Scientific and technical human capital: an alternative model for research evaluation. International Journal of Technology Management, 22(7), 716-740.

ABSTRACT. Policy makers are pushing to increase minority participation in science, technology, engineering, and mathematics (STEM) education (National Science and Technology Council, 2013). Policy makers seek to ensure that the STEM workforce reflects the population by increasing the participation of traditionally underrepresented groups, including both the previously mentioned racial groups and women, in post-secondary STEM education. The National Science Foundation produces a biennial report presenting data on women and minorities in post-secondary STEM fields in the form of “percent by field” and “percent of degrees awarded.” This report, however, does not compare these numbers with each group’s proportion of the college population (National Center for Science and Engineering Statistics, 2015). Similarly, a study by the National Academies examined the percentage of underrepresented groups across a wider range of sub-groups than the NSF report, including the college-aged population and the K-12 population (National Academies, 2011). Yet, this study presented the percentages from a single year, therefore lacking the longitudinal scope necessary to examine the STEM education pipeline. Analyses emphasizing proportional representation have been used in limited scope before. Research on the effects Historically Black Colleges and Universities (HBCUs) and Women’s Colleges have on STEM participation among underrepresented groups has included some use of proportional analyses. For example, research found that while HBCUs accounted for 22% of all bachelor’s degrees awarded to Africa-American students, these schools produced 30% of all STEM bachelor’s awarded to Black students (Perna, 2007), suggesting the importance of HBCUs in broadening STEM participation. This method of analysis, however, has not to date been applied across all gender and racial groups. 

This paper will examine proportional representation in postsecondary STEM education. A novel metric will be presented that can be to quickly understand the proportional representation of a group in both the college population relative to the college-aged population and in STEM degrees awarded relative to all degrees awarded. Representation will be presented as an expected value comparing two proportions. When examining representation of a group in STEM fields, this expected value will represent their share of STEM degrees relative to their share of all bachelor’s degrees, with the expectation that a group should have equal proportions in both. Similarly, when examining representation of a group in college graduates, the value will represent their share of bachelor’s degrees relative to their share of the US college-aged population. This expected ranges in value from 0 to infinity. A representation value of 1.0 denotes proportional representation, while a level below or above denotes under- and overrepresentation, respectively.

We will argue that this metric provides policy makers a new insight into representation in postsecondary STEM education. Policy discussions will be well served by using this new metric in conjunction with traditional measures of STEM participation. 

Preliminary analysis with this metric shows that traditionally underrepresented minority (URM) groups are still underrepresented in post-secondary STEM fields, with a STEM representation metric of 0.85 in 2012. Furthermore, the analysis shows that the representation gap between URM students and non-URM students has widened since the early 2000s. Until the late 2000s, URM men were actually over-represented in STEM fields relative to their proportion of bachelor’s degrees. Declining to a representation metric of 0.9 in STEM fields, URM men have now reached parity with non-URM women, who have seen a steady increase in their relative representation in STEM since the early 2000s. Additionally, URM women have remained level in their relative representation in STEM fields since the mid-1990s at approximately 0.75 representation.

National Academies. (2011). Expanding Underrepresented Minority Participation National Academies Press. National Center for Science and Engineering Statistics. (2015). Women, Minorities, and Persons with Disabilities in Science and Engineering. National Science Foundation. National Science and Technology Council. (2013). FEDERAL SCIENCE, TECHNOLOGY, ENGINEERING, AND MATHEMATICS (STEM) EDUCATION--5-YEAR STRATEGIC PLAN. Perna, L. L.-W., V.; Drezner, N.D.; Gasman, M.; Yoon, S.; Bose, E.; Gary, S. (2007). The Contribution of HBCUs to the Preparation of African American Women for STEM Careers: A Case Study. Research in Higher Education, 50(1), 1-23.

ABSTRACT. College personnel often encourage undergraduate research engagement as one important way to involve students in campus life, increase retention, bolster marketable skills, and increase underrepresented groups in science and technology (STEM) fields. But how do undergraduates themselves view the personal and professional outcomes of research participation? How do their perceptions vary compared to other college activities?

In this abstract we present our overall research problem, an abbreviated literature review, and some sample results. We utilize a situated learning perspective in reviewing institutional studies on student engagement. Student research experiences enable learning in context. Although concentrated in the life and physical sciences, research activities potentially could benefit all majors.

Under the auspices of a public southeast major Doctoral Research University Office of Undergraduate Research, in mid-2013 we invited 30,744 undergraduates to complete an online survey on “undergraduate experiences”. Items tapped demographics, research attitudes and several campus experiences . Respondent demographics, 13 perceived outcome Likert items (including a two-factor confirmatory analysis solution loadings) describing Personal Development and Professional Development during college were examined. 7,469 students completed virtually all questions. Multivariate predictors include gender, race/ethnicity, college level, academic major, and first generation student status. Space here precludes describing all analyses although supporting tables are included in the presentation.

27% of these undergraduates reported a research experience (more common among seniors, Asian Americans, continuing generation, and STEM students). Tables include t-tests on the perceived collegiate outcome items by research activities, a two factor Confirmatory Factor Analysis, and multiple regressions on the Personal and Professional Development factors

Surprisingly, we found that research participants rated their personal and professional college outcomes less positively than other students, although overall differences were small, reaching conventional statistical significance in part through the large n. When we conducted paired t-tests to compare research participants only on how they assessed research versus non-research college activities; they were more satisfied with their research experience outcomes than comparable “college experiences” on 9 out of 13 items.

We next examined a general hypothesis that the students involved in research were less engaged with other campus activities and perhaps more isolated than other undergraduates, possibly because of their concentration on scholastic and pre-professional activities. However, in fact, those undergraduates engaged in research participated more in other campus activities as well. Indeed, students felt these other activities (e.g., academic clubs) more positively promoted their personal and professional development, but research experiences (and being a STEM major) had more negative outcomes. In further analyses, the quality of the research experience (e.g., mentorship or social networks) was important in student perceptions of their education outcomes.

Although campus experiences other than research participation are important, we note that these results are student self-perceived outcomes; faculty and staff may view these outcomes quite differently. Complicating multivariate analyses, diverse student populations differentially participate on campus, e.g., women more often joined academic clubs or “Greek” houses, or engaged in international study. First generation students less often joined fraternal organizations or clubs or engaged in international study.

Thus, how students viewed their college outcomes by research participation was unexpectedly mixed. Obviously these activities are one route among several to campus integration. But, because activity participation differs by gender, ethnicity, and first generation college status, such variables moderate how undergraduates view research and other experiential outcomes . In some respects, the sexes, different ethnicities or majors, and first versus continuing generation students occupy unique campus niches. By reaching out to all student populations with a broad message, research activity may hold the potential able to positively influence student self-perceived college outcomes.

ABSTRACT. Hello, my name is Tiffany ray and I am the Executive Director of Generation Infocus, a STEAM (science, technology, engineering, art and math) focused K-12 educational program. I would like to propose a session on closing the gender gap in STEM careers through the implemetation of STEAM learing opportunities for girls.

Abstract: During the Wearable Technologies: Mobilizing Girls into the World of STE(A)M presentation , Tiffany Ray will discuss the relevance of girls engaging in STEAM at an early age. Gi will provide data-driven best practices that explore the use of fashion and technology as a means to attract young women into STEM fields.

Strand: Equity, Geneder and Diversity, STEM Education/Workforce

Session Format: Focus45

Session Description:

Tomorrow’s workforce will demand talented STEM/STEAM-related skilled workers regardless of gender. Projections from government agencies, educational institutions, and think tanks on the changing workforce are exposing the critical need to build interest within girls today. With only 25% of STEM degree holders being women; society is now wondering how to recruit and prepare the emerging, diverse labor force to fill the rising number of positions.

During the Wearable Technologies & E-Textiles: Mobilizing Girls into the World of STE(A)M presentation, Generation Infocus will provide methods to engage girls in STEAM to encourage communities to take part in training and mentoring young girls. Participants will be introduced to real world STEAM project based learning opportunities that are closely aligned with the Common Core and Next Generation Science Standards for girls in areas to include circuitry, fashion design technology, and 3D printing.

The purpose of the presentation is to raise awareness and advocate for policy expansion of female-centered STEAM programs in urban education. The intended outcome is to provide leverage for more diversity in the future workforce by equipping girls with the technical skills to become the next generation’s problem solvers, critical thinkers, and inventors in fields in which they are underrepresented. The research presented will be based on data collected by Generation Infocus’ 21st century grant funded learning program located within an urban school which serves 140 2nd-8th grade girls during after school. The targeted audience is school district leaders, school administrators, teachers, parents and anyone who wants to be a part of a movement to inspire and authentically engage girls in STEAM.

ABSTRACT. Science, technology, and innovation is a critical component of our efforts to tackle societal challenges we face today. Since the 1980s, when the Bayh-Dole Act was introduced in the United States, many industrialized countries have introduced policies to encourage innovation through technology transfer academia to the private sector, often with exclusive agreements on intellectual property rights. While some successful cases have been reported in transforming knowledge created by university researchers into industrial products, the existing models of university-industry-government collaboration tend to focus on narrowly-defined technical issues, mainly targeted to commercial applications. For tackling societal challenges, a new approach would be required to promote innovation, involving a wider variety of stakeholders with more diverse knowledge and expertise in scientific and technological fields.

Smart cities and communities are particularly considered to be one of the key areas in which a variety of science and technological knowledge need to be integrated effectively through collaboration among multiple actors. A smart city or community would involve an advanced technological system for efficient electricity supply and applications, incorporating all the behavior of the actors involved, including generators, distributors, technology developers, and consumers, through an intelligent information network. As a smart city integrates a diverse mixture of hardware as well as software in a complex way, different approaches would be possible to introducing and implementing the concept of smart cities and communities in practice, depending on the economic, social, and environmental conditions and purposes, such as energy efficiency, operating cost, environmental impact, resilience to external shocks and disturbances, and accessibility and inclusiveness to end users.

Recently, leading research universities around the world have started to apply their expertise and sources of innovation to the goal of building smart cities and communities. In this paper, we examine approximately 80 cases of the leading initiatives led by universities through stakeholder platforms in Japan, Europe, and the United States. A particular focus is placed on smart cities and communities, involving relevant stakeholders in academia, industry, and the public sector. A close examination of the leading initiatives reveals the important functions of university-driven stakeholder platforms in implementing innovation to address societal challenges. These include the creation of future visions based on science, setting of concrete and practical goals and targets, joint scenario making with stakeholders, securing active participation and serious engagement of stakeholders, collection and analysis of data on societal needs and demands, development of new technologies and systems through social experimentation at universities as living laboratories, assessment of impacts with transparency, objectivity, neutrality, legitimation of innovation in society, provision of effective feedback to decision makers, incorporation into institutional design, and contribution to agenda setting at regional, national, and global levels.

On the other hand, difference are also found with regard to the direction and process of technological development on smart cities and communities between Japan, Europe, and the United States. The Japanese approach is characterized by a strong focus on sophistication of application technologies for extensive use of home appliances and electric vehicles. In Europe an emphasis is placed on establishing a basic infrastructure in which information about the behavior of all the stakeholders is collected and distributed among the stakeholders appropriately so that the various objectives of the electricity grid are achieved in a more equitable way. In the United States a strong interest can be observed in creating and maintaining security through improvement in resilience against physical as well as virtual threats. These asymmetries in conceptualizing and implementing smart cities reflect the differences in how knowledge development, stakeholder networks, and institutional environment interact in dynamic and systemic manners.

The cases highlighted in this paper provide valuable insights into potential ways forward for collaboratively designing and creating knowledge and implementing innovation to tackle societal challenges. For follow-up efforts and new projects in the future, however, we still need deal with remaining challenges. These include how to navigate differing motivations and incentives for serious engagement and fruitful collaboration among stakeholders, to promote joint initiatives and networking that contribute to achieving desirable goals and targets and developing complementary skills and capacities, and to identify the factors and conditions required to promote their successful implementation.

ABSTRACT. Firms that develop from other organizations, often called spin-offs or spin-outs, are a mainstay in industry creation, market growth, innovation, employment and economic development (Callan, 2001). Spin-offs are created when intellectual property or knowledge is transferred from one source to form a new organization (Clarysse & Moray 2004; Parhankangas & Arenius 2003). For example, much of the early semiconductor industry in Silicon Valley was created through spin-offs from Fairchild Semiconductor (Saxenian 1994). As such, spin-offs are important to economic growth and the development, diffusion, and commercialization of nascent technologies. Parent organizations play a large role in the founding conditions, resource stocks, and strategies of their progeny spin-offs. It follows that a firm’s lineage is important for its performance (Burton et al 2002; Shrader & Simon 1997). Indeed, the resources capabilities provided to spin-offs at founding from their parents influence the spin-offs’ ability to survive (Powers & McDougall 2005; Klepper 2002). But not all parental organizations are the same; a wide range of organizations create spin-off firms. Spin-offs can be created when IP or knowledge is moved from one organization and used as the basis of a start-up (Clarysse & Moray 2004; Parhankangas & Arenius 2003). A spin-off depends on the IP or knowledge on which it is founded, the organization from which the IP is transferred, and the team transferring the IP and starting the new firm. Each provides opportunities for heterogeneity in the resulting spin-offs. Parent organizations vary from private firms to public institutions. Founding teams can include a range of people including professors, serial entrepreneurs, and researchers. However, much of the current research tends to either examine spin-offs as a relatively uniform group or focuses on either academic or corporate spin-offs. Only recently has literature emerged that compares the attributes and performance of different types of spin-offs. This stream of research has focused on the knowledge-based view that specifies that the foundation of knowledge imprinted on spin-offs by their parent organizations at the beginning influences their outcomes (Lockett et al 2005). For example, Zahra et al (2007) find that academic and corporate spin-offs differ in knowledge conversion capabilities, which leads to differences in productivity, profitability and revenue growth. Wennberg et al (2011) argue that the commercial knowledge inherited by corporate spin-offs results in higher growth and survival rates when compared to academic spin-offs lacking such experience. Similarly, Clarysse et al (2011) argue that since academic spin-offs lack market knowledge, they may benefit from having a broad technology with many applications, while corporate spin-offs use their superior industry knowledge to benefit from narrow technologies targeting specific markets. This work has provided insight into the knowledge-based differences and similarities of academic and corporate spin-offs; however, many questions remain. Yet, there is a dearth of research on spin-offs from other types of organizations. In 2014, the U.S. federally funded 25 R&D centers, 10 systems analysis centers, and six systems engineering centers. Spin-offs from these include Amtech from Los Alamos, MEMX from Sandia, and Sion Power from Brookhaven. Between 1997 and 2008, Los Alamos enabled over 50 spin-offs (Engardio 2008). However, government spin-offs have been treated as the same as other spin-offs or generally neglected in the literature. Given their active role in entrepreneurship and technology transfer, government spin-offs warrant additional consideration. This study examines the characteristics and outcomes of de novo firms and three types of spin-offs: those from universities, firms, and the government. The article begins with a review of work on academic and corporate spin-offs. The insights from these literature streams are integrated with findings from the limited work on government spin-offs. Then, I empirically examine how a firm’s origins influence its outcomes using a database of nanotechnology firms founded between 1981 and 2001. The data are analyzed using five types of firm outcomes: closure, acquisition, liquidation/bankruptcy, venture capital and government grants. The findings indicate that academic spin-offs are more likely to obtain venture capital and government grants while government spin-offs are more likely to be acquired. Start-up firms with a high proportion of former business executives on the founding team are more likely to close, while university spin-offs are more likely to obtain venture capital funding. The findings suggest that, indeed, the characteristics of a firm’s lineage influence its success, however in distinct and provocative ways that are not well explained by current theory.

ABSTRACT. The paper presents status and major changes in relationships between science and business that have occurred due to introduction of new government policy measures. These measures are aimed at improvement of knowledge flows between universities, research institutes, and industry. In Russia government is in fact mediator between science and industry. Total expenditures on R&D as percentage of GDP are at the level of 1.12% - lower than in most technologically developed countries. At the same time, the share of government in the total expenditures on R&D is unprecedented high – about 70% and there was steady growth during the last decade (OECD 2014). This is higher than in other BRIC countries, to say nothing about industrially developed countries. Government participation is growing not only in financial terms – regulatory function is also widening. In the paper, two major policy instruments implemented in the last 5 years in Russia are analyzed: matching grants for industry to cooperate with universities, and technology platforms as a communication instrument between science and industry/ Since 2006 the Russian government assigned to leading universities three major functions: to be centers of fundamental research, to have strong linkages with industry and thus produce highly demanded applied research and developments, and to become internationally visible. In order to strengthen cooperation in R&D between universities and industry the Russian government has introduced an instrument of matching grants. In this instrument government money are provided to industry but financing is earmarked: it should be transferred to universities that then develop new technologies for industry. The results of evaluation of this measure reveal a number of positive side effects (Dezhina, Simachev 2013), such as: • Improved universities’ awareness of industrial/business needs • Widening of contacts – creation of consortia (consisting from several universities around one company) • Stabilization of linkages (proved by continued partnership in other projects) • Involvement of companies in development of new lecture courses at universities. Since 2013, budget allocations for applied research started to grow. In the conditions when severe budget cuts due to economic crisis are expected, orientation towards applied research and closer linkages with industry is well justified. However, in a long run underestimation of the value of fundamental research is dangerous because the basis for further genuine technological breakthroughs may be lost. Finally, paper presents analysis of the new and not typical for Russia instrument aimed at establishing linkages among all actors of innovation system – technology platforms. Initiative was announced in 2010, and its concept was adopted from the EU experience. Government called platforms “communicative instrument aimed to activate creation of new technologies and products due to synergy of business, science, government, and civil society”, as stated at the “Strategy for innovation development till the year 2020”. The results of case studies are presented, showing the pace of developments within technology platforms. To the date, by judgment of the Russian Ministry of education and science, there are not more than 7 cases of success among 35 platforms that were established during 2010-2014. In predominant number of platforms, governmental research institutes and universities play a major role and industry is not active. Most of platforms have weak international linkages. At the same time, the successful are platforms where large and medium-sized companies take an active part. In these platforms a number of R&D projects were initiated that have led to development of new technological products. Then, platforms recently started to play an important role of “collective expert” helping to define and evaluate perspective R&D projects (Gutaruk 2014). Therefore in terms of influence on knowledge flows platforms may become a key instrument. In conclusion, positive and negative developments in government actions are summarized and the view on prospects for further development is suggested. Most likely government will widen approach according to which support of selected universities will be a priority, due to decrease in federal funding. Substantially better flows between science and industry will not occur in the near future; at the same time, the intent of the government to support import substitution may become an impulse for companies to invest more in R&D. References:

Dezhina I., Simachev Yu. (2013) Matching Grants for Stimulating Partnerships between Companies and Universities in Innovation Area: Initial Effects in Russia // Journal of New Economic Association, №3. P.99-122 (in Russian) Gutaruk E. Technology platforms as technology and platform for international cooperation. May 28, 2014 (in Russian) Main Science and Technology Indicators (2014). Paris: OECD.

ABSTRACT. Digital technologies such as computers, smartphones, and wireless networks are becoming ubiquitous in many parts of the world. We use digital technologies to innovate the ways we work, live, and play. These technologies are enabling or embodying new products, processes, business models, and industrial infrastructure, leading to significant social and economic benefits while posing serious challenges. These digital innovations, when successfully developed and deployed, are crucial to organizational performance, economic prosperity, and social well-being. However, success is anything but guaranteed when innovating with digital technologies. Innovations that are not so successful or have failed are rarely noticed, yet they waste substantial financial and human resources. Therefore, making policies on digital technologies, be it adopting big data or supporting net neutrality, is often a dicey challenge. At the core of this challenge are the widely differentiated outcomes of digital innovations, which beg a fundamental question: why do some digital innovations succeed while others do not?

To address this question, research has been conducted largely within disciplinary boundaries. Within each discipline, research has been so focused on one aspect of innovation (e.g., adoption) that other aspects (e.g., development) are assumed unproblematic. This division of labor has been increasingly challenged within each discipline and a consensus has emerged from the disciplinary boundaries that, for an innovation to succeed, both development and adoption matter.

Recent research on innovation ecosystems is promising in developing a much-needed holistic perspective to bring innovation developers, adopters, others, and their activities together to explain and shape innovation outcomes. Innovation ecosystems have been conceptualized as aggregations of interdependent actors and activities supporting innovations at organizational, industry, and national levels. Yet, beyond basic concepts, innovation research has thus far used the term "ecosystem" just a metaphor.

In this study, we take the "eco" in innovation ecosystems seriously, by applying organizational ecology theory and its modeling techniques to explain the evolving outcome of digital innovation. We conceptualize that the outcome of a digital innovation relies on a community, consisting of organizations playing different roles (e.g., technology provider, adopter, consultant, and researcher) and sharing interdependent ties (e.g., adoption, collaboration, and competition). Organizations are free to enter or leave any innovation community. Thus entries help the community grow; whereas exits hinder growth. Both entry and exit depend on two ecological processes: symbiosis means that the coexistence of organizations in a community signals legitimacy, which attracts new entrants (Hypothesis 1); competition for resources among community members deters new entrants (Hypothesis 2).

To illustrate this ecological perspective, we empirically studied Customer Relationship Management (CRM), a class of enterprise software that enables innovations in marketing and customer services, and used the ecology of CRM to explain the dynamic evolution of the CRM innovation community. Specifically, we collected and analyzed 200 articles on CRM published in the trade magazine Computerworld during 1998-2007. In these articles, we identified organizations, their relationships, and the roles played in the CRM community. We then focused on the organizations playing two types of roles (adopters and technology providers) as they are the most populous groups in the community. Employing a classic ecological modeling technique (density-dependence modeling), we formulated the rate at which organizations entered the community to play each role as a function of legitimation and competition already in the community. We found that, on the one hand, legitimation of CRM innovations indeed attracted new entrants to offer or adopt CRM; and on the other hand, competition for resources deterred entrants. Further, the legitimation among CRM adopters seemed to correspond to the competition among the providers of CRM, a surprising result that bridges the demand and supply of the CRM innovation.

These findings suggest: (1) ecological processes can explain substantial variance in innovation outcomes and thus ecology theory is a valuable addition to the repertoire of theoretical frameworks to guide innovation policy; (2) ecological modeling techniques provide a much sounder base than metaphors for making policies on innovations, especially digital innovations that involve novel technological infrastructure that often breaks traditional industry boundaries; and (3) policy-makers need to monitor actors and their activities closely in innovation communities so as to optimize symbiosis and competition for innovation success.

ABSTRACT. Trade in counterfeit products accounts for some 5-7% of global trade, amounting to more than $600 billion annually, and presents an important challenge to producers of legitimate goods, especially in developing countries with poor intellectual property rights protection. Yet, due to a lack of relevant data, there currently exists only scant empirical evidence on how counterfeit producers interact with legitimate producers in the market.

Using a novel dataset of medicinal sales in India between 2007 and 2013 and trademark litigation during the same period, this paper addresses this gap in the literature by examining how strategic market responses of heterogeneous producers (authentic producers, counterfeiters, potential counterfeiters, other firms present in the market) evolve when counterfeiting is discovered and acted upon by a legitimate producer.

Empirically, we disentangle the effect of counterfeiters on heterogeneous producers during every step of pre- and post- legal adjudication process. These periods include: (a) period during which either the authentic producers were not aware of any counterfeit sales; (b) period during which the authentic producer undertake private negotiations with the counterfeiters to curb their market activity; (c) period during which the authentic producer undertakes legal procedures to counter the counterfeiters, and (d) period after the judgement in favor or against the counterfeiters. Using a number of identified trademark infringement cases spanning a variety of pharmaceutical products sold in India in the period 2007-13, and adopting a difference-in-difference regression approach, our analysis provides insights into strategic behavior regarding pricing, sales, and advertising strategies adopted by heterogeneous producers during major steps of an evolution of trademark infringement cases against counterfeiters.

Contrary to popular beliefs about the overall enforcement weaknesses in emerging economies, we find India’s judiciary ability to curb counterfeiting to be more effective than private enforcement strategies adopted by the legitimate producers of pharmaceutical products. In addition, we find that different types of producers develop complex strategic responses that change as a trademark case progresses. Specifically, we find the infringers, in order to maximize profits in anticipation of court proceedings that could curb their activities, to flood the market with their products using a variety of strategies to increase sales. Additionally, a court’s verdict (interim injunction) against them forces them to reduce their prices and withdraw their product from the market. On the other hand, a court’s verdict in favor of them results in increased prices and sales of the infringing product. Producers of legitimate products, however, reduce their prices to make their product attractive to their consumers during the period before court proceedings. During the duration of the court case, however, they increase their prices to signal their product’s quality. Our results suggest a need for further study to better understand the complexities of strategic behavior of firms in the presence of counterfeiters.

ABSTRACT. Innovation is the main engine of a sustained economic growth that leads to a vibrant economy. However, innovation depends significantly on overall conditions in the economy, governance, education, and infrastructure. In such framework conditions, governments have traditionally played an important role in promoting innovation through various incentives, policies and laws. Having the policies aligned is a crucial matter in order to have an organic growth within the economy. Therefore, the challenge remains on how to have an assessment of the efficiency of the policies which transform the country to be more innovative. In this regard various approaches had been taken for measuring innovation and therefor indices had long been constructed known as Science, Technology Innovation (STI) indicators which is dominantly used. In the matter of measurement practices, efforts in the European Union have been made to measure country innovation capabilities. For instance, these assessments have been done through objective economic measures, such as Oslo manual series, European Community Innovation survey (CIS) and the European Innovation Scoreboard (EIS).

Meanwhile, it is increasingly evident that innovation is a social-economical phenomenon and it’s a cause of Top-down and Bottom-up interaction in a society. Our research concern is how to capture the efficiency of the government efforts throughout innovation policies into the innovation aspects which can be seen in society at large. This has been well illustrated in the Global Entrepreneurship Monitor (GEM) index, which captures an entrepreneurial activity across countries and is the best representative of its kind so far. Entrepreneurial community as such is playing a key role to stimulate economies, solve pressing problems and create fulfilling employment opportunities. According to the recent studies, the venture capital backed companies have a positive impact on the U.S. economy far outweighs its relative size. Also a venture backed jobs contains 11% of U.S. private sector employment (Heesen 2011). Hence the challenge is how these attempts for promoting innovation, mainly by governments had been transformed into more vibrant entrepreneurial culture.

Our approach is to look at the perspective in the form of an input and output where inputs translated to be the governmental efforts for promoting innovation and the output is measured by the activity of the entrepreneurial community. Therefore the variables which can interpret the transformation from input to output has been carefully selected. The study is limited to the European countries scale. Input variables: indicators illustrating government efforts in fostering and flourishing innovation from sources such as Global Innovation Index and European Innovation Scoreboard.

Output variables: the attempt is to capture the entrepreneurial activity in the countries, therefore new data sources have been selected and variables had been constructed. With this vision we looked at CrunchBase and AngelList as a service which profiles new businesses, startups, their investment rounds and job vacancies on a global scale.

Methodology: in order to assess the productivity of input efforts to be transformed into output, we utilized a non-parametric methods dominant in operation research and economies known as Data Envelopment Analysis (DEA). It is used to empirically measure productive efficiency, which first was proposed by chance, Cooper and Rhodes (1978). DEA method has been used extensively by scholars in course of National Innovation System (NIS) performance measurement (Abbasi et al. 2011; Pan et al. 2010; Cai 2011; Chen et al. 2011; Guan & Chen 2011; Hsu 2011).

Expected results: The study gives an overview of the activity level of entrepreneurship in European countries. Moreover an efficiency measure will be calculated considering governmental support for promoting innovation as an input with respect of entrepreneurial activity as an output. This indeed will generate a new scale and ranking accordingly, which emphasis on efficiency rather than proficiency of innovation policies.

ABSTRACT. Science policy is shifting towards an increasing emphasis in societal problems or grand challenges. Previous experiences of mission-oriented research, such as Nixon's war on cancer, suggest that while such mission-oriented emphases are useful in legitimising research, the extent to which mission-oriented policy succeeds in shifting research funding priorities and research practices towards the missions it claims to pursue is controversial (Sarewitz, 1996). The controversies are due to both the ambiguity in framing of the societal problems and their socio-technical solutions, as well as the uncertainty about the research options that can deliver those solutions (Stirling and Scoones, 2009; Wallace and Rafols, 2014). Tackling complex global problems or grand challenges – such as climate change, food security, poverty reduction, the risk of global pandemics – requires not only to increase R&D expenditure, but also to exploration and eventually to coordinate of a variety of stakeholders with different areas of expertise, pursuing diverse research avenues and often with contrasting world views. Typically these challenges benefit from the understanding of the physical and biological phenomena underlying the problem (e.g. the virus and its genes), but also demand an understanding of the environmental and social contexts in which they occur, and the policy networks and instruments available in those contexts (Ely, Van Zwanenberg, & Stirling, 2014).

In this context, public debate is seen as necessary on democratic grounds, given that both problem framing and priority setting are value-ladden and politically charged (Stirling, 2014). This debate can be enriched by an understanding of the plurality in societal demand for a given mission such as fighting obesity and the various research options that may contribute tackle it (Sarewitz and Pielke, 2007). In this article, we present quantitative tools based on semantic analysis of scientific and policy documents that aim to facilitate understanding and foster debate about science supply, societal demand and the relationship between the two for two case studies: rice and obesity research.

Most conventional bibliometric studies have focused on the mapping of knowledge production side in terms of mapping research areas and measuring scientific influence, in a way that implicitly assumed an "internalist" view of science (see Rotolo et al., 2014, for a different view). In contrast, mission-orientation and grand challenges demand to investigate the articulation of societal needs. We present an exploratory investigation of rice research (a major staple crop) and the grand challenge of obesity (an emerging health problem with enormous social costs). We illustrate a potential approach, showing how scientific publication can be used to describe existing science supply by using semantic networks and clusters based on publications' abstract. In the case of rice, the diverse focus of research is analysed for various countries and related to different societal demands and contexts --e.g. rice exports in Thailand and crop yields in Indian. In the case of obesity, we also use question records in the French and EU parliaments as an instance of social demand and we compare between the publication maps with parliamentary maps to explore (mis)alignments between societal concerns and scientific outputs.

In summary the paper contributes to the discussion on the role of societal demand in science by illuminating existing research priorities for a given topic, and contrasting them with socio-economic drivers and discourse on needs for that issue.

References Rotolo, D., Rafols, I., Hopkins, M. M., & Leydesdorff, L. (2014). Scientometric mapping as a strategic intelligence tool for the governance of emerging technologies. Available at SSRN 2239835. http://papers.ssrn.com/sol3/papers.cfm?abstract_id=2239835 Sarewitz, D. (1996). Frontiers of Illusion: Science, Technology and the Politics of Progress. Philadelphia: Temple University Press. Sarewitz, D., & Pielke, R. a. (2007). The neglected heart of science policy: reconciling supply of and demand for science. Environmental Science & Policy, 10(1), 5–16. Stirling, A. (2008). “Opening Up” and “Closing Down”: Power, Participation, and Pluralism in the Social Appraisal of Technology. Science, Technology & Human Values, 33(2), 262–294. doi:10.1177/0162243907311265 Stirling, A. C., & Scoones, I. (2009). From Risk Assessment to Knowledge Mapping: Science, Precaution, and Participation in Disease Ecology. Ecology and Society, 14(2). Retrieved from http://www.ecologyandsociety.org/vol14/iss2/art14/ Wallace, M. L., & Rafols, I. (2014). Research portfolios in science policy: moving from financial returns to societal benefits. http://papers.ssrn.com/sol3/papers.cfm?abstract_id=2500396.

ABSTRACT. Realising the societal gains from publically funded health and medical research requires an approach towards developing a reflexive evaluation precedent for the consideration for the societal impact of research. Whereas a number of conceptual models and theories have been proposed to understand the process of impact realisation and, in turn, guide its evaluation by peer review panels, actual empirical investigations of the assessment of societal impact, where the results are linked to funding outcomes are rare. Problems associated with access to peer-review panel deliberations, and a lack of formal frameworks incorporating criteria of the societal impact of research, have made conducting this type of empirical research difficult. This research provides the first empirical study of societal impact evaluation, by exploring UK Research Excellence Framework (REF2014) evaluator values and opinions about the process of assessing societal impact. Two intertwined, semi structured interviews (one before the evaluation process began (pre-evaluation interview); and the other after the evaluation process was complete (post–evaluation interview)) were conducted with 62 REF2014 evaluators from the health, medical and biomedical fields. Themes expressed by evaluators during interviews concerning the societal impact and its method of assessment were compared before and after the evaluation process, in order to gain an understanding of the process of societal impact evaluation during peer review panel deliberations. In the absence of prior studies, we discuss research of panel based peer review processes with which to interpret our results. Evaluators described their inexperience when dealing with societal impact assessment, and this inexperience led to a number of values and ideas about the value of different societal impacts being weighed against each other during the assessment process. At times, these ideas and definitions about societal impact ran counter to the existing guidelines designed to aid the evaluation process. Specifically, we discuss the characteristics of two different modes of evaluation – the “quality-focused evaluation” and the “societal impact-focused”. These represent two extremes of considering how evaluators might assess impact. We represent these two modes as two poles on a conceptual evaluation scale. We show how evaluators had a wide range of views about impact, and that these views could be conceptually reflected in a range of different positions along this scale, somewhere between the two poles of evaluation. We describe the characteristics of these poles in detail, and also discuss the different beliefs evaluators had which could influence where they positioned themselves along the scale. These decisions, we argue, when considered together, form a dominant definition of societal impact that influences the direction of its evaluation by the panel. In addition, considering evaluators’ views about societal impact using the evaluation scale aids the discussion of the tensions faced by evaluators regarding the valuation and conceptualisation of societal impact for assessment.

ABSTRACT. The relative newness of ‘impact’ as a criterion for research assessment has meant that there is yet to be an empirical study examining the process of its evaluation. In 2014, the UK Research Excellence Framework (REF2014) was the world’s first mandatory national research assessment framework, linked to funding allocation, to dedicate a formal proportion of its peer review evaluation to considerations of societal impact. Labelled “impact”, the newness of this as a criterion of assessment means there is a need to understand more clearly how evaluative decisions are made about this measure during peer review. This is particularly true give the current contention about how societal impact should be defined, evidenced, demonstrated and measured. This paper is part of a broader study which is exploring the panel-based peer and end user review process for societal impact evaluation using the UK’s REF 2014, as a case study.

In particular, this paper explores the different perceptions REF2014 evaluators had regarding societal impact, preceding their evaluation of impact as part of REF2014, and their different characterisations of impact as a criterion of measurement. Data is drawn from 62 interviews with evaluators from the health-related Panel A and its subpanels, prior to the REF2014 exercise taking place. The results showed that a majority of evaluators perceived impact as an “outcome” of research, though these views varied with regards to how much value they placed on different “research outcomes” when it came to impact assessment. In contrast, rather than viewing societal impact as an outcome, a small minority of interviewees viewed impact as a process and placed value on research activities which promoted research outcomes rather than the outcomes themselves. We note the similarities between these views and the SIAMPI model of assessment.

We discuss the implications of our findings for future impact evaluation frameworks in terms of the complexities’ involved with evaluating impact formally. We also postulate a series of hypotheses about how evaluators’ different perceptions going into an impact assessment could potentially influence the evaluation of impact submissions. Using REF2014 as a case study, these hypotheses were tested in an additional series of interviews with REF2014 evaluators post-assessment.

ABSTRACT. With the interdisciplinarity of the technology and the explosion of information, it is more difficult to identify potential technology opportunities from large and varied volumes of scientific and technology information. Most traditional technology forecasting techniques are subject to limitations in that they usually depend on qualitative approaches by intuition or simple processes such as extrapolation of past data. Several emerging tools—e.g., patent map, bibliometric analysis and text mining—have been proposed to produce substantial output based on empirical evidence like patents or science literatures, however they cannot reflect key technological findings or structural relationships among components. Meanwhile, conventional morphology analysis (MA), which has been utilized to identify technology opportunities, is subject to limitations in that there is no scientific or systematic way in establishing the morphology of technology. Subject-Action-Object (SAO) semantic analysis, as a fact-oriented modeling, can explicitly describes the structural relationships among technological components and fully reflects the specific key-findings. In this paper, we propose an SAO-based MA method that is supported by a systematic procedure and quantitative data for concluding the morphology of technology. By listing the occupied configurations of collected patents, the unoccupied territory of configurations are suggested as technology opportunities. Moreover, all possible alternatives can be identified, and the potential partner(s) for R&D collaboration can be examined by SAO semantic structure map. A Dye-Sensitized Solar Cells(DSSCs) case is exemplified to illustrate the detailed procedure of proposed method. The overall process involves the following five steps: (1) Selection of patent documents The main objective of this step is to select patent documents from a patent database to make the result containing more relevant patents and less irrelevant ones. To collect patents from the Derwent Innovation Index database in WOS, a multi-step Boolean retrieval query, which consists of keywords related to the target technology and International Patent Code, is used. (2) Identification of key technical problems and key technology components The main objective of this step is to build the framework for technological morphology matrix consists of key technology components (dimensions) and key technical problems (shapes). First, we identify the key technical problems by tech mining and comprehensive document surveys. Second, we identify the key technology components based on term clumping analysis and principal components analysis by use of VantagePoint software. We build the framework for technological morphology matrix by taking the key technology components as dimensions and the key technical problems as shapes. (3) Extraction and optimization of SAO semantic structures An SAO semantic structure can be extracted from any description in textual format including the title, abstract, claims and so on. In general, titles and abstracts are precise and have been regarded as the most meaningful parts. To extract SAO semantic structures, the abstracts are processed with Goldfire software from Invention Machine. The optimization of SAO semantic structures is to eliminate noise SAO models and to get the valuable SAO semantic structures. This process entails three tasks, 1) to filter unnecessary action words related to the key technology composition, 2) remove noise SAO models that were extracted due to the limitation of NLP technology, 3) to normalize SAO models by defining the academic vocabulary/terms, acronym or domain ontology for the particular field. (4) Generation of technological morphology matrix and technology opportunity analysis Based on the result of PCA analysis and the experts’ opinion, we classify the SAO semantic structures extracted into key technology components or key technical problems, and list the SAO semantic structures into the framework for technological morphology matrix, so we can form technology morphological matrix. SAO-based morphology analysis produces a number of undeveloped morphological combinations, which will help to identify technology opportunities. (5) Formulation of a SAO semantic structure map We propose a coordinate figure to illustrate the structures in a key technical problem, entitled the SAO semantic structure map with a horizontal axis referring to time and a vertical axis referring to key technology components. From the proposed SAO semantic structure map, we can identify the innovation pathways, key players (Inventors) and key organizations (Assignees) of the specific key-findings denoted by SAO semantic structures. It is useful to identify the potential partner(s) for R&D collaboration in order to exploit technology opportunities. Finally, the proposed method will be explored and applied in DSSCs sector to verify the validity and applicability.

ABSTRACT. Interdisciplinary research is increasingly recognized as the solution to today’s challenging scientific and societal problems, but previous studies found inconsistent results in terms of the relationship between interdisciplinary research and scientific impact.  One possible explanation is that they adopted different measures for interdisciplinarity.  Wang, Thijs and Glänzel (2015) used factor analysis to uncover distinct aspects of interdisciplinarity and investigated their independent dynamics with scientific impact.  They uncovered three factors corresponding to variety, balance, and disparity and found that long-term (13-year) citations (1) increase at an increasing rate with variety, (2) decrease with balance, and (3) increase at a decreasing rate with disparity.  Furthermore, interdisciplinarity also affects the process of citation accumulation: (1) although variety and disparity have positive effects on long-term citations, they have negative effects on short-term (3-year) citations, and (2) although balance has a negative effect on long-term citations, its negative effect is insignificant in the short run. 

Building on Wang, Thijs and Glänzel (2015), this paper makes two contributions.  First, instead of looking at the volume of citations, we examine the disciplinary profile of impact.  Using all 2001 Web of Science research articles, we extract variety, balance, and disparity factors for (1) on the one hand the degree of interdisciplinarity based on references (i.e., how many disciplines are sourced for the research, how equal are the roles played by these sourced disciplines, and how distant are these resourced disciplines) and (2) on the other hand the disciplinary profile of impact based on forward citations (i.e., how many disciplines are influenced by the focal research, how balanced is the impact across these influenced disciplines, and how distant are these influenced disciplines).  Subsequently we study the relationship between (1) different dimensions of interdisciplinarity and (2) different aspects of impact.

Second, in addition to scientific impact, we also examine the technological impact as embodied in patent citations to the scientific literature.  While citations from scientific literature are widely used to measure scientific impact, patent citations are widely used to indicate technological impact and economic values.  Therefore, patent citations allow us to examine the technological impact and economic values of interdisciplinary research.  We extract the number of patent citations to our sampled papers in the EPO PATSTAT database and study the relationship between interdisciplinarity and patent citations.

ABSTRACT. In the past decades, comparisons between universities, may it be in the course of national evaluation exercises or in so called university rankings, have become abundant. One of the problems of such comparisons and their interpretation is the lack of knowledge concerning the significance of differences between two scores or ranks. Recent developments in the field have tried to amend this by introducing stability intervals through bootstrapping of the actual publication set of a university, as e.g. for the Leiden Ranking. However, this method is only sensitive to the size of the publication set and the effect of outliers. In this paper we introduce a way of transforming a different source of uncertainty concerning impact indicator values into a stability interval that quantifies the coarseness of bibliometric evaluation: The internal heterogeneity of disciplines subsumed as Subject Categories (SC) in the Web of Science database (WoS), which are the foundation of classical field-normalized citation indicators such as the Mean Normalized Citation Score (MNCS). Field normalization has been developed in order to neutralize the obvious diversity of publication and citation practices between fields and subfields, as a corrective to otherwise unfair comparisons between the citation impact results of corpora with varying subject distributions. Various methods for field delineation have been proposed including many proposals for clustering methods and arguments to determine the correct levels of aggregation. So far, however, no classification system other than those provided by the database vendors could be established as standard throughout the bibliometrics community. However, it is evident that the SCs diverge in size and specificity. Van Eck et al. (2013) provide furthermore strong evidence of heterogeneity within the medical SCs along the characteristics of clinical and experimental research: The substructures displayed with the help of term clouds differ significantly in their citation rates along these dimensions. The authors draw the conclusion that the impact of clinical research is systematically underestimated by normalized citation indicators. Solving this problem by introducing finer classification systems is not straight forward, as only further normalization options are created, while the legitimacy of these new classification schemes is itself questionable.
We introduce a way to quantify such heterogeneity within the SCs. The approach thus combines two advantages: on the one hand, we continue to work at the level of a standard classification system and on the other hand, underlying structures on a secondary level are made transparent. In this proof of concept paper we use keyword terms of all articles and reviews published in journals of several medical subject categories of one publication year (2008) in WoS. A stemming procedure is carried out which neutralizes different inflexions. These keyword pairs are used for a coupling procedure of the publications; Salton’s Cosine is used to neutralize differing amounts of keywords. Ward’s method is used to cluster the SC’s substructures and the clusters’ Mean Citation Score (MCS(cluster)) are calculated. Sets of publications affiliated with several German universities were selected.On the basis of the resulting cluster distributions, a bootstrapping approach is utilized:For each institutional publication corpus, a random set of MCS(cluster)s equal to the institutional corpus’ size is drawn with replacement from the clusters’ MCSs with the probabilities equal to the clusters share in the SC. The arithmetic mean of this combination is calculated and serves as the Expected Citation Score (ECS(i)). Each raw citation score is then divided by the ECS(i) and the arithmetic mean of the results deliver the MNCS(i). 10’000 iterations of this procedure are executed. Finally, the 2.5% and 97.5% quantiles of this distribution express the stability of the university’s MNCS. Preliminary results show that clusters can be delineated and differing topical foci can be identified. While the medical SCs scrutinized display clusters in clinical and experimental research and the MCS(cluster) differ by up to a 100%, thus corroborating Van Eck et al.’s results, other facets of heterogeneity also emerge. The stability intervals show ranges smaller but comparable to the bootstrapping method utilized for the Leiden Ranking.
These two stability calculation methods could be combined and account for both the robustness of the citation scores given the size and distribution of the publication sets, as well as the underlying uncertainty of the expected citation rates. We believe that such methods that display the imprecision of bibliometric point estimates, which especially clients of evaluative bibliometric analyses are prone to disregard and thus revel or despair at minute changes of their scores and ranks, are an important step to the correct interpretation of bibliometric indicators.

ABSTRACT. Faster technological progress has long been considered a key potential benefit of agglomeration. Physical proximity to others may help inventors adopt new ideas in their work by increasing awareness about which new ideas exist and by enhancing understanding of the properties and usefulness of new ideas through a vigorous debate on the ideas' merits (Marshall, 1920). We test a key empirical prediction of this theory: that inventions in large cities build on newer ideas than inventions in smaller cities. We analyze the idea inputs of nearly every US patent granted during 1836--2010. We find that a larger city size provided a considerable advantage in inventive activities during most of the 20th century but that in recent decades this advantage has eroded.

Full paper available at: www.nber.org/papers/w20920

ABSTRACT. This paper tests empirically the different factors associated with tie formation and tie persistence of network co-authorships using a longitudinal dataset of the National Research Foundation (NRF) of South Africa. We highlight three factors: status homophily and cognitive proximity (1), organisational proximity (2), and social and community structure (3). We find that tie formation is favoured for researchers that are well connected in the co-authorship network, have high cognitive and organisational proximity, and are part of the same social communities. In contrast, our results show that nothing matters apart from cumulative advantage and social structure to make a tie persist. 

Publications are increasingly important, if not the most important, measure of research productivity. Understanding the determinants of collaboration and research outputs at individual and collective levels can help in the design of research policies focusing on how to boost the quality and quantity of publication outputs. This is an issue of particular concern in Middle Income Countries (MICs) as their science systems are typified by scarcity of resources for research. Social network analysis is a tool that can contribute theoretically and empirically to the understanding of these determinants. 

Moreover, in recent years there has been an increase in interest in the economics of science. Most of the work, however, is done in the context of rich countries, with well-developed science systems. Very little work is directly concerned with these issues in middle-income (much less poor) countries. But the constraints under which science systems operate in these countries, and the important role they might play in technological upgrading, and thus in economic growth in general, implies that there is an important lacuna in our understanding. This paper aims at this gap. 

We use data on rated researchers from the National Research Foundation (NRF) of South Africa. We analyse individual NRF-rated scholars on the basis of their peer-reviewed publications and the networks of collaborators evident from those publications, as well as their NRF ratings and scientific domains at moment of application. We use a longitudinal dataset to analyse the formation and persistence of ties in the period 2000-2011 for a total of 3,532 researchers. 

We analyse two sets of econometric models, one estimating tie formation; and a second one for tie persistence. We use a probit model to estimate the probability of a first collaboration, and for those ties that were formed; we use a Cox proportional hazard regression model in order to estimate time to consequent collaborations.

Our paper gives an important place to the effects of social structure on scientific collaborations. In particular, we are interested in whether or not membership in a community of researchers has an impact on tie formation and persistence. We adopt some of the relatively new techniques for community detection in networks generally to investigate the meso-level effects of social structure on scientific collaboration. 

We find that tie formation is favoured by researchers being well-connected in the co-authorship network; for most disciplines, the plurality of collaborations are within discipline. Status homophily is not a significant predictor of tie formation. In fact, our results suggest the importance of heterophily, hinting that researchers look for complementary skills and knowledge when collaborating. Cognitive proximity is an important factor for tie formation, but not for tie persistence, whereas difference in status is positive for tie formation, but is not significant for tie persistence. Being part of the same social community is however significant for both tie formation and tie persistence.

What the results suggest is that tie formation and tie persistence are driven by very different types of forces, and that once a good fit between partners is found a tie will naturally persist, since most of the explanatory variables do not change with time. Interestingly, the crude interpretation of these results is that nothing matters apart from cumulative advantage and social structure to make a tie persist.

ABSTRACT. National initiatives to foster innovation have become more complex over the last few decades, reflecting our more advanced (yet still far from complete) understanding of how innovation works. Despite this surging complexity in policy, our evaluation tools have not followed suit and frequently lack a way of accounting for their complex goals and global linkages. In this paper, we focus on understanding the impact of one such complex policy instrument combining network analysis with a systems architecture approach. The MIT-Portugal program (MPP) began in 2006 and is a partnership between MIT, eight Portuguese universities and various industry partners. This program facilitates research collaboration between researchers in Portugal and at MIT, and supports the education of Portuguese and international graduate students, many who spend an extended research period at MIT. In particular, we aim to answer the question: How do the goals and structures of MIT-Portugal Program dictate the collaboration patterns of participating researchers?

To evaluate the impact of MPP on researchers’ collaborations, we observe how individual and collective collaboration patterns change, how new collaborative ties are formed and strengthened, what impact these collaborative ties have on research focus and research output, and how this relates to the partnership goals and architecture. Here we analyze the network of academic journal article co-authorship between the over 300 faculty participating in MPP. Data was collected from the ISI Web of Science and Scopus databases on the publications of each faculty member to generate the network of co-authorship throughout time. We first used a difference-in-difference method to compare our results to a control group, which is as similar as possible to the MIT-Portugal cohort except for the participation within the program. The control group is selected from researchers from the same 5 universities and from three of the same departments as the treatment group. Further, researchers were selected so they have a similar distribution of number of pre-2006 number of publications and the same spread of junior and senior faculty. We then used propensity score matching and coarse exact matching to refine this control group and eliminate any selection bias.

We were able to assess MPP in terms of some of its more complex, less tangible major goals. These goals include 1) Encourage Portuguese Universities to work closer together build domestic critical mass; 2) Encourage Portuguese collaboration with MIT; and 3) Conduct high-impact research. In addition, we find a number of interesting effects on the impact of MPP on the formation of collaboration networks between participating researchers. MPP-affiliated researchers formed 40% more ties within Portugal than the control group, the majority of which were with an MPP-affiliated university, research center or private partner that was not their home institution. We see a disproportionate impact on the collaboration networks of Junior and Senior Faculty. We find that researchers can only maintain a relatively constant number of active research connections at one time and therefore, in order to form new collaborations with MIT faculty as a part of MPP, we see that participating researchers have to break existing collaborative ties. However, Junior Faculty can add these new collaborators from MIT without severing existing collaborative ties, and as such quickly build up a much larger collaborative network than their control-group counterparts. As such, we see a dramatic difference in the research output of junior faculty, where they produce over twice as many papers per year as similar researchers in the control group (as opposed to a modest 25% and 17% increase in output of mid career and senior faculty respectively). Finally, we examine the formation of future collaborations after an initial research connection is made, and find that MPP-participating researchers are more likely to be introduced to and integrated into a new co-authors network of previous collaborators than the control group.

The MIT Portugal Program case is particularly interesting because it’s part of a broader trend, where governments use these Complex International Science, Technology and Innovation Partnerships (CISTIPs) to build domestic capacity in science, technology and innovation through collaborations with globally recognized experts. Future work will focus on developing further cross case comparisons to CISTIPs that have a wide variation of goals and structures to better understand how such policy decisions can be used to shape the formation of collaborative networks in research. Such work will better help us understand how to design future CISTIPs and how to better manage existing ones, ensuring that investments in such innovation policies are put to effective use.

ABSTRACT. Several studies on research collaboration have indicated the importance of self-organization of research community in finding collaborators, exchanging ideas and organizing interactions. Literature on ‘invisible colleges’ suggests importance of self-organization in scientific communication and formation of research networks (Crane, 1972; De Solla Price & Beaver, 1966). Some studies of research collaboration associate self-organization with freedom to choose research collaborators and find it to be an important element of successful collaborations (Melin, 2000; Wagner & Leydesdorff, 2005). However, the process of ‘self-organization’ in science policy studies is still rather black-boxed and there is relatively little elaboration about how it works. Often the idea of self-organization in the existing collaboration literature is present but implicit.

To address this research gap, this paper will analyze evidence from seven long-term international research collaborations (Ulnicane, 2014) focusing on why and how, and under what conditions self-organization works. The paper aims to explicate the process of self-organization in formation and organization of research collaborations. Important focus is on interaction between self-organization of research collaborations and factors such as individual and institutional characteristics and epistemic properties. Questions addressed include factors facilitating and hampering self-organization, the role of trust, communication and leadership in self-organization, advantages and risks of self-organization as well as interaction of self-organization with attempts from research institutions and funding agencies to steer collaborations. It is particularly interesting to analyze self-organization in case of international collaborations, where researchers often have to overcome longer geographical distances.

To provide an in-depth understanding and rich insights, and to develop empirically relevant conceptualization of self-organization process, the paper draws on multiple research methods and data sources. Seven longitudinal case studies integrate evidence from 61 semi-structured interviews with collaborating researchers with data on their projects, publications, organizations and CVs. Collaborations studied involve researchers from five European countries (Germany, the Netherlands, Belgium, France and the United Kingdom) in and emerging field of nanosciences and technologies. Tracing and comparing the formation and development of collaborations lasting over 10 and 20 years enables a thorough examination of the role of self-organization process, its key features and its interaction with individual, institutional and epistemic factors.

References: Crane, D. (1972). Invisible Colleges. Diffusion of Knowledge in Scientific Communities. Chicago: The University of Chicago Press. De Solla Price, D. J., & Beaver, D. (1966). Collaboration in an Invisible College. American Psychologist, 21(11), 1011-1018. doi: 10.1037/h0024051 Melin, G. (2000). Pragmatism and self-organization - Research collaboration on the individual level. Research Policy, 29(1), 31-40. doi: Doi 10.1016/S0048-7333(99)00031-1 Ulnicane, I. (2014). Why do international research collaborations last? Virtuous circle of feedback loops, continuity and renewal. Science and Public Policy. doi: 10.1093/scipol/scu060. Published online October 13, 2014. Wagner, C. S., & Leydesdorff, L. (2005). Network structure, self-organization, and the growth of international collaboration in science. Research Policy, 34(10), 1608-1618. doi: DOI 10.1016/j.respol.2005.08.002

ABSTRACT. This paper contributes to the scholarship of science-policy linkages through a meta-synthesis of published case studies on boundary organizations and boundary spanning processes. This research is important as humanity faces increasingly intractable environmental problems characterized by high uncertainty, complexity, and swift change. How science is developed and applied to policy-making is one major factor influencing “humanity’s environmental future” (Caldwell 1990). Yet, science is currently underutilized in environmental policy despite the growing call for effective scientific engagement in public policy (NSF 2002). 
Scholarship from the natural sciences to humanities on science-policy linkages is dominated by boundary organization theory which argues boundary organizations (formal organizations which are designed to bridge science and policy) are the best way to bridge the science-policy divide (Guston 2001; O’Mahony and Bechky 2008). Classic examples of boundary organizations include the EPA, the Sea Grant Program, and the Intergovernmental Panel on Climate Change as outlined in Guston et al. (2000). Yet, boundary organization research lacks systematic measurements of key processes, variables, and outcomes while rarely engaging in theory building and cross-case comparisons (McNie 2007; Crona and Parker 2012). It also assumes, but fails to critically assess whether, boundary organizations are the most effective means to span science-policy boundaries and overly privileges the formal features of boundary organizations and boundary spanning processes as explanatory variables. This paper addresses these gaps through a meta-synthesis (systematic evaluation of published case studies) to evaluate the strength of the relationship between the formal features of boundary spanning processes and science-policy influence. 
      Over fifty environmental science-policy case studies were evaluated through the method of meta-synthesis. The five phases of meta-synthesis, as outlined by Sandelowski et al. (1997) and Timulak and Creaner (2013) were completed. Phase I determined the research question and corresponding hypothesis. Phase II created the research article inclusion and exclusion protocol and executed the research strategy. Phase III coded each article (n=51) and determined comparability between articles. Phase IV built upon the construction of the preliminary conceptual framework and Phase V provided analysis and results of the meta-synthesis.
Overall, 6,057 articles were screened on title and when possible, abstract. 5,779 articles were excluded from initial screening based on topic which left 278 full articles to be considered for review. The 278 articles were analyzed utilizing the inclusion criteria protocol which led to the exclusion of 227 articles from meta-synthesis analysis. 51 articles were left to compare the features, roles, strategies, and outcomes reported in each study. This analysis yielded 81 explanatory variables which may have potentially impacted the uptake of environmental science in policy in this sample size. However, to ensure that (1) there was not spurious significance; and (2) correlated variables were not needlessly analyzed, excess variables (n=68) were discarded from the calculations. To decipher what features, roles, and strategies of boundary organizations and spanning processes result in effective policy influence the strength of relationships between the 13 identified variables and science-policy outcome were evaluated using the Chi-square Test (and Fisher’s Exact Test when necessary).
Unexpectedly, some of the key features, roles, and strategies hailed in the literature as most important and in boundary organization theory did not significantly influence the uptake of environmental science in this sample size (e.g., the co-production of knowledge, providing accountability and responsibility, utilizing a boundary organization, and creation and use of a boundary object). These results should not dismiss, however, the scientifically observed importance of these explanatory variables in successfully integrating science and policy. In fact, a key attribute of this meta-synthesis should be the demonstration of the high-volume of variables working simultaneously to enhance the uptake of science in policy. Yet, when tested individually against a successful science-policy integration outcome, two explanatory variables were found to be significant in this study. When case studies reported utilizing a boundary spanner or a clear focus on creating salient, credible, and legitimate knowledge for a multitude of stakeholders, these variables most significantly influenced the uptake of environmental science in policy.

.

ABSTRACT. Introduction

The analysis of the policy-making process related to science, technology and innovation (STI) in developing countries is a quite demanding exercise in terms of the limitation in gathering data and literature. If additionally, the STI analysis is focused on a specific topic such as regional integration and collaborative research between countries, the difficulty increases significantly. This is the case of the analysis of the role of STI policies in the Central American integration process.

One of the key steps of such analysis in the CA context consists on framing the STI policy rationale in relation to the economic integration process. However, when the scope of the analysis includes the dimensions of economic openness and economic integration in a “multi-scaled” regional context (national level, regional level, supranational level), in countries with socioeconomic asymmetries and different degree of institutional development, few tools are left to determine which STI policy rationales would support certain policy design and the niche in which STI policy fits.

Methodology

This paper aims to explore the role of STI policies in the economic integration process of the relatively small and open economies of Central America and the Dominican Republic. In order to do so, this paper focuses on answering two research questions: 1) What role has the definition of STI policies played in the frame of the Central American integration process? 2) What lessons derive from the attempts to build a regional STI policy in Central America, in the frame of the Central American Integration System (SICA)? The first question highlights the different efforts carried out through time at the regional level, at the second question implies the analysis of the efforts deployed after the creation of the SICA.

The analysis has been made using two complementary methodological approaches: 1) interviews to key STI policy makers of the region; and 2) literature review including selected input and output indicators of STI. The input from key STO policy makers was obtained through the format of “structured interviews”, which consist on a set of predetermined questions asked in the same order within a verbal interchange, to explore the perceptions about science, technology and innovation policy models at the national and regional levels, understanding by ‘region’ the CTCAP´s influence área.

Findings

Including a regional STI policy as a key component of Central American integration process implies to explore other STI policy rationales to create a mix of policies and instruments that will allow to overcome the current regional STI policy inertia.

In regards to the first research question, it could be concluded that the Central American integration process doesn´t yet have a specific rationale that supports the design of STI policies for integration, and therefore, STI policies have played a marginal role in the overall setting of policies for integration. In fact, it is possible to find some specific STI policy rationale at the national level but not at the regional level. The absence of a regional STI policy rationale related to the Central American integration process, is historically determined by an STI institutional setting with a very narrow perspective about knowledge production, regardless of the personal characteristics of policy makers. Fortunately, there are some clues about a new dynamics of policy making on STI that would move the regional organizations towards an STI policy mix that will allow managing the regional variability, support organizational synergies and harness national STI capabilities: a partial and encouraging answer for the second research question.

Finally, the current regional STI mechanisms (CTCAP- Commission for the Scientific and Technological Development of Central America and Panama) and the main economic policy mechanism (SIECA- Central American Economic Integration Secretary) should increase their formal collaboration in order to build a regional and comprehensive STI policy frame to deal with the regional variability and specially to overcome the ´policy of things´. This would imply the definition of a more structural and comprehensive perspective on STI policy and development.

ABSTRACT. Introduction From its very outset a common science and technology policy was an integral part of the ‘European project’. The First Framework Programme for Research (FP) was launched in 1984. Cross border collaboration is required in most EU R&D initiatives. In 2000 the FP was complemented by the ‘European Research Area’ initiative (ERA). Its aim is to achieve an internal market, in which researchers, scientific knowledge and technology circulate freely. Strengthening the transnational research cooperation within the EU is one of the ERA objectives. To monitor cross border scientific collaborations bibliometric indicators based on co-publications have been developed. Although in the past the European Commission used these indicators to monitor the development of the EU R&D system, an analysis of the evolution of EU cross border publications is not included in the 2013 and 2014 ERA Progress Reports. Research questions, methodology and data It is well know that over the last decades the business of science has become more international and cross-border collaboration is on the rise especially for natural and life sciences and the basic disciplines of engineering It is against the background of this globalisation that the realisation of the ERA takes place. Given the different degree of economic and scientific development of the countries publishing papers in journals covered by the WoS and their geographic location comparing the evolution of the overall fraction of internationally co-authored papers processed for the WoS with the fraction of intra-EU publications yields little information. In this study the United States is used to benchmark the ERA. The US science and innovation system gradually developed over more than two centuries. The Federal Government does not use policy tools to strengthen research collaboration between the states of the Union. Moreover in this benchmarking exercise an administrative, not an economic or scientific partitioning of the country is used. The United States Census Bureau defines nine divisions grouping states and the District of Columbia by their geographic location (https://www.census.gov/geo/maps-data/maps/pdfs/reference/us_regdiv.pdf). The WoS contains the information of the addresses mentioned in the by-line of the publications; in the US addresses the state code is one of the elements. This information is used to assign the US publications to the divisions. For each year in the period 1980-2014 the symmetrical co-publications matrix of the US divisions is calculated with on the diagonal the number of publications of each US division and on the off-diagonal entries the number of co-publications between two divisions. To quantify the collaboration strengths between the US divisions the Salton Index (SI) is calculated. For the EU based on each country intra-European co-publications the values of SI are calculated for each year. In this abstract only results for the six EU founding members (EU06) are discussed. To compare the evolution of the collaboration strength between the US and the EU06, for each year the average, the median, the maximum and the minimum value of the SI are used. These indicators provide an overall picture. The same analysis is carried out at the level of research fields using the OECD Revised Field of Science and Technology (FOS) Classification. In this classification WoS subject categories are mapped on the scientific fields. Beside the total number of publications, the top 10% most cited publications are also used to compare the evolution of the collaborative strength between the USA and the EU06 overall and for different scientific fields. Results Over the whole period the average and median values of SI of the EU06 remains below those of the USA. For both geographic regions an increase can be observed. For the period after 1990 the value of the slope of the linear fit for the USA and the EU05 is 0.0028 and 0.0022 (Pearson correlation > 0.99). A multiple linear regression analysis shows that for the period 1990-2014 the slope of both curves differ significantly at 5% level. For the 10% most cited publications similar results are obtained. In a next step this analysis is made for the research fields. Except for chemistry for all disciplines the average and the median values of the SI of the EU06 are below these of the USA over the entire period. Only for Mathematics, for Earth and related Environmental Sciences and for Chemistry a ‘catching up’ of the EU06 is observed as the value of the slope of the linear fit is higher than for the USA. Only for Chemistry a very different picture emerges. In the beginning of the 80’s of last century both the average and the median values of the SI for the EU06 were less than half that of the USA. Over the next three decades the growth rate of both indicators was considerably higher for the EU06 than for the USA and at the end of the period in some years the average and median values are even higher.

ABSTRACT. This paper argues that Europe – as other parts of the world - is confronted by fundamental changes in its external environment as well as internally, giving rise to several daunting policy challenges, the solution of which requires a completely new policy stance.

First, there is the economic challenge manifest in slow growth or even stagnation in many developed countries. This is related to a number of factors. One is the globalization of the world economy and the rise of new economic powers in Asia and elsewhere, posing a threat to the competiveness of many national industries throughout the developed world while at the same time opening up new markets and opportunities. A second is the extensive and long-lasting impact of the 2007-8 financial crisis and the measures taken to counter its effects. In recent years, economic stagnation has been especially pronounced in Europe, where it has led amongst other things to increased differences between rich and poor countries and to very high levels of unemployment (especially among the young) in large parts of the continent. Thirdly, there are underlying and persistent structural problems in certain countries that need to be urgently addressed. Austerity programmes currently underway in a number of European countries are unlikely to provide a satisfactory solution to these problems. Rather, as we shall argue in this paper, what is required is a continuous upgrading of skills, technologies and industrial structures in the countries concerned, and long-term policies – backed up with substantial investment – to make all this possible.

Second, there is the challenge posed by the climate crisis, the solution of which requires nothing less than a fundamental transformation from carbon-based growth to a new, sustainable economy. This may be technologically feasible but is nevertheless extraordinarily difficult to bring about as it challenges fundamental beliefs regarding the economy and society and our very notion of ‘progress’. Yet without a successful transformation to a sustainable economy, future generations will be in dire straits. Like the economic challenge, the climate challenge points to the need for long-term policies supporting the development of skills, innovation and structural change in Europe.

The third challenge concerns the governance crisis now facing Europe and the difficulties this poses for policy making and implementation. These difficulties were already becoming apparent before 2008, but the financial crisis has made them much more severe. A European Union that is struggling with dwindling popular support for its institutions is finding it difficult to rally its citizens behind adequate responses to such long-term problems. That task has been made all the more difficult by the hegemonic influence of neo-liberal ideology over the last two decades, with its blind faith in the supremacy of ‘market solutions’ and its constant belittling of any efforts by ‘the state’.

The paper argues that a completely new policy stance is required, one that simultaneously addresses the challenges brought about by economic stagnation, by climate change, and by the governance crisis, as these are indeed closely inter-related. Without growth, greenhouse gas emissions may continue to decline, although almost certainly not at the pace necessary for their elimination by the end of the present century. But employment would undoubtedly continue to suffer. A return to the type of growth that prevailed before the 2007-08 financial crisis, on the other hand, might be good for employment but certainly not for the climate, as emissions would continue at unsustainably high levels. Is there a way out of this dilemma? It is argued that challenges, while often daunting, are also opportunities for doing things differently, and may call forth - and indeed even require – innovation and entrepreneurship at various levels of society, not only in private enterprises but also in the public sector and among policy-makers, for their successful resolution. What is required is a reconceptualization of what is intended by ‘economic development’ and a fundamental transformation of the economy to a new ‘green’ trajectory based on rapidly diminishing emission of greenhouse gases. This entails a much greater emphasis on innovation in all its forms (not just technological). Innovation policy must be placed at the very heart of industrial policy and indeed of economic policy more broadly. While this may well entail certain costs in the short term, it will undoubtedly bring considerable benefits in the longer term in the form of strengthened industrial competitiveness, enhanced exports and new jobs. It may also encourage countries in other parts of the world to follow suit, thereby ensuring that climate change is kept within manageable bounds.

ABSTRACT. Emerging technologies, especially when they raise significant combinations of societal, economic, environmental health and safety, and ethical implications, are today subject to increased deliberation, at times generating concerns about how (and if indeed whether) such technologies should be developed. During the past decade or so, “responsible research and innovation” (RRI) (alongside related concepts such as responsible innovation, responsible research and responsible development) has emerged as a strong theme in new initiatives to rationally govern emerging technologies. RRI has gained growing visibility in European public and policy discourses related to science, technology and innovation. RRI has been put forward in various ways, but all definitions have more or less contained considerations of reflexivity, anticipation, participation, responsiveness and inclusiveness. However, the connotation of RRI continues to be debated, with a range of frameworks and implementation strategies proposed by various stakeholders. Some are general, conceptual frameworks aiming to assist the governance of research and innovation with RRI taken into account. Others are technology-specific approaches, such as those delimitated for synthetic biology. Thus far, the available literature mostly focuses at high-levels, adopting normative and policy perspectives. Broad governance and regulatory frameworks have been a focus of discussion. Yet, what is less-well documented is how RRI is being implemented in practice, and how it is shaped by various factors including the incentives and interests of key participants and stakeholders.

This paper contributes to a better understanding of the real-time application of RRI by examining the case of newly-established network of synthetic biology research centres in the UK. Under the framework of a UK roadmap for synthetic biology, six multidisciplinary synthetic biology research centres were established in 2014 with sponsorship from the Biotechnology and Biological Sciences Research Council. A national synthetic biology innovation and knowledge centre is also supported by the Engineering and Physical Sciences Research Council. Thus far, the UK has committed about £200m ($295m) in public investment to generate a critical mass of synthetic biology research. The aim is to build a knowledge base to commercialise new applications of synthetic biology in a series of sectors including agriculture, chemicals, energy, and medicine. At the same time, there is recognition that synthetic biology raises concerns about the ethics and potential environmental, health and safety risks associated with the engineering of nature, as well as issues such as the ownership and control of synthetically-engineered organisms and the effects on existing sectors and workforces. Each of the new UK synthetic biology centres incorporates a responsible research and innovation component. Significantly, while overall frameworks are set out by the research sponsors and the UK synthetic biology roadmap, each centre has flexibility in how it interfaces RRI considerations with its research and commercialisation strategies.

The variations in RRI approaches adopted by the UK synthetic biology research centres not only reflect differences in scientific research targets and sector orientations but also in how each centre interprets the meaning and practice of RRI. In the still early evolution of RRI, this process of experimentation offers an important opportunity for exploration and reflection. The paper takes up this opportunity by examining how the UK synthetic biology research centres understand and translate broad RRI frameworks into specific practices and activities, how key participants and stakeholders shape these strategies, and what early implications can be drawn. We adopt a comparative case study research design, building on our own role and access as contributors to RRI processes within one of these research centres. The research involves (1) scoping the contexts of both synthetic biology and RRI, using bibliometric and social network analysis to gain an understanding of who is working with whom in the areas of synthetic biology and RRI in the UK; (2) identifying key policy and centre statements and documents related to research and commercialisation strategies and RRI; and (3) interviews and focus groups with researchers and practitioners associated with the UK synthetic biology research centres. Analysis of this information will allow us to better understand both the varieties and communities of practice of RRI in synthetic biology and to probe key factors and stakeholders driving the operationalization of RRI. Our assessment of these practices, and the interactions between normative framework mandates and centre- and researcher-level translations, will offer important early insights related to the governance of synthetic biology (and other emerging technologies) through a networked embedded centre-based approach.

ABSTRACT. The concept of Innovation for Development in science, technology and innovation (STI), development and policy discourse, is now a major focus in both developing and developed countries academic and government institutions, including the African Union, the OECD, the United Nations and the World Bank. The multiplicity of organisations has led to the generation of many definitions of the term inclusive innovation and the related concept of innovation as a tool for inclusive development.

In spite of the differences in current definitions, terminologies and frameworks, there is consensus that innovation for development can be a very important tool in making STI relevant for the wider society and communities (Marcelle 2012, 2014, Marcelle et al, 2013; UNCTAD, 2014) particularly to those outside formal institutional settings, thereby contributing to growth and socio-economic development (Kraemer‐Mbula and Wamae, 2010). In spite of huge technological advances and innovation, “increasing inequalities spread starvation, violence and despair around the globe” (Soares and Cassiolato, 2013, p1).

In this paper we make the case for a critical assessment of the current understanding and framing of innovation for development in order to provide greater conceptual clarity and facilitate the emergence of innovation ecosystems. We start with literature review and discussions on the choice of innovation for development terminologies, which we observe, reflects North-South dichotomies. We explore how innovation for development is defined, theorised and conceptualised in scholarly and policy literature, and referred to by terms which include: grassroots (Gupta et al., 2003; Gupta, 2012), frugal (Zeschky et al., 2011) and pro-poor (World Bank, 2010; Iizuka and SadreGhazi 2011). Others are, innovation for social purposes, community-based and bottom-up, (Fressoli et al., 2011; Smith et al., 2014); base/bottom of the pyramid (BOP) (Prahaland and Hart, 2002; Prahalad, 2006; 2012), below the radar (Chataway et al., 2009), and reverse innovation (Agarwal and Brem, 2012).

We show that the substantial conceptual differences that exists in literature (Daniels, 2014; Smith et al., 2014) have led to ambiguities in our knowledge and understanding of innovation for development resulting in nuances in policy focus and emphasis, as well as lack of conceptual clarity. We submit that a revisit of this area of research is important in enhancing conceptual clarity on innovation for development.

Next we examine the proposals made for innovation ecosystems that can facilitate development outcomes. This includes delineating the types of claims made about institutions, landscape and innovation dynamics, as well as the roles and performance of actors (UNCTAD, 2011; Foster and Heeks, 2013). The comprehensive review will distill and analyse the implications of this variation. Some of the relevant material to be reviewed, will include but not be limited to the treatments of key actors, practitioners; intermediaries and institutions (formal and informal) in areas such as learning, knowledge generation, capacity building (Muchie et al., 2003; Marcelle, 2004; Berdegué, 2005; Bell, 2007; Oyeyinka, 2012); inequality and poverty reduction (Cozzens and Kaplinsky 2009; Arocena and Sutz, 2010; Cozzens, 2010; OECD, 2012; World Bank, 2013).

There is a small but emerging body of work based on empirical studies that attempt to demonstrate the contribution of innovation to socio-economic development and these case studies and illustrative examples convey evidence relating to the potential and actual contribution of reconfigured innovation to development and impacts on selected poor and marginalised sectors such as youth, women, disabled citizens (Gupta, 2003; Lorentzen and Mohamed 2009; Cozzens and Sutz, 2012). The analysis of this material permits drawing fresh insights obtained, we map the challenges and gaps identified and make recommendations for improvements on capability, knowledge production and dissemination, learning, innovation ecosystems and relevant institutional frameworks.

In the final section we address issues related to public (including innovation) policies and policy support for innovation for development which has been observed to be weak or non-existent in most countries (Gupta, 2013; OECD 2013; Daniels, 2014; Marcelle, 2014; UNCTAD, 2014).

In conclusion, our findings indicate that innovation for development offers substantial development opportunities and can be a useful mechanism for integrating various groups into the broader innovation ecosystem and national economic activity. If it is done successfully, it will facilitate social cohesion, entrepreneurship and job creation thus providing opportunities and tools for addressing global inequality and poverty, development challenges and policy goals and ensuring that innovation responds to societal needs.

ABSTRACT. We live in a fast changing world in which inequality, unemployment, poverty, low rates of life expectancy, child malnutrition and high rates of infant mortality represent tremendous social challenges, especially in developing countries. Science, technology and innovation (STI) policies could be means to trigger a dynamic integration of science, technology and innovation in the development of sustainable processes of social inclusion. The concern for determining priorities in scientific and technological research has been key since the beginning of the STI policies. It can be seen the way in which more developed countries have successfully established their policy priorities. What happens with the STI policies in developing countries, like those of Latin America? Certainly, attempts at creating programs or blueprints to plan scientific activity have been polemical, especially among the main actors of corporate economy. The persistency of social inequality in the region can be seen as a sign of the fact that, to date, social and STI policies have not been effective. In the specific case of STI policy in Mexico, it can be said that it is an isolated policy, determined without targeting poverty and social inequality. This is why is important to ask how were built STI policies in countries in which they have been instrumental to lower poverty and social inequality? This paper seeks to analyze and compare different approaches used in the selection of priorities in STI policies in several Latin American countries (Argentina, Brasil, Colombia, Chile and Mexico), with particular focus in the case of Mexico. In each case, it will be important to identify their particular strengths and weaknesses. The hypothesis to be defended is that there exists coincidences in the majority of selected priorities with developed countries as per developing countries. The definition of these priorities does not consider the overall dynamics and fundamental aspects: i.e., the context of production vis-à-vis the context of application, the actors involved, the expectations and the problems that look forward to solve. These dimensions are, beside, part of different economic and social dynamics that is necessary to consider in order for the priority selection to be effective. After proving this hypothesis, it will be proposed and supported the use of the analytical category of Strategic Priorities, considering the elements necessary for its definition. As will be shown, Strategic Priorities allow us to establish a comprehensive process to improve living standards, addressing real social problems from the outset. The analysis I propose will be focused on the public policies. It’s important to notice that efficacy in public policies is based upon the ability they have to adapt themselves to the systems of science and innovation. The research this paper is part of, attempts at contributing to de design of theoretical instruments that allow all the participants in the planning, implementation and assessment of public policies determine a set of Strategic Priorities according with its scientific and technological capacities and needs, in a way that allow to solve social problems within the frame of an inclusive and sustainable economic and social development. This paper has three parts. In the first one, it will be presented a brief evaluation of diverse experiences in developing countries in different times. In the second one, it will be presented a comparative analysis of five Latin American countries. In the final part it will be proposed and analyzed the category of Strategic Priorities.

ABSTRACT. Grassroots innovation (GRI) has sparked increasing interest among scholars in various countries, who are engaged in describing, analysing and conceptualising this phenomenon. Aggregating the research arguments of Hilmi (2012), Bhaduri and Kumar (2011), Hua, Jiang, and Lin (2010), Smith, Fressoli, and Thomas (2014), and Gupta (1989, 1992, 2006 and 2014) allows a very general definition of GRI as innovation by individuals, households, and communities to improve quality of life under conditions of limited resources and lack of access to formal innovation. The research has also identified some of the basic elements of GRI development in low- and middle-income countries: a large informal sector of the economy and a high rate of poverty as the environment for GRI generation; and shortage of resources (e.g., water, food, energy) or lack of access to them as the driver of the demand for GRI. Among the most important forces for GRI development are the existence of infrastructure for GRI promotion with active participation of the government, the academic sector, and business, and recognition of GRI by the government in order to provide grassroots innovators with needed legal, financial, infrastructural, and policy support. These elements and GRI itself need to be conceptualised in their social, economic, cultural, and historical context in relation to national innovation systems (NIS) in order to understand the position of GRI in the NIS of a given country, the connections and influences these innovations experience there, and the relationship GRI does or should have with other stakeholders in the country’s NIS.

GRI concepts applied to developed and developing countries may have limited application for describing grassroots innovative activity in Russia. The small size of the informal sector (about 17-18 per cent of the Russian economy), a low rate of absolute poverty, and a large proportion of the population being highly educated set Russia apart from the GRI environment, triggers, and demands present in developing countries. A weak civil society, narrow public policy in science, technology, and innovation development, and lack of interaction between society and government in this area also distinguish the context of Russian GRI from the European or American ones.

This paper considers GRI in Russia in its multifaceted context and identifies its specifics, its environment, and its place in the NIS. It examines the innovation policy of Russia, the degree of its inclusiveness, and the extent to which needs of grassroots people are supported or targeted in this policy. The paper also discusses the extent to which conditions for emergence of GRI are present in Russia. The prospects and implications for GRI development and policy interventions in it are summarized, based on selected case studies, expert interviews, and analysis of policy documents.

ABSTRACT. Singapore and Hong Kong are often placed to compare with each other due to many similarities. They are both small cities developed from British colonies and experienced an upgrade of industrial structure from labor intensive industries such as textile and clothing to high tech industries such as electronics and to banking and financial services (Young 1992). They are both limited by natural resources and comprised primarily of Chinese population. Their economic performance and competitiveness are rather comparable, which often places them next to each other in various rankings such as Global Competitiveness Report (WEF 2014), World Competitiveness Yearbook (IMD 2014), Global Innovation Index (GII 2014), etc.

Despite of the many similarities, the differences in the development trajectory of these two cities are also noticeable, one of which is the role of the government. Singapore is known as government-made city-state for the high level of government intervention in various aspects of the society (Mok 2005). By contrast, Hong Kong adopts “positive non-intervention” policy that favors free economy and minimizes the power of government in influencing the market. The difference is reflected in the innovation policy in these two regions. Singapore government has been actively implementing a series of strategies and allocated substantial fund to promote innovation and entrepreneurship. While a similar set of policies and programs are in place to drive research and innovation in Hong Kong, the scale is much smaller. For example, the Singapore government provided US$2.1 billion (S$2.7 billion) for R&D in 2012, accounting for 0.8% of GDP (ASTAR 2013; Table 1.6) . In the same year, the Hong Kong government financed US$0.8 billion (HK$ 6.8 billion) for R&D, only 0.3% of GDP (C&SD 2013).

Given the similarities and differences, Singapore and Hong Kong comprise interesting cases for comparison. In this paper, we will analyze innovation activities in these two cities and investigate how government intervention may influence innovation dynamics. It is generally in agreement that a capable government is important for national economy. Government is one of the determinants for innovation capacity according to the National Innovation System theory (Nelson 1993) or the Triple Helix theory (Etzkowitz & Leydesdorff 2000). However, how government should function is still debatable, especially regarding questions such as what role the government could play and how government could get involved in innovation. Some argue government should actively advise industrial policy and fund innovation activities. Others believe government’s main responsibility would be regulating and offering infrastructure and institutional support, while leaving innovation decisions to the market. Singapore and Hong Kong are representatives of these two streams. The comparison of innovation activities and performance in these two cities could shed light on the implication of government involvement in innovation.

ABSTRACT. In response to global climate change and energy security concerns, increasing the share of renewable energy in the electricity supply has been proposed as a promising solution in the United States. Among all types of renewable energy technologies, US has experienced tremendous technological progress in wind power over the past two decades and has become the world’s largest wind power country in terms of annual wind generation since 2008.

This paper examines the determinants of technological change of wind power in US from a learning perspective. Based on technological learning theories and collaboration theories, this paper uses a panel of 576 wind projects between 2001 and 2012 to test the effects of state level energy policies, as well as the impacts of different learning channels – learning through R&D in wind turbine manufacturing (learning-by-searching), learning from a wind farm operator’s previous installation and operation experience (learning-by-doing), learning from the experience of other firms (knowledge spillovers), and learning through collaboration among wind turbine manufacturer, project operator, and the transmission distribution system owner (learning-by-interacting) — on technological change in the US wind industry. This paper focuses on two aspects of technological change, which are the performance improvement in wind farm installation and wind farm operation respectively. Preliminary results suggest that a turbine manufacturer’s R&D improves the wind farm performance right after the installation stage. However, experience from the wind farm operator matters more for the operational performance over time. Particularly, the wind farm operator’s interaction with turbine manufacturers and the transmission distribution system owners greatly improves wind farm productivity. In addition, the evidence of knowledge spillovers within a state and the impacts of production-based policies including production tax credits and the RPS policies provide supports for state level energy policies that encourage wind power generation.

This study contributes to literature on state renewable energy policies and technological learning literature. Most existing empirical studies on state renewable energy polices focus on testing whether a particular policy, such as the renewable portfolio standards (RPS), and tax credits, has led to the diffusion of wind power in the form of installed capacity or share of electricity generation. The only one that examines the impacts of renewable energy polices on the technological change of wind power is the work of Nemet (2012), which looks at the learning-by-doing effects in wind projects within California. With a more representative sample of wind projects in US, this research provides a cross-state analysis for the technological change in wind power, and tests the impacts of other different channels of learning in addition to the learning-by-doing.

In addition to its contribution to the literature, this paper will be of interest to utility regulators and other policymakers for energy technology policies. Focusing on the roles of various actors in wind power industry, this research will increase the understanding of the learning process in the US wind industry. As a result, it will help policy-makers better target policies to different learning channels so as to improve the performance of wind farms.

ABSTRACT. The role of institutional agents such as universities, businesses and governments in increasing innovation has been underlined in the literature on national innovation systems (Nelson and Winter, 1982; Nelso, 2006; Freeman, 1987; Edquist, 1997), regional innovation systems (Cooke, 2001; Malerba and Nelson, 2012) and in the triple helix model (Etzkowitz and Leydesdorff, 2000).

Innovation policy can be defined as public action that influences the process, because the development and diffusion (product and process) of innovations and their results can be looked at from an economic standpoint, as well as from a cultural, social, environmental, or military one (Chaminade and Edquist, 2006)

Where the literature on innovation practice, innovation policy and innovation theory come together there is a constant search for ways to improve the effectiveness and efficiency of innovation policy and innovative practice. Within this dynamic process, the ideas, reasoning and tools of innovation policy appear as a result of the ongoing interaction of the agents involved in innovation practice, innovation-related public intervention strategies and innovation research and theory (Kuhlmann et al., 2010). The authors recognize the importance of learning processes, including policy learning. It documents the close interaction between analytical work on innovation and the planning and implementation of innovation policy.

The Brazilian development characteristics and the integration of innovation policy within this context after the enactment of the federal Innovation Law, in 2004, the states also began to propose and approve their own state laws in support of innovation, to also allow state budget resources earmarked for R&D to be allocated to companies for the purpose of innovation.

A singular characteristic of Brazil’s innovation system is that the university, rather than the business – as in The Organisation for Economic Co-operation and Development (OECD) countries – is the locus of innovation, particularly in high-tech areas (Póvoa, 2008).

A survey of the level of innovation in Brazilian industry indicates that the innovation rate is 35.7%, covering a total of 128,699 businesses employing ten or more people. The greatest innovation was in internal processes (18.3%), mainly the replacement of machinery and equipment, followed by innovation in both products and processes (13.4%), while innovation in just products alone was 3.9%. The distribution of innovation by sector shows that in the service sector, 36.8% of the companies carried out some kind of innovative activity, while in manufacturing the rate is 35.6% (IBGE/PINTEC, 2013).

The aim of this paper is to further the study of theoretical propositions related to innovation policy and its assessment by means of indicators, and to analyze how the strategy for the planning and implementation of innovation policy in Brazil since 2004 with a view to making an additional contribution to the analyses on this topic that have been carried out to date.

References

Chaminade, C. and Edquist, C. 2006. From Theory to Practice: The Use of the Systems of Innovation Approach in Innovation Policy. In: Hage, J. and Meuus, M. (Ed.). Innovation, Science, and Institutional Change. A Research Handbook. Oxford: University Press. Cooke, P. 2001. Regional Innovation Systems, Clusters, and the Knowledge Economy. Industrial and Corporate Change, 10(4): 945-974. Edquist, C. (ed.). 1997. Systems of Innovation: Technologies, Institutions and Organizations, London: Pinter/Cassell. Etzkowitz, H e Leydesdorff, L. 2000. The dynamics of innovation: from National Systems and ‘‘Mode 2’’ to a Triple Helix of university–industry–government relations, Research Policy, 29(2):109–123. Freeman, C.1987. Technology Policy and Economic Performance, London: Pinter. IBGE (Instituto Brasileiro de Geografia e Estatística). 2013. Pesquisa Industrial de Inovação Tecnológica – Pintec - 2011. Rio de Janeiro: Instituto Brasileiro de Geografia e Estatística/Ministério do Planejamento, Orçamento e Gestão. Kuhlmann, S., Shapira P., and Smits, R. 2010. A Systemic Perspective: The Innovation Policy Dance In: Smits, R., Kuhlmann, R. S., Shapira, P. (editors). The Theory and Practice of Innovation Policy. An International Research Handbook, Cheltenham, UK: Edward Elgar. Malerba, F. and Nelson, R. 2012. Economic Development as Learning Process. Variantion Across Sectoral Systems. Cheltenham, UK - Northampton, MA, USA: Edward Elgar. Nelson, R. R. and Winter, S. G. 1982. An evolutionary theory of economic change. Estados Unidos: Harvard U. P. Nelson, Richard R. 2006. As fontes do crescimento econômico. Campinas, SP: Editora da Unicamp. Póvoa, L. M. C. 2008. A crescente importância das universidades e institutos públicos de pesquisa no processo de catching-up tecnológico, Revista de Economia Contemporânea, 12(2): 273–300.