ABSTRACT. Emerging technologies are new technologies which promise much for firms, industries, and economies. As one of the well-recognized emerging sectors, the biotechnology industry has received longitudinal policy and development attention since the 1970s. Gathering international collaboration data regarding joint publication, co-patenting, clinical trials, and commercialization activities (marketing) from multiple sources along the innovation value chain to examine the international collaborations in the past three decades, this paper aims to analyze the dynamic structures of the global innovation networks in the biotechnology sector in order to study the role of policies in enabling the international engagement of the biotechnology sector . Particular attention will be paid to the role that technology policies play in the transformation of scientific knowledge into commercial technologies across continents to observe the trend of shifting global technological power over the past three decades. The findings show that in the biotechnology sectors, the leading technological power has remained in the U.S. and has not shifted to East Asia, though East Asian countries have tried hard to promote biotechnology sectors through international engagement. The co-evolvement of the vertical and horizontal policy instruments is crucial to support the development of emerging sectors. In the context of the biotechnology industry in Taiwan, several important policies were implemented rather late. In particular, policies aimed at enhancing the global engagement of emerging technology sectors are still rare. While facing the rise of China, firms and institutes in Taiwan have to strengthen their engagement with foreign technology resources. In fact, compared to the manufacturing sector, Taiwanese actors and Chinese actors in the biotechnology sector have built their own respective international networks, which means the Taiwanese actors are less dependent on the Chinese economy while establishing emerging sectors. While emerging technologies are mostly science-based and originally came from Western countries, policies for enhancing international technology transfer from technology frontier to the technology followers should be implemented earlier to shape the structure of the emerging industries at an earlier stage. Only then can the effects of the policy implementation be more efficient in enhancing the development of emerging technologies.

ABSTRACT. 1. Background and hypotheses

Economics of innovation literature has extensively studied how scientific knowledge is transferred to productive sector and society and how the private sector absorbs and applies scientific knowledge for the development of innovations. However, this is usually a bidirectional relationship and, in most cases, the engagement with the productive sector also impacts on academic activity in different ways: opening new research questions, training human resources, giving additional research funding, opening new research contacts, etc. So, it is to be expected that technology transfer/economic engagement impacts on research by increasing academic productivity. Anyway, the literature is not conclusive about the presence of a positive or negative relation between both dimensions. Some contributions show that collaboration with private sector has a positive effects on academic productivity (Gulbrandsen and Smeby, 2005; Van Looy, Callaert and Debackere, 2006; Zucker and Darby, 2007; Lowe and Gonzalez-Brambila, 2007; Perkmann and Walsh, 2008; Breschi, Lissoni and Montobbio, 2008). Then, there are some results that show the opposite relationship (Lin, 2006; Czarnitzki and Toole, 2010). In spite of these contributions, relatively little attention has been paid to this topic from developing countries perspective (Povoa and Rapini, 2010; Rivera-Huerta et al, 2011; Garcia et al, 2011; Rapini, 2007; Rapini et al, 2015; Chaves et al, 2015, Barletta et al 2017; among others). Thus, the main contribution of this article is to discuss the effect Argentinean private engagements research groups have on the academic productivity in multipurpose technologies (Nano, Biotechnology and ICT). So, as hypothesis 1 we propose that academic productivity of multipurpose university R&D research groups –centered on nanotechnology, biotechnology and ICT- are higher in groups involved in technology transfer activities. This motivation is particularly relevant in the current context of the scientific Argentine system, in the framework of a debate on how to value the university- industry interaction in the evaluation of the researcher's career and on how to generate incentives for these linkages to happen. So, evidence on the impact of academic- industry collaboration on scientific outputs is fundamental for shaping research policy design, particularly in terms of evaluation system and incentive scheme. Another contribution of the paper is that we explore academic productivity of research groups instead of considering individual researchers productivity on the basis that research is a collective and not an individual task (Acosta et al, 2019; Barjak and Robinson 2007). The empirical exercise is carrying out on research groups in multipurpose technologies due to the available information. However, we expect different results according to the knowledge discipline and we divide between bio and nano, for one hand, and ICT for the other hand, for the following reasons. Firstly, in the case of Argentina, bio and nano research groups are usually related to high-tech productive sectors such as chemical, pharma and GMO seeds industries which have strong path dependence in Argentina. Secondly, bio and nano research groups have strong incentives and more pressure to reach scientific outcomes and have publications since more than 40% of the groups belong to the National Council of Scientific and Technological Research (CONICET). So, they are subject to performance rules pondering the publication in scientific journals. By the opposite only 11% of ICT research groups belong to CONICET. For these reasons, we propose the following hypothesis 2: when comparing Bio-Nano by one hand and ICT by the other, we expect transfer activities had a higher impact on productivity.

2. Empirical analysis

The empirical analysis is based on a set of data that includes bibliometric information for each one of the members of research group from two sources of information. Firstly, we have used a survey of 250 Argentinean research groups on multipurpose technologies that capture the existence of technology transfer activities to private firms along the period 2013-2014. Secondly, we have used a bibliometric database for each one of the research group’ members based on data about Elsevier's Scopus scientific publications. The integration of these two databases resulted in a dynamic panel made up of 237 research groups between 2000 and 2017. It is an unbalanced panel, with an average frequency of observations per group of 13 observations. Thus, the academic productivity of each group is estimated as the ratio between the number of papers published during 2014-2017 and the number of researchers of each group. In order to identify the causal effect of technology transfer activities on the academic productivity, we design an identification strategy based on the application of a Lagged Dependent Variable model (LDV). The identification assumption is that the treatment-free potential outcome (papers per member) for both groups is the same in expectation conditional on the past outcomes and observed covariates (Angrist and Pirschke, 2009). Under this assumption, research groups with similar outcomes in the pre-treatment period would be anticipated to have similar potential outcomes in post-treatment periods after conditioning on observed covariates (O’Neill et al. 2016). While this seems to be a strong assumption, following Garone et al. (2017) we include a rich set of interaction terms between yearly dummies and research group’s characteristics that allows us to control for different trends at the group level and unobserved cofounders that change over time and affect groups in different disciplines, institutions or locations in different ways. We consider alternative identifying assumptions to check the robustness of results and employ a fixed effect model. If the unobserved heterogeneity is fixed in time, but we mistakenly use LDV, the positive estimates of the effect of technology transfer will show an upward bias. Conversely, if the unobserved heterogeneity is time-variant, but we use a Fixed Effects (FE), the positive effect of technology transfer will tend to show a downward bias. In this direction, Angrist and Pischke (2009) demonstrate that FE and LDV bound the actual causal effect.

3. Results Results show a positive effect of university-industry linkages on the average academic productivity of research groups. Thus, its support hypothesis 1: the estimated effect of university-industry linkages lies in the range between +0.14 (LDV model) and +0.43 (FE model) papers per member of research group. This verifies the robustness of the empirical strategy. The model also showed differences between the disciplinary areas. Despite of the model considered, the positive effect on Bio and Nano more than duplicates the effect on ICT. Thus, results also support the second research hypotheses. Results point out the need to discuss the Argentine scientific policy, historically oriented to support basic research with scientific publication ends. In this context, this study suggests that publication and university- industry collaboration are not excluding activities but reinforce each other. Results also can raise new questions for future research. By one hand, it could be interesting to explore if the results will be maintained for other knowledge disciplines. Due to the available information, this study has focused on multipurpose technologies but it could be important to increase the scope in future research. By the other hand, it is expected that the impact on academic productivity differ according to the type of university- industry interaction. So, future research can explore if the increase of academic productivity derived from economic engagement can be explained by the predominance of research driven consultancy over opportunity driven ones.

ABSTRACT. Background

Sustained development and deployment of clean energy technologies is expected to make a significant contribution towards climate change mitigation goals. Fostering technological innovation in energy is especially challenging due to the scale and complexity of the modern energy and technological systems. Technological innovation is intimately related to knowledge creation and recombination, and effectively solving scientific and technological problems often involves research at the intersection of different science and technology (S&T) domains. As a matter of fact, many recent innovations in energy devices and systems have been achieved with research at the intersection of the domain of energy technology and the domains of nanoscience and nanotechnology, and electrical engineering and computer science. In order to devise policy instruments to promote energy technology innovation, it is important to know the type and extent of related research so that the allocation of RDI funds can be strategically directed. Advances in methods that aim to define the extent and intensity of boundary-spanning research and illuminate the processes which underlie knowledge creation, recombination and diversification are crucial for the design of effective technology innovation policies. Here, we present a combined statistical and network-based approach to study collaboration in scientific authorship within and across various sectoral, regional, and national research networks. We apply this approach to Greece and its broader European and international environment as a case study.

Methods

Statistical analysis of information about scholarly publication of research has served as a transparent means to improve decision-making in knowledge management. Defining the scope of a S&T domain based on scholarly publication is a methodological challenge. We embed the collection of publications with simple search terms in a workflow for information retrieval that serves to enlarge an initial set of core publications by a form of query expansion. On the basis of the collected publications, we construct and analyze coauthorship networks to map boundary-spanning research, highlighting research clusters (teams) as the elementary unit, as opposed to studies of team formation and evolution which examine the internal composition of teams and their size. In doing so, we complement previous work that aims to define diversity-based indicators of interdisciplinarity by introducing time-dependent indicators to quantify knowledge diversity the networks. Our approach offers a multi-level view of knowledge organization on the basis of the research interests and collaboration partners of individual scientists.

Results

Using our methods we attempt to illuminate the processes which underlie knowledge creation and diversification in these research networks. An important question we attempt to answer is to which extent independent energy-related research efforts have contributed to shape existing knowledge diversity into a distinct field of inquiry. We establish a (positive) relationship between gross expenditure on research and development (GERD) and the extent and diversity of team-based research in renewable energy technology and its intersections with the domains of nanoscience and nanotechnology with focus on materials, and electrical engineering and computer science. It could be argued on the basis of this finding that increased R&D spending would result in higher levels of such boundary-spanning research. However we also find that research is carried out mostly by independent research teams –thus a distinction needs to be made between fortuitous and deliberate boundary-spanning research. The fact that the boundary-spanning research network comprises mostly disconnected clusters (teams) at all times is indicative of barriers to the advancement of deliberate interdisciplinary research. Our findings suggest that policy initiatives could, for instance, target specifically the interface of the domains of renewable energy technology and electrical engineering and computer science to create opportunities, if desirable, that promote coherence of the research network and interdisciplinary activity. We also identify regional research hubs in interregional and international collaboration which support the two knowledge bases and are in close proximity, introducing a component of regional specialization.

Significance

Our specific findings collectively provide insights into the collaboration structure and evolution of energy-related research activity in Greece and contribute towards an improved understanding of the Greek innovation system. With a discussion of international collaboration, the implications of our specific findings necessarily extend beyond national borders. We expect that our findings pertaining to Greek research networks will also be relevant to other national RDI systems of similar extend, within which the main contributor to GERD is the public sector, and when there is an interest in incentivizing interdisciplinary research to promote innovation. More generally our approach can be used for operationalizing boundary-spanning research to support the design, monitoring, and evaluation of interdisciplinary research programs on the basis of empirical evidence.

ABSTRACT. While the systems of innovation approach has been used extensively as an analytical framework to understand innovation activities at the city, regional, and national levels, less attention has been devoted to cross-border regional innovation system, which is defined in our study as adjacent regions divided by a nation-state border line. One the one hand, technological transformations have redefined the boundaries of regional innovations; on the other hand, the institutional differences between cross-border regions may continue to impede the interaction and collaboration. This study applies the systems of innovation approach to examine the dynamics of cross-border regional innovation system in the context of Hong Kong-Shenzhen metropolis. Being adjacent to each other and having close social and economic linkages, Hong Kong and Shenzhen present an intriguing case for studying cross-border regional innovation. On the one hand, Hong Kong has its distinctive legal, economic and administrative system as a special administrative region of China, presenting a range of institutional differences that are common for cross-border regions; on the other hand, recently announced Greater Bay Area Initiative calls for greater integration and synergetic development in technological innovations between Hong Kong and Shenzhen. In this research, we will use publication and patent data from Web of Science and USPTO respectively to study the degree of integration in the innovation areas and actors in the cross-border regional innovation system. We will also apply social network analysis to illustrate the existing and potential collaboration network. Our study seeks to shed lights on potential complementarities, institutional barriers, and collaborative mechanisms for fostering cross-border regional innovation system.

ABSTRACT. 1. Introduction

This paper contributes to the growing literature on changing research careers of scientists. It focuses particularly on how geographic mobility affects career progression across Europe. The paper utilizes a novel data feature of a 2016 European wide survey (MORE3) covering an estimated 1.4+ (*) million researchers. The focus is on a comparison between mobile and non-mobile researchers in terms of their progression through a succession of career stages. Scientists at universities in 31* European countries reported their progress through four career stages defined in terms of the acquisition of sets of professional competences (EU, 2011).

The inclusion of careers stage information for both mobile and non-mobile researchers creates a unique dataset for investigating whether scientific mobility affects the career path of scientists in different contexts. By analyzing differences between these two populations the paper sheds new light on the following questions: - How does research career progression differ in different (country and disciplinary) contexts? - How does international mobility vary at different career stages? - Does international mobility affect the rate of career stage progress in different contexts?

The paper reviews recent literature at the intersection of scientific mobility and changing careers in the academy. In doing so, it emphasizes aspects of the formative European labor market that confronts today’s scientists. In this context, different conceptions of scientific career are presented, factors that may affect appointment and promotion at different stages are identified, and strengths and weaknesses of different measures are discussed. The paper then models mobility-career relationship using available data and presents results about the effect of long-term (> three months) mobility on career progression rates.

2. Background

International research mobility is understood to play an important role in the scientific community (see Guena, 2015 for a recent review). In Europe, where numerous academic traditions meet in different national systems (with differing labor markets, languages, etc), the context has been changing for some time (e.g. Enders, 2001). A first stream of the researcher mobility literature explicitly includes a theoretical model of research career stages (e.g. Laudel (2005), Laudel et al. (2018)). This literature illustrates how cognitive, peer and organizational careers interact and how mobility functions within this tripartite conceptual framework. A larger body of the research mobility literature does not (explicitly) operate with a theoretic understanding of the career stages of a scientist. The focus—as well as the results— of this literature are mixed. For example, this work finds that research mobility:

a. Is perceived by researchers to have positive effect on career development (e.g. Bauder (2002), Børing et al. (2015) Musselin (2004) Stephan et al. (2014) ); b. Has been mainly found to have a positive effect on labour productivity (Fernández-Zubieta 2009, Franzoni et al. 2015, Lu & Zhang 2015, Moed et al. 2013, Veugelers & Bowel 2009), but with some exceptions (Cañibano et al. 2008, Halevi et al. 2016); and c. Has mixed effects on promotion (e.g. Lawson & Shibayama (2015), Marinelli et al. (2015)).

The paper discusses these results in specifying what we mean by research progression, how we measure it, and how our results might contribute to policy intentions given the significant policy focus the issue has in Europe (see below). In doing so, we take into consideration work in adjacent fields, such as the labour economics literature on 'promotion rate', which also sheds light on timing of promotion for different populations (Blau and DeVaro, 2006).

3. Data and Approach

The paper utilizes data from the latest round of the MORE family of surveys , which has become a centerpiece for policy studies of mobility and career paths in Europe. The MORE studies support ongoing development of the European Higher Education Area (AHEA) and the European Research Area (ERA) and the positive value placed on open labour markets and mobile researchers in both these policy frameworks. It utilizes a two-stage stratified sampling strategy based on three fields-of-science (FoS) for the 31 countries. The sample of 10,400 scientists is designed to be representative of the overall research population as calculated and published by Eurostat. The questionnaire involved a total of 104 questions.

3.1. The focal measure: Progression through careers stages R1-R4 for mobile and non-mobile respondents

Official statistics do not routinely include information about current career stages, and this poses a challenge to our understanding of factors that affect career progression. The MORE3 survey takes steps to advance knowledge on this important issue. It pools available information about subpopulations and tests post-stratification weights to frame as correctly as possible researchers at different career stages. This advancement however is not straightforward and poses some challenges for our analysis (see Idea Consult & WIFO 2016: section 5.4.1).

Our focal measure segment the research career into four according to standardized stages of the scientific career (EU, 2011). Respondents reported the year they passed from being a doctoral student or “Early Career Researcher” (R1), to being a “Recognised Researcher” (R2), an” Established Researcher” (R3) and a “Leading Researcher” (R4). These transitionary phases are coordinated with other information (e.g. entry into current employment) to establish baseline progression in how populations in different country and disciplinary settings accumulate experience, credentials, and recognition during their career. Other information about contract type, teaching loads and levels of satisfaction provide ancillary information to help better understand these stages. The paper then analyzes the effect of mobility experience on our measure of career progression in light of other factors and a set of controls. We test different models to deal with different issues (e.g. simultaneity) in the data.

4. Preliminary Findings

Our (unweighted) findings indicate strong variations in the temporal structure of careers in different parts of Europe. We observe:

a. European researchers start as R1 in their late 20s, transition to R2 at 33, transition to R3 at 38, and transition to R4 at 43. However, there are some country, gender and FoS differences; b. Researchers in some countries and regions change jobs more often than in some others; c. The time taken (years) to reach that Leading (R4) career stage differs. In many countries the duration increases for each career stage (R3> R2> R1). In another group of countries, R2 is shortest career stage. d. Respondents in the Leading (R4) career stage who were mobile during the last 10 years move more quickly through R3 (6.2 years) than non-mobile respondents (6.8 years), however no difference was observed in rate of progression through the R2 stage and into R3.

The findings to date therefore suggest that some relationship exists between international research mobility and career stage progression rates, although this effect appears to be modified in different ways by factors including country, FoS and gender.

ABSTRACT. Background & Rationale I will present recent work quantifying the impact of two types of mobility in science: cross-disciplinary and cross-border. Importantly, both types of mobility are important mechanisms for promoting scientific convergence — be it transdisciplinary integration as a strategic means for addressing grand challenges or national innovation system integration as a means for increasing international competitiveness.

In the first case, we use the Human Genome Project as a case study to measure the convergence between computing and biology - a ‘perfect marriage’ giving rise to the field of genomics. By analyzing the temporal evolution of the collaboration network formed between 4,190 biology and computing faculty from 155 departments in the U.S. and the citation patterns across their 413,565 publications, we are able to quantify the role of cross-disciplinarity in the important context of socio-cultural integration between distinct fields and its impact on combinatorial innovation at this critical biotechnological frontier.

In the second case, we analyze patterns of international collaboration before and after the 2004/2007 European Union enlargement to highlight the incremental yet counterintuitive dis-integration of European science following the enlargement. Incorporating inter-European high-skilled mobility data into our quantitative framework, we identify causal channels linking subsequent decreased rates of international collaboration to the brain-drain following the enlargement. Based on further analysis of researcher mobility at the individual career level, we confirm the obvious candidate mechanism linking mobility and collaboration dynamics. We conclude by estimating the impact of researcher mobility in physics by comparing mobile to non-mobile researchers, appropriately matched according to geographic and prior researcher profile characteristics.

Methods Descriptive analysis of large and comprehensive data samples collected specifically for each study motivate each facet of the overarching research topic: precision measurement of the impacts of mobility in science. We use various statistical techniques — difference-in-difference, synthetic control method, and propensity score matching — to identify causal effects.

Results Collaboration rates (cross-border and cross-disciplinary) and relative citation impact of individual publications are the primary units of measurement.

In the first case, in which we compare matched samples of cross-disciplinary to mono-disciplinary publications, we estimate a roughly 20% increase in citation impact for genomics publications featuring cross-disciplinary collaboration between faculty from both biology and computing departments — measuring this differential relative to the mono-disciplinary research published by the same researcher (i.e. estimated using an author fixed-effects specification).

In the second case, we estimate that most 2004/2007 entrant EU member states would have had higher rates of international collaboration had they not joined the EU. To be specific, we estimate a ~8% decrease in cross-border collaboration attributable to the combined effects of entering the EU and the subsequent brain-drain primarily from eastern to western entrant countries.

We provide additional direct evidence for the causal channel by analyzing roughly ~26,000 researcher profiles form physics, thereby estimating the rate at which individual collaborations disintegrate when one researcher moves. By way of example, in the most extreme scenario, we find that researchers completely curtail all prior collaborations 11% of the time they migrate. Despite the destructive features of researcher mobility, giving rise to substantial churning in collaboration networks, we measure a 17% increase in citations for mobile researchers relative to their non-mobile counterparts. This result can be explained by their simultaneous increase in diversity of co-authors, topics, and geographical coordination in the period immediately following migration.

Significance The results of our analysis will be framed in terms of their implications for immigration and national innovation system policy.

[1] National Research Council. Convergence: facilitating transdisciplinary integration of life sciences, physical sciences, engineering, and beyond. National Academies Press; 2014 Jun 16.

[2] Fleming L. Recombinant uncertainty in technological search. Management science. 2001;47(1):117-32.

[3] Marx M, Strumsky D, Fleming L. Mobility, skills, and the Michigan non-compete experiment. Management Science. 2009 Jun;55(6):875-89.

[4] Chessa A, et al. Is Europe evolving toward an integrated research area?. Science. 2013;339(6120):650-1.

[5] Petersen AM, Puliga M. High-skilled labour mobility in Europe before and after the 2004 enlargement. Journal of The Royal Society Interface. 2017;14(128):20170030.

[6] Petersen AM, Majeti D, Kwon K, Ahmed ME, Pavlidis I. Cross-disciplinary evolution of the genomics revolution. Science advances. 2018; 4(8):eaat4211.

[7] Arrieta OA, Pammolli F, Petersen AM. Quantifying the negative impact of brain drain on the integration of European science. Science advances. 2017;3(4):e1602232.

[8] Petersen AM. Multiscale impact of researcher mobility. Journal of The Royal Society Interface. 2018;15(146):20180580.

ABSTRACT. The accessibility of knowledge determines innovation, which further influences productivity and economic growth. Thus, understanding knowledge flows is an important topic for economists. One particular factor to consider is the mobility of foreign-born scientists, because they connect foreign knowledge flows to national scientists. But how does scientific mobility impact knowledge diffusion? In literature, it is controversial to claim that scientific mobility increases knowledge flows due to two reasons. First, the observed positive effect might result from the fact that one institution consistently favors the knowledge from another (e.g., because of its reputation). Second, local absorptive capability, inter-institutional cooperation, knowledge distance and administrate barriers further complicate the question. In such cases, it is hard to quantify the impact from scientific mobility. Thus, it is an indispensable option to conduct a comprehensive study that addresses all the aspects above via hypothetical experiments. In particular, while the existing studies focus on quantifying self-generated knowledge of foreign scientists, we also account for whether they transfer their colleagues’ knowledge in a broader way.

In our study, we measure the effect on knowledge flows by counting the number of forward citations (citations made to the focal paper by other papers in the future). This research presents new pairing techniques on new data, and thus contributes to the existing literature. Specifically, we first identified moving scientists who were enrolled in the Chinese Youth 1000 talent program from 2011-2015. We traced approximately 1100 moving scientists and identified around 300 distinguishing pairs of origin-destination institutions related to the moves. Then we examined the citations of papers from origin institutions by destination institutions before and after the moving year with consideration of the long-term effect (we observe three years before and after the moving year). Most importantly, we constructed a paper-twin control group by coarsened exact matching (CEM) as the identification strategy and difference-in-differences as the econometric tool to analyze the causal effect of scientific mobility on knowledge diffusion across institutions. The examination on multiple interaction effects verified our existing understanding of knowledge diffusion by further probing into the mechanism that underlies this relationship.

To correct for any potential unobserved heterogeneity, we conducted CEM based on a set of covariates for papers from the same journals, volumes and issues, number of authors, authors' country group, and similar citation trends before the mobility event (a proxy for pre-existing differentials). To construct these paper-twin groups, we also tested the difference of citations to papers without authors from China. We also controlled variations in locations of destination institutions and scientific fields.

Our results show that, citations to focal papers by new institutions do not positively increase than controlled papers, and papers heterogeneously diffused across fields. Also, returning to their alma mater could increase the influence of scientific mobility on knowledge flows. However, we did not find a significant effect that scientists bring knowledge created by their peers to destinations as expected. These findings contribute to the previous work by revealing the mechanism that generates the effects of scientific mobility on knowledge diffusion, not only the moving scientists' knowledge, but also their peers' knowledge. Overall, the results suggest that the return of foreign-trained scientists does not benefit new institutions in terms of knowledge diffusion as we expected. These findings contribute to the literature and present an integrative perspective of scientific mobility, knowledge diffusion and the globalization of expert knowledge.

ABSTRACT. Background and Rationale: There exists a large body of research on which mechanisms predict entrepreneurial entry, yet the amount of research conducted on which mechanisms affect entrepreneurial exit is considerably less. Prior literature assumes that the firms that performed the best were the most likely to survive, and these studies indicate that those who produce low financial returns simply fail altogether and exit entrepreneurship. Other research has found that the determinants of performance and survival may differ. They claim that many entrepreneurs persist in entrepreneurship for long periods of time without making large financial returns and without any intention of growing. This finding implies that producing only low wages does not mean an entrepreneur is necessarily going to leave entrepreneurship; they may choose to remain in entrepreneurship for the non-pecuniary benefits of entrepreneurship. A model is needed that combines both the impacts of pecuniary and non-pecuniary benefits. The goal of this paper is to demonstrate that studies that do not include a measure of non-pecuniary benefits with research on transitions in and out of entrepreneurship are missing out on a fundamental component. One of the problems with the current literature is that while studies claim that people may remain in entrepreneurship for the non-pecuniary benefits of entrepreneurship, there is little empirical evidence examining the relationship between non-pecuniary benefits and entrepreneurial exit. Additionally, the literature has begun to acknowledge that human capital and traits of the individual entrepreneur influence financial performance in entrepreneurship, yet there has still been little work done on the relationship between entrepreneurial human capital and non-pecuniary benefits. This opens the question who exactly is exiting entrepreneurship. Is it those who are unsatisfied in it, since non-pecuniary benefits are a main motivator into entrepreneurship, or is it those who are not earning high salaries in entrepreneurship? I argue that it is a combination of both mechanisms that influences entrepreneurial exit and that this relationship is largely dependent on the entrepreneur’s human capital. While pay and non-pecuniary benefits will drive exit, these in turn reflect underlying skills and how appropriable and valuable these skills are for entrepreneurship versus outside options in wage work. It is important to observe the entrepreneurial activities in which they participate to gain a better understanding of their entrepreneurial abilities, as entrepreneurs’ do not know their entrepreneurial abilities pre-entry. I argue that human capital of entrepreneurs should be directly related to their levels of pecuniary and non-pecuniary benefits in wage work, which should impact entrepreneurial exit. In the sample of scientists and engineers for this study, having specialized R&D skills should have a greater comparative advantage for wage work in science and engineering, than in entrepreneurship which should require a greater mix of technical and business skills. On the other hand those with a more generalized skill set with a mix of R&D and non R&D skills may experience high non-pecuniary benefits in self-employment. This group will be more likely to persist in entrepreneurship because they not only derive utility from entrepreneurship in terms of non-pecuniary benefits, but may have less outside options to re-enter wage work in science and engineering. Time in entrepreneurship may lead to a depreciation of the specialized human capital previously gained in wage work. I argue that those who have a diversified skill set and work in entrepreneurship by working in non-technical and management activities in entrepreneurship and who have allowed their specialized skill set to depreciate, may experience more disadvantages by returning to wage work with a generalized skill set. However, those who have not lost specialization in their skill set should be able to return to wage work and not be at a disadvantage. Some entrepreneurs continue to focus on specialized technical skills and work in teams or hire others to do non R&D activities, and therefore their skill set should not depreciate. Those who continue to work in specialized skills in entrepreneurship should be more likely to be those who have high salary but possibly lower levels of non-pecuniary benefits in entrepreneurship. This group should have more employment opportunities available to them outside of entrepreneurship and therefore should be able to improve their salary when they exit. Methods and Results: My empirical analysis uses data from the National Science Foundation’s Scientists and Engineers Statistical Data System (SESTAT). SESTAT is an integrated database that includes demographic, employment and educational information about scientists and engineers in the U.S. This paper uses data collected in the years 2003, 2006, 2008 and 2010. I use proxy measures for both non-pecuniary benefits (job satisfaction) and skill variety (R&D and Non R&D skill counts) to better analyze the relationship between entrepreneurial human capital, the non-pecuniary benefits of entrepreneurship and entrepreneurial exit. I then study what happens to entrepreneurs’ work outcomes once they return to wage work. My final sample of entrepreneurs includes 8,513 entrepreneurs, with 1,413 who transition out of entrepreneurship. I find that there is a positive but non-linear (inverted u-shaped) relationship between salary and entrepreneurial exit and a significant negative linear relationship between job satisfaction and entrepreneurial exit. I find that those with lower job satisfaction are more likely to exit entrepreneurship, regardless of salary. Those with greater R&D skill sets are those who are more likely to report lower job satisfaction but higher salaries in entrepreneurship compared to other entrepreneurs. They are the most likely to exit entrepreneurship, with their greater opportunity costs to remaining in self-employment and the marketability of their specialized skill set in wage work. On the other hand, those with a higher number of non R&D skills, and who have greater skill variety are more likely to be those with high job satisfaction. I contribute to the skill variety literature and entrepreneurial human capital literature by finding support for the idea that skill variety is beneficial in entrepreneurship as a contributor to job satisfaction. For those who exit entrepreneurship and return to wage work, I find that they are able to significantly increase their salary but not their job satisfaction, compared to remaining in self-employment. This paper adds to the understanding of entrepreneurial exit in three important ways. First, this study highlight that decisions made by scientists and engineers to exit entrepreneurship and return to wage work are largely influenced by not only their salary but also the non-pecuniary benefits they derive from entrepreneurship. Second, this paper speaks to the emerging literature on human capital and entrepreneurial exit by highlighting the importance of human capital in entrepreneurial exit decisions. While prior work in this domain has considered how the entrepreneur’s human capital relates to the financial performance of the firm which influences exit, this paper expand this to consider how the entrepreneur’s human capital relates to the non-pecuniary benefits an individual gains from entrepreneurship, which may influence exit. Finally, this study uses longitudinal data to analyze the changes in pecuniary and non-pecuniary outcomes for those who re-enter wage work.

ABSTRACT. Introduction The use of scientometric indicators for individual research assessment has been severely criticized over the years due to their limited capacity to discriminate between different scientists and capture differences in a statistically reliable manner (Costas, van Leeuwen, & Bordons, 2010). Nevertheless, science managers and policy makers make use of these indicators for recruitment of scholars, promotion or allocation of funds. This has provoked strong reactions from the academic community, such as the San Francisco Declaration (DORA, 2014), a specific mention warning on the dangers of using bibliometrics for individual assessment (Hicks, Wouters, Waltman, de Rijcke, & Rafols, 2015), or even a whole body of literature discussing the pros and cons of the H-index (Rousseau, García-Zorita, & Sanz-Casado, 2013), the most renown indicator for assessing individual research performance. We argue that the greatest threat of the current use of bibliometric indicators for the assessment of scientists goes beyond technical or methodological decisions, and is more related to the irreflexive use of metrics at the individual level. We claim that this irreflexive use of metrics endangers the diversity of the scientific profiles researchers exhibit. This diversity is not only evident (Larivière et al., 2016), but needed to ensure scientific progress (Milojević, Radicchi, & Walsh, 2018) and a breadth of societal and scientific outcomes (Woolley & Robinson-Garcia, 2017). Some evaluation models for individual assessment have been proposed in the literature (e.g, Bozeman, Dietz, & Gaughan, 2001; Wildgaard, Larsen, & Schneider, 2014). But they have not been able to prevent the irreflexive use of bibliometric indicators. In our belief, there are three reasons behind this failure: 1) they propose the introduction of a wide range of indicators, of which not all are necessarily operational; 2) they are framed in such terms that are difficult to operationalize; or 3) they deny the use of quantitative indicators without offering a viable and cost-efficient alternative. By linking with the current literature and our own experience on conducting research evaluation, we here present a tentative valuation model which tries to balance between a conceptually-informed framework and a methodological viable operationalization. The model is designed so that it can be operationalized by making use of bibliometric indicators, although we acknowledge that it is sufficiently broad as to give room to non-bibliometric indicators. Main pillars of the valuation model The model is structured into three distinct parts. The first and main one has to do with the actual performance of the individual in a set of five dimensions of the scientific practice. The second one addresses confounding effects derived from the individual’s context, such as work environment, institutional logics or national policies shaping their performativity. The third pillar of the model relates to personal features of the individual. In principle, these characteristics hold little relation with researchers’ performance, but can be of special interest for policy makers. For instance, science managers may be interested in promoting young researchers within a given programme, reduce gender inequality by encouraging the recruitment of women, or try to integrate and promote foreign born scholars. Evaluative dimensions We consider five dimensions as key factors to value the research performance of individuals. Scientific engagement, social engagement, capacity building and trajectory look into diverse aspects of the individual’s academic activities. However, the research practices dimension is represented as an overarching dimension which affects the other four. In the following, we describe each dimension. Capacity building refers to the capacity of the individual to create new knowledge, train new scholars or develop novel applications. Some indicators operationalizing this dimension could be number of publications, normalized citation score, but also number of PhD students supervised or generation of patents. Scientific engagement includes activities and actions reflecting a proactive engagement of the individual with the scientific community. This not only refers to scientific collaboration or division of labour, but also to reviewing papers, editing journals or organizing and participating in conferences and seminars. Social engagement is conceived here as outreach and interaction with societal actors. For example, different modes of engagement would be considered (D’Este, Llopis, Rentocchini, & Yegros-Yegros, 2015) as well as social outreach for instance by written for non-academic audiences. Trajectory reflects aspects related to the academic background of the individual such as geographical mobility, disciplinary changes or previous work experience. Research practices are conceived here as an overlapping dimension which modulates each of the other four based on how open or closed these are. For instance, share of OA publications would reflect openness in capacity building, while diversity of stakeholders could apply in the case of social engagement. Conclusions and further steps We propose a new evaluation model of scientists which considers the wide variety of profiles and activities researchers perform. The model captures the heterogeneity of activities and roles researchers perform into five dimensions by which they can be profiled, also quantitatively. Furthermore, the model considers confounding effects mediating on individuals’ performance as well as personal features which might be of relevance for science managers. The model is still under-development and still many caveats need to be solved as well as to the application of such a model on real case scenarios. For this we intend as future steps to carry out a series of case studies that can portray how each dimension interacts with each other and portray complementary profiles of scholars within research teams. References 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(7-8), 716–740. Costas, R., van Leeuwen, T. N., & Bordons, M. (2010). A bibliometric classificatory approach for the study and assessment of research performance at the individual level: The effects of age on productivity and impact. Journal of the American Society for Information Science and Technology, 61(8), 1564–1581. D’Este, P., Llopis, O., Rentocchini, F., & Yegros-Yegros, A. (2015). Star vs. Interdisciplinary scientists? Exploring distinct patterns of engagement in university-industry interactions. Presentado en University-Industry Interactions Conference, Berlin. DORA. (2014). San Francisco declaration on research assessment. Recuperado a partir de http://am.ascb.org/dora Hicks, D., Wouters, P., Waltman, L., de Rijcke, S., & Rafols, I. (2015). The Leiden Manifesto for research metrics. Nature, 520(7548), 429-431. Larivière, V., Desrochers, N., Macaluso, B., Mongeon, P., Paul-Hus, A., & Sugimoto, C. R. (2016). Contributorship and division of labor in knowledge production. Social Studies of Science, 46(3), 417-435. Milojević, S., Radicchi, F., & Walsh, J. P. (2018). Changing demographics of scientific careers: The rise of the temporary workforce. Proceedings of the National Academy of Sciences, 115(50), 12616-12623. Rousseau, R., García-Zorita, C., & Sanz-Casado, E. (2013). The h-bubble. Journal of Informetrics, 7(2), 294-300. Wildgaard, L. E., Larsen, B., & Schneider, J. (2014). ACUMEN DELIVERABLE D5.4b – Consequences of Indicators: using indicators on data from Google Scholar (p. 9). Recuperado a partir de http://curis.ku.dk/ws/files/124046907/4._Deliverable_5.4b_consequences_of_using_Google_Scholar_data.docx Woolley, R., & Robinson-Garcia, N. (2017). The 2014 REF results show only a very weak relationship between excellence in research and achieving societal impact. Impact of Social Sciences Blog. Recuperado a partir de https://blogs.lse.ac.uk/impactofsocialsciences/2017/07/19/what-do-the-2014-ref-results-tell-us-about-the-relationship-between-excellent-research-and-societal-impact/

ABSTRACT. Novelty is one of the core values in science, and as such, it is highly regarded in the recognition system of science (Dasgupta and David, 1994; Gaston, 1973; Hagstrom, 1974; Merton, 1973; Stephan, 1996; Storer, 1966). However, novel research also faces a higher level of risk, and there is increasing concern that the current science system fails to support enough novel research but favors conventional and safer research (Boudreau et al., 2016; Nicholson and Ioannidis, 2012; Wang et al., 2017). Since novelty is the driving force of scientific progress, it is imperative to create an environment that encourages risk-taking and novel research.

In spite of the importance of novelty, we do not know whether or how an individual scientist can acquire skills for generating novel idea and become more creative as an individual. The source of creativity has long been studied in multiple lines of literatures. For example, there is an consensus in the psychology literature in adopting the product definition of creativity, which highlights two criteria of the creative product: novelty and usefulness (Amabile, 1983). This literature has uncovered a variety of personality and group determinants of creativity (Amabile, 1983; Ford, 1996; Woodman et al., 1993). Building on these general theories about creativity, previous research has also explored how creativity in science is affected by traits of individual scientists, as well as characteristics of scientific teams (Cummings et al., 2013; Lee et al., 2015; Simonton, 2003). Others have explored institutional factors, such as funding schemes, leadership, organizational structure and culture (Andrews, 1976; Heinze et al., 2009; Hollingsworth, 2004; Wang et al., 2018). While our knowledge regarding the source of creativity has been expanding at a fast rate, we still do know how novelty is built into the norm of scientific communities or to the practices of individual scientists.

This study focuses on academic training as a key mechanism for fostering tastes and skills for novel research in junior scientists, thereby passing it down to the next generation. Junior scientists – e.g., postgraduate students and postdocs ("students" hereafter) – typically go through learning-by-doing by engaging in an actual research project as apprentices under senior supervisors ("supervisors" hereafter) (Delamont and Atkinson, 2001; Latour and Woolgar, 1979). In this setup, supervisors could offer students with novel projects, let them learn skills needed for novel research, and socialize them into the norm of pursuing novelty. Anecdotes also suggest that creativity transmits across generations: Nobel laureates are often apprentices of Nobel laureates (Zuckerman, 1967).

We test this training or learning effect using a large sample of Japanese doctoral students and their supervisors. We found that supervisors’ novelty has a positive effect on students’ novelty during their PhD studies. However, it is questionable whether this positive association is because of the training/learning effect, as the PhD project might be to a large extent decided by the supervisor. To address this issue, we replicate the analysis and still find a significantly positive effect of supervisors’ novelty on students’ novelty when (a) analyzing only students those who played a leading role in their PhD studies and (b) examining students novelty after they are tenured. Under both circumstance, it is unlikely that the supervisor will choose the project for the student. Another possible explanation is that more creative scientists are more likely to have access to and select creative students to train. We use another dataset to test the training effect after correcting for this selection effect. Specifically, we test the effect of postdoc supervisors’ on postdocs’ novelty after controlling for postdocs’ novelty during their PhD studies, the results confirms significant and positive training effect.

A more intriguing finding is that supervisors’ novelty has a long-term positive effect on students’ novelty after the students are tenured, regardless whether the student engaged in novel research during the PhD studies or not. However, this positive training effect is large when the student do engage in novel research during their PhD studies. On the other hand, when a student engages in novel research during PhD under a supervisor who is not inclined to novel research, his/her research will be less novel after tenure. These findings highlights the importance of supervision: skills for novel research can only be gained or sustained under proper supervision.

ABSTRACT. 1. Introduction

For decades, achievements of university-industry collaboration (UIC) around the world motivated governments to issue various UIC (Larsen et al., 2016, ch.3). However, the effectiveness of government-funded UIC program for promoting a firm’s innovation is still less clarified. Some studies, on the one hand, found that the public program of UIC triggered the positive effect of the firm’s innovation performance (Scandura, 2016; Wirsich et al., 2016). On the other hand, some researches addressed the negative effect on the product innovation of the firm affected by government-sponsored UIC either (Maietta, 2015). For re-understanding the puzzle between firms’ output and public-funded UIC, this article combines the assessment of a novel UIC program issued by the Taiwanese government on the national science parks and the theory of institutional entrepreneurship to an analysis. The research question, thus, is marked that how does the institutional entrepreneurship on a different part of the firm organization drive the innovation performance of the firm through exploiting the chance of granted UIC.

2. Conception and Hypotheses

The institutional entrepreneurship, referring the discussion by Teece (2010), is conceptualized as the professionality of individuals shaped by their past professional experience. This article distinguishes the different direction of the work of the professionality into top management team (top-down way) and general employee. In short, two hypotheses are here: a) The professionality on firm’s top management team will increase the innovation performance of the firm affected by government-sponsored UIC, and b) The professionality on firm’s employee will increase the innovation performance of the firm affected by government-sponsored UIC.

3. Variable, Data & Method

The innovation performance of the firm affected by government-sponsored UIC, explained variable, is measured by the number of US granted patent of firms. Otherwise, the explaining variables on the first and second hypothesis are respectively operationalized as the degree of education of board members and the number of doctoral employees interact with UIC policy. Various control variables are set to elaborate on the effect of independent variables.

The program, The Ministry of Science and Technology’s Project for Industry-Academia Collaboration on Innovative R&D in Science Parks (PIACIRD), and 253 public high-tech firms located in the science parks of Taiwan for 13 years (2005 to 2017) are the empirical data of this paper.

In order to estimate the hypotheses, this article adopts the empirical strategy suggested by Heckman (1979) to infer the causality. I implement the strategy with three parts. First, I used propensity score matching (PSM) to reduce the self-selection within PIACIRD. Next, I used the difference-in-difference approach to evaluate the basic impact of PAICIRD on the innovation performance of the firms. To gauge the core explaining variables, finally, the interaction terms were disposed between the impact of PAICIRD and the professionality of the firms.

4. Result & Conclusion

The results supported the second hypothesis and rejected the first one. The number of doctoral employee in firms positively enhances the number of granted patents significantly. The assessment of the three-way fixed model showed that, statistically, one additional doctoral employee hired by funded firms could add 0.5 granted patents on average than the firms without public UIC subsidy. This addresses that though the institutional entrepreneurship can enlarge the effectiveness of government-sponsored UIC, the intra-organizational process of knowledge production may hinder the work of the institutional entrepreneurship. These findings contribute to future UIC research and implication for UIC policy design.

ABSTRACT. In competitive market economies, technology is one of the main assets that provide a competitive advantage to the firms developing and applying them on new products and processes. Firms need to be up-to-date on the existing technological regimes to remain competitive and be able to survive, despite the characteristics of the environment (i.e., rules, consumption habits, competition laws, sectors growing and declining, educational characteristics) may differ over time. This is one of the main messages provided by evolutionary economists. According to this theory, firms, and the territories in which these are located, follow evolutionary processes as a result of (technological) change, leading to innovations that shape societal change. This line of reasoning lead during the 1990s to the emergence of the so called field of innovation studies, which approaches technology and innovation from both micro and macro perspectives. One of the key questions that has for long brought the attention of innovation scholars is understanding which is the most effective means by which firms and territories can diversify their knowledge and technology bases. In this regard, one of the concepts with the highest echo not only in the academic world but also in the policy realm is the concept of related variety, which is anchored in the literature on relative competitive advantage. This classic concept, central to the studies of business and economics, shows how a firm/sector/territory should specialize in those areas of activity and technologies in which its knowledge bases have the largest scope for exploitation (i.e., strengths). The underlying idea behind the concept of related variety is that when a firm/sector/territory specializes in a particular composition of complementary sectors to those in which it shows its main strengths, then it will experience higher growth rates than those that tend to specialize in other sectors that do not complement each other. The related variety approach finds its rationale in that complementary sectors share technological capabilities and competences with those already present in the firm/sector/territory. Accordingly, when two firms/sectors are close to each other, but still are slightly different, it becomes easier for both of them to understand and absorb the new knowledge/ technologies developed and transferred by the other. So far the literature has shown that related variety is a more effective strategy for diversification since the emerging new branches of activity are substantiated on assets and capabilities already present in local firms or in the territory. However, in latest years there have been some critical voices who argue that unrelated diversification strategies could be even more effective, particularly in the long run. Unrelated variety is the opposite of related variety. According to it, the greater the variety of exploitable niches in the economy, the more potential innovation opportunities there will be, so existing and new technologies can migrate from niche to niche to increase their market exposure and acceptance. From this point of view, limiting the variety of an innovation system would also limit the future opportunities for entrepreneurs, challenging the future development of the system as a whole. The advantage of this approach would be that when one sector is hit by an economic downturn in a territory with a high degree of unrelated variety, this will not negatively affect the other sectors. In contrast, if the same territory would rely on related diversification, since the sectors in it are ‘closer’ to each other, the probability for negative externalities to spread throughout the entire system would be much larger. Taking this scholarly discussion as a starting point, the paper wonders whether technological diversification is only a matter of related or unrelated varieties. It adds to the literature by providing a discussion of the extent to which territories, firms and sectors should follow related and/or unrelated diversification strategies. The paper focuses on the evolution of the technological diversification followed by the Basque Country, one of the leading industrial regions in relation to innovation not only in Spain but also in Europe. The Basque region is often regarded as one of the few truly comprehensive regional innovation systems and one of the territories with a more articulated and comprehensive innovation policy making process. We here analyze the logic behind the technological diversification shown by the world patents (WIPO) granted to companies located in the Basque Country between 1991 (year of the first granted PCT patent) and 2011. By focusing on this 20 year period we aim to observe whether related or unrelated diversification strategies are more effective in periods of economic expansion (years 2003-2007 where periods of high growth in Spain and in the Basque Country) or in periods of high uncertainty and economic crises (2008-2011). Our hypothesis is that diversification strategies should not be either related or unrelated, but should rather naturally evolve between the two, depending on the conditions that are shaping the environment over time. The paper provides a twofold perspective on technological diversification, by joining sectoral and territorial views. This constitutes a relevant contribution to the literature, which provides either case studies in particular sectors or evidence at the territorial level. The extant evidence suggests that all sectors, regardless their location, may follow similar dynamics, and correspondingly, that all sectors within a territory may also follow the same dynamics as the territory as a whole. This however, does not seem a very plausible rationale, as sectors, technologies and territories show divergent paths in their evolution. The evidence to date does however not help to disentangle the different paths that sectors may follow in a territory. This paper aims to provide a new perspective on technological diversification, by focusing on both sectoral and territorial levels. We introduce a novel approach for the measurement of technological diversification of territories over time based on the Herfindahl–Hirschman index, which has been extensively used to assess concentration patterns. We also develop a quality indicator for measuring technological diversification based on Boolean logics. The paper focuses on the evolution of the technological diversification followed by the Basque Country. We analyze all the world patents granted to Basque companies in the period 1991-2011 (n=1102) and their technology classes (m=1177). Our results evidence that for the Basque Country as a whole technological diversification starts under a related variety paradigm, but it continuously shifts to get closer to unrelated variety logics In turn, from the sectorial perspective they show that both related and unrelated variety play a significant role in explaining the technological evolution of Basque firms.

ABSTRACT. Background and rationale

For most countries, national R&D statistics is found for volume (money spent) and distribution of activity across sectors (public, private, etc.). These numbers, however, do not tell us anything about how many different organizations that are involved in a nation’s R&D activity or how the activity is distributed across these organizations. An interesting empirical question is who accounts for the R&D activity of a country within each sector. To the best of our knowledge, there is no available database where one can decompose national numbers below the sector grouping. That is, to study the R&D activity per higher education (HES) institution, government sector research institution (REC) or private company (PRC).

This study makes two contributions to the scientific literature. First – empirically - we use proposals submitted to the European Framework Programs (EU FPs), FP7 and Horizon 2020, to make country profiles of 40 European countries’ R&D profiles. Second – theoretically – we use these profiles to see whether they correspond with R&D performance indicators, i.e. we aim at understanding what characterizes countries that perform especially well on R&D performance indicators.

To the best of our knowledge, no similar study has ever been performed, because no relevant data is available. Having spent more than two years on standardizing the institutional names of applicants to the framework programs for the period 2007-2017, we have now finally been able to perform such an analysis.

Methods

We use a unique dataset covering all HES, REC and PRC institutions that have applied for funding to the EU FPs in the period 2007-2017, taken from the European Commission’s data warehouse ECORDA. At the Nordic Institute for Studies in Innovation, Research and Education (NIFU), all institutional addresses and their sectorial belonging in ECORDA have been standardized (approximately 1,4 million institution names). In total there are 111,238 unique institutions from the 40 countries that have participated in proposals, with Germany (13,652), Italy (13,404) and UK (12,400) having the most.

For each country in total – and per sector – we have calculated a ‘skewness’ indicator using the Gini Coefficient calculation, i.e. we have calculated ‘R&D Gini Coefficients’ based on each participating institution’s share of the nation’s (or sector’s) total volume of proposals to the EU FPs. We then compare the R&D Gini values with R&D performance indicators from OECD (Gross Domestic Expenditure on R&D (GERD) as percentage of GDP), the European Innovation Scoreboard (we have constructed an ‘innovation’ indicator made up of patents, trademarks and design) and citation scores (using Web of Science). We do not claim to make a ‘complete list’ of all R&D performing units per country. Rather, we believe that the contributing units from each country in sum will provide national profiles of the general magnitude and dispersion of the R&D performing units.

Results

Any causal mechanism from high R&D investments to high performance indicator values, must be transferred through a set of R&D performing actors. Our results suggest that the transition of funding to R&D output goes through highly dispersed R&D systems in a large bulk of the most successful R&D nations.

All six countries with R&D investments above 2.5 per cent (GERDs) except Israel, have either top scores on both innovation and citations (Denmark, Sweden and Switzerland) or second highest score on citations and highest on innovation (Finland, Austria and Germany). The R&D Gini coefficients in these countries are high, clearly indicating that the most successful R&D countries of Europe do so with a work-force of R&D institutions that are log-tailed, i.e. the efforts are pulled by a relatively small group of leading actors, not by a generally moderate activity level across all R&D actors. The latter is more typical in countries with low R&D investments. For countries with low investments, there is a clear tendency that there are higher citation scores in countries with high R&D Gini Coefficients compared to countries with low coefficients, across the same GERD levels.

Do highly cited R&D systems have a more skewed distribution of R&D actors? Yes, the pattern clearly indicates a moderate decline in R&D Gini Coefficients as the citation index drops (R²=0,654).

Are the systems skewed in both academic sectors and in the PRC sector? The evidence for this is mixed. On the one side, all the most cited European countries (except Iceland) have relative high sector R&D Gini Coefficients, but we find the most skewed systems among the large countries that are not among the most cited countries. The general finding of the sector results is that is not so much a matter of a gradient, but of a divide between Western and Eastern European countries.

What is the extent of R&D intensive locomotives? The results are undoubtedly associated to citation scores. Although large countries like Germany, France, Italy and Spain for obvious reasons have much higher numbers of R&D locomotives (i.e. institutions with more than 500 proposal contributions), the per capita numbers of locomotives are lower than in smaller, yet higher cited countries.

Conclusion/significance

Our study shows that the concentration of R&D actors at country-level and within the sectors differ across European countries, with the general conclusion being that well-performing countries seem to combine a) high R&D investments with b) a highly skewed R&D system, where a small part of the R&D performing actors account for a very high share of the national R&D performance. This indicates a dual R&D system which combines a few large R&D performing institutions with a very large number of small actors.

An interesting question is a) whether a skewed set of PRC actors is consistent with strong innovation output, while skewed sets of HES and REC actors is consistent with strong citation scores, or b) whether skewed sets of actors in all three sectors are consistent with strong performance output indicators overall. In the first case (a), one may claim that industry ‘innovates’ and universities ‘publish’. In the other case (b), strong presence of both types of actors may benefit a country’s capability of both innovation and research, as the science-industry cooperation may involve more mutual set of benefits than what is the focal point in innovation studies; mostly describing the knowledge transfer from universities to private companies only. The empirical evidence related to case a) above, is that the countries with the most skewed sets of PRC actors are not the best performers on innovation output. In the second case, skewed sets of HES or REC are also only weakly associated with good citation scores.

By contrast, three out of four countries (Iceland is the exception) with highest citation scores (Switzerland, Denmark and the Netherlands) combine the following elements: very high overall R&D Gini coefficient values, high innovation scores and no particular dominance of any sector. Switzerland stand out as the most consistently skewed R&D nation, with very high R&D Gini coefficients observed across all three sectors. We believe our study actualise future studies that investigate how locomotives from all three sectors may be catalysts for R&D processes both leading to strong research and innovation output and which mechanisms of skewed R&D systems that are beneficial compared to more egalitarian systems.

ABSTRACT. This paper proposes to measure and explain technological competitiveness based on the driving forces of technological innovation by originality and generality indicators. Competitiveness has been associated in the innovation studies literature as a desired outcome of technological innovation which has become pivotal for a country’s economic growth in the era of global knowledge economy. Attention has increasingly shifted towards the long-run perspective on building a solid foundation for continuously generating technological innovation, which suggests that progress comes in waves, each one originating with a major breakthrough or generating potential wide application. The central components of national systems of innovation are the innovative activities performed by practitioners (mainly corporates). However, little is known about evaluating the effects of corporate R&D on technological competitiveness outcomes at aggregation levels of industries and countries.

This paper seeks to assess whether industry or a country is likely to be (or become) more competitive in terms of technological originality and generality of corporate R&D. Patent data have been used as indicators of technological innovation as well as the technological competitiveness of an industry or a country. Therefore, changes on the corporate trends of patented technologies in an industry or a country will affect the research and development strategies of all involved actors in an innovation system. This paper argues that originality and generality are both vital features of technological development to trigger innovation, and becoming increasingly important for long-term competitiveness. Therefore, this paper proposes the extension of originality and generality indicators derived from (Trajtenberg, Henderson, & Jaffe, 1997) for measuring the technological competitiveness at the industrial and national levels based on corporate R&D.

The technological capacity of the world’s economies contributes to achieve sustained economic growth (McArthur & Sachs, 2001). From a competitive perspective, it appears that industries and countries with higher levels of technological development may be perceived as better performing in terms of economics than those with lower levels of technological progress. However, to what extent would the technological competitiveness of an industry and a country last longer than the others? This question needs further study with a long-run perspective. From the perspective of long-term economic growth, the global comparison in terms of the technological competitiveness today is the one between industries, countries, and that are able to create or absorb technological innovation rapidly and those that are not (McArthur & Sachs, 2001). This paper argues that technological originality and generality of corporate R&D not only can lead to better exploitation of technological progress made, but also inspire exploration towards a new territory. Technological originality can be regarded as the knowledge base of the technological innovation, can serve as a direct source of idea for initiating new possibilities or generating more efficient or effective engineering designs. On the other hand, technological generality has become the most vital endogeneity for a country or an industry to sustain its growth with its erogeneity as technological originality (Rosenberg, 2013)(Gilson, 2010). Originality and generality index firstly proposed by (Bronwyn H. Hall, Jaffe, & Trajtenberg, 2001a) enable researchers to understand the distribution of the technological origin and fields which are influenced afterwards by the calculation of backward and forward citations. However, there have been no concerted effort on extending originality and generality indicators for measuring technological competitiveness at industry and country levels through the use of patent citation information.

The indicators of originality and generality were firstly developed and constructed with patent citation data by (Trajtenberg et al., 1997) and regarded as measures of basicness and appropriability. Basicness indicates fundamental features of technologies, such as originality, closeness to science, and breadth when appropriability refers to the ability of inventor to gain the benefits from their own innovation, such as generality. Existing studies of innovation process largely emphasize the importance of links between basicness and appropriability. Therefore, basicness and appropriability have been both regarded as the sources of heterogeneity which take the prominent place in the economics of technological change as well as key dimensions of technological basis and outreach, in which accumulated and formed the knowledge foundation of an economy (Trajtenberg et al., 1997). Appropriability have been featured by the technological development and the legal mechanisms for patent protection (Teece, 1986).

In order to study, measure, and rank technological competitiveness, a few models have been proposed as good measures that capture the nature of competitiveness. For considering both the origin of technological capability and the magnitude of technological influence for better measuring technological competitiveness, the heterogeneity of analytical components and the importance of long-term perspectives need to be recognized, and this recognition is the major contribution of this study to both the international competitiveness literature and patent citation indicators.

The objectives of this paper is to make a further step to the extension of technological originality and generality proposed by (Trajtenberg et al., 1997) and further developed by (Schmoch, 2008) in order to analyze industries and countries’ technological competitiveness based on corporate R&D. This paper, which reports the results of an empirical investigation based on patent data obtained from USPTO, provides new findings bearing on this issue. By extending the originality and the generality indicators to include knowledge basis and technological influences, a more holistic picture of the effects of technology on a country’s competitiveness can be obtained, than by examining trade flows and foreign direct investment only.

ABSTRACT. Entrepreneurs have been known to play vital roles in developing and transition economies (Brasil, India, China and South Africa) because they played crucial role in structural transformation from low tech based labour intensive economies to high technology, knowledge based economies through adoption, assimilation and adaptation of foreign technologies from developed countries (Etzkowitz et al., 2015). Thus, the entrepreneur is the indispensable machinery for knowledge creation and knowledge diffusion (Braunerhjelm, Ács, and Audretsch 2010). There are lessons to learn from newly industrialising countries. For instance, China’s rapid industrialization over the last few decades disconnect with the popular widely accepted path in which developed countries in Europe took – having some financial standard and framework (Allen, Qian, and Qian 2005). In particular, the vast majority of small firms had little or no access to credit from state-owned banks (Ruan and Zhang, 2009; Firth et al 2009; Lin, 2011). Despite this major constraints, China has achieved industrialization in less than a quarter of the time it took Europe. How was China able to do this? Through the clustering mode of production (Allen, Qian, and Qian, 2005; Fisman and Love, 2003a; 2003b). It can be assumed that by dividing a cohesive production process (which agglomeration offers) into several incremental phases, clustering can lower capital entry obstacles, creating a pathway for more entrepreneurs to join production industries. Also, clustering promotes inter-firm trade credit therefore reducing the pressure of looking operating capital. Nigerian entrepreneurs are taking a peek from the Chinese story and trying to follow the shorter part towards industrialization. The Nigerian government too has been supporting the development of clusters as through its industrial policy and programmes. The Nigeria Industrial Revolution Plan (NIRP) was designed to facilitate the development of industrial cities, parks, and clusters while focusing on making hard infrastructure available within these industrial zones. Currently Nigeria ranks second in West Africa in industrial cluster development (World Competitiveness Report, 2015). The top cluster destinations in Nigeria include: Nnewi (Automotive), Otigba (ICT), Yaba (Technology), Onitsha (Plastics), Aba (Footwear) and Kano (Leather). Asides these major clusters, there exists several hundreds of other smaller clusters. The government of Nigeria has also established several free trade zones in different geopolitical zones of the country even though not all of these free trade zone in Nigeria were created by government at the Federal and State level, but these free trade zones have enjoyed the support of government. Within the free trade zones, goods may be landed, handled, manufactured or reconfigured, and re-exported without the intervention of the customs authorities.The study outlined how the microenterprises overcome the barrier of capital and credit through ‘open development’ characterised by collective sharing of knowledge, tools, equipment and workforce amongst competing microenterprises. Also by belonging to trade/professional associations that ensure that knowledge becomes a public good that can be accessed by everyone who needs it. The study also found out that majority of the microenterprises experienced scaling-up within few months to around three years from inception, while only a few scaled up only after three years. Scaling-up of microbusinesses was captured with four proxy variables: increase in number of employees, increase access to capital, increase in gross earnings and increase in gross sales. A major reason for the rapid growth in the cluster was knowledge sharing not access to finance. There was serious emphasis on sharing knowledge, skills and ability that each enterprise had to collectively compete. The nature of open collaborative innovation in the cluster was through cooperation to collectively compete as a cluster (with international market). In the cluster, knowledge acquisition was achieved either through formal methods (university education and trainings) and informal method (apprenticeship system and indigenous knowledge systems). However, apprenticeship system of education was the most used channel of acquiring knowledge in the cluster. Knowledge diffusion was communal in the cluster. especially with the presence of trade associations/union guiding wide dissemination of knowledge in the cluster. This is because of the monitoring role played by trade association/unions such as CAPDAN and COMPTON evidently present in the cluster. Amongst other reasons, knowledge sharing and diffusion influenced scaling-up in the cluster this manifested principally in increase in number of employees within enterprises. Improved access to finance, increase fine gross earning and increase in gross sales Hence, knowledge sharing was seen an instrument of collective advancement (open development). The practice of open collaborative innovation among knowledge-based enterprises/networks has been found to be highly productive in overcoming the barriers to accessing finance in the cluster. It is therefore highly recommended that government, unions, professional bodies, trade associations and self-help organizations buy into this, as knowledge sharing and collaborative problem solving approaches represent the currency in which enterprises in clusters need to trade with. How much an enterprise knows, how fast they can learn something new, and how much knowledge it is willing to volunteer determine the vibrancy of the enterprise not really how much it has.

ABSTRACT. Background and rationale The evaluation of the performance of research organizations is an important issue in science policy. The challenge of the scientific management of a research organization is to evaluate the research work of the organization and the diversity of research achievements towards the owners, and to develop visions for the future research agenda. However, it is a complex challenge to identify, monitor and benchmark the competences of a thematically diversified research organization on one hand and to position it in the scientific community and develop a future research agenda based on the international research findings on the other hand. The diversity of research skills can be very broad in a research organization, such as a university or a research institute, and can span over many disciplines. Therefore, the impact of research in the scientific community is difficult to be measured. Classical performance indicators often used in the bibliometric evaluation only allow quantitative statistical statements. Examples for such indicators are the number of publications or citations, the H-index or impact factors. They only allow a rough insight view in the target achievement of the organization. Statements about fulfilling the mission, as well as about its impact on research or contributing to tackling major challenges such as climate change, environmental degradation or emergent technologies are not possible. This presentation proposes an approach that closes this gap. The variety of competence fields of an organization is displayed in a science map based on the publication output of the organization. The individual fields of competence are sub-fields of content-clustered publications. To capture the impact fields, all publications that cite at least one publication of the organization are analyzed and clustered in subfields in a second science map. The contents of the fields are collected from the bibliographies, the author's keywords and the publications. The research work was part of a strategic project of the Austrian Research Center for Industrial Biotechnology ACIB in Vienna and Graz and was funded by the COMET Program of the Austrian Research Promotion Agency FFG. ACIB is an important Austrian player in basic and applied research in industrial biotechnology. The project aimed to develop a decision-making basis for the future research agenda and results were included in the report to the scientific advisory board. Objectives for this project were: Analysis and evaluation of the research and competence portfolio of ACIB and of the impact of published research as a starting point for future research fields. Methods and Data We used bibliographic coupling and a science mapping approach to visualize, identify and analyze the research fields and expert workshops to evaluate the results. In the first task “Research Portfolio of ACIB” all 675 publications (retrieval date: 29.5.2017) with at least one authors from ACIB were downloaded from the Web of Science database referring to the affiliations, the funding agency and the Comet Program of the FFG. The documents were mapped and clustered by bibliographic coupling to identify and describe the competence fields. The results were displayed with a 2D and 3D surface map. The relative publication activity and the knowledge growth were used as two dimensions to position the competence fields in a research portfolio. The results were evaluated with the chief scientific officer of ACIB. 4899 citing publications (retrieval date: 06.09.2018) that cite at least one author of ACIB served as the database for the identification of the impact fields. They were retrieved with the help of the Create Citation Report feature of Web of Science. The procedure to delineate the impact fields was the same as for the competence fields. Indicators to evaluate the impact fields were the number of citing publications, average citations per document, average year of references (as proxy for the novelty) and the closeness of acib to the impact fields measured by the number of common references. Results The quantitative analysis of the research output as competence fields of the research organization ACIB confirmed that the research organization is strong in traditional topics like recombinant protein production in Pichia Pastoris (yeast; recombination protein, whole-cell biotransformation, secretion, …), and others. The published work is dominated by established researchers. Impact fields with the same research issues as the organization allowed the international positioning of research done. There are also impact fields that do not consist of the same research than that of the organization. As the organization works on technologies in biotechnology the results are used in other fields of biotechnology that are of growing interest but have not net yet been considered in the organization. From the centers scientific management perspective new remarkable potential issues could be identified from the impact fields like cell engineering reprogramming and directed differentiation of pluripotent stem cells where ACIB has competences in the technology. The Lignin depolymerization by white rot fungi wa identified as a hot issue for the future research agenda. Relevant impact fields have been assigned to the new areas for the future funding period together with the management of the research center. With the help of this evaluation procedure the future establishment of new research fields for the new funding period were considered and could be argued in a systematic way for the future funding period. Of course, it was to be linked to the demand of the industry that is represented by present and future industrial partners of the center. The publication activity of the center and particularly the kind of journals and the active participation in conferences shed light on its mission. It is a flagship center with a certain but not strong orientation to fundamental research. However, it is strongly oriented to perform applied research projects with industrial partners. That is the reason why the published work is not strongly to be found in basic research journals. A bigger amount of publications in top ranked journals would have needed another orientation of the mission.

ABSTRACT. Overview and motivation

Interdisciplinarity is increasingly used as a criterion for the evaluation of research proposals by research funders, e.g. in Germany through the DFG Priority Programs or DFG Research Centers (DFG 2015a, 2015b)). Many research organizations also use the fact that they have "interdisciplinary" research teams as a marketing instrument, implying that there is an expected benefit of interdisciplinary research that is commonly shared. Yet, the possibilities to publish interdisciplinary research are limited. Many top-tier journals are limited in their focus and interdisciplinary research often does not "fit" into their scope, which basically results in penalties for interdisciplinary research as lower-ranked journals have to be targeted. Before making judgements about this phenomenon and argue for changes in one or another direction, however, we first of all should be able to answer the question whether more interdisciplinarity is always "better", or whether this is true only up to a given threshold, in certain dimensions etc. Therefore, we need methods to measure the interdisciplinarity of a research project, a research group or a research organization. In the literature, there have been several attempts to measure interdisciplinarity with the help of bibliometric indicators. This is highly useful, as scientific papers include lots of bibliographic information - scientific disciplines, citations, reference lists, etc. - that can be used to assess the interdisciplinarity of the given paper, a research team or an organization. However, this might also lead to a one-sided assessment of interdisciplinarity, which is further aggravated by the described phenomenon of rather homo-thematic high-tier journals. In this paper, we therefore propose several non-bibliometric measures of interdisciplinarity - alongside some traditional bibliometric measures - and apply them to the German Fraunhofer Institutes, i.e. at the level of research organizations. In a further step, their validity is tested against common bibliometrics based measures. Once the indicators are established, they will be used to create an "interdisciplinarity index", i.e. a composite indicator with which we can assess interdisciplinarity in a broader sense. This also enables us to relate the interdisciplinarity indicator to certain performance measures. The data used are partly exploratory in nature and some of the data are only available as "internal" data from the Fraunhofer Society, e.g. the educational background of the researchers. However, they shall serve as a "proof of concept" regarding the measurement of interdisciplinarity.

Defining and Measuring Interdisciplinarity - A quick look at the literature

By investigating interdisciplinarity, we examine the question how competences, perspectives, knowledge etc. from different disciplines are used for the work of an organization. This corresponds to the generally used definition of interdisciplinarity, namely bringing together and applying different aspects from several disciplines (van den Besselar/Heimericks 2001) and sets it apart from multidisciplinarity, e.g. different perspectives on the same topic taken by different disciplines, or transdisciplinarity meaning basic research questions (and sometimes also methods) that cannot uniquely assigned to a single discipline (e.g. Alvargonzález, 2011). Depending on the applied concept, there are different operationalizations of interdisciplinarity. The vast majority of empirical literature, however, uses bibliometric data for its measurement (Mugabushaka et al. 2016, Nagaoka/Kwon 2006, Shafique 2010, Small 2010). The measures range from differences in cited references, shares of references to publications outside of one's own discipline to publications with more than one subject category or combinations of indicators (e.g. Levitt/Thelwall 2008, Rinia et al. 2001, Steele und Stier 2000, Stirling 2007). Many researchers further use Stirling's (2007) conceptual decomposition of interdisciplinarity (diversity, variety & balance) as the basis for more refined bibliometric measures, also regarding the measurement of a relation between interdisciplinarity and scientific impact in terms of citations in publications (Wang et al. 2015). In this paper, we include further data sources for the measurement of interdisciplinary research. Though earlier research has shown that information contained in scientific publications can help in measuring interdisciplinarity and other data sources are scarce, we believe that only using bibliometrics delivers a one-sided picture of the phenomenon. Kwon et al. (2016) and Nichols (2014) are notable exceptions that experimented with other data sources. Kwon et al. (2016) used patent data, which we will also use for further analysis. We also propose to look at patent information. In addition, we will use data on the disciplinary backgrounds of the involved scientists as well as data on the projects in which the researchers were involved.

Data & Indicators

Since we use the Fraunhofer Society as an example to test our indicators, the single Fraunhofer Institutes (FhI) serve as the level of analysis. To create indicators and calculate the interdisciplinarity of each FhI, we use data from several sources. For the bibliometric indicators, we employ Scopus by Elsevier and generate five bibliometric measures: average number of scientific fields of a publication (10 fields aggregate of the Scopus classification), share of publications in "multidisciplinary", share of citations from publications from other fields, share of references to publication in other fields. For the patent indicators, we use data from the EPO Worldwide Patent Statistical database (PATSTAT) to generate a Herfindahl index of the spread of patent of an FhI by IPC classes. A matching of PATSTAT data and the Orbis database by Bureau van Dijk further allows us to generate a spread of an FhI's patents across industry sectors. Furthermore, we use funding data from the German "Förderkatalog" (Funding Catalogue of the Federal Government) that lists public research spending by all German ministries. We hereby analyze how many partners on average were involved in all FhI's research projects. We also analyze in how many projects from different fields an FhI was involved on average. Furthermore, we make use of internal data on the educational background of the FhI employees, which can be analysed on aggregate to find out about the average number of different disciplinary backgrounds within an FhI. Finally, we aim to incorporate further bibliometric based indicators by taking a more in-depth look at the titles and abstracts of publications and see whether a keyword based approach can be used to assess a paper's interdisciplinarity. With a similar approach, we aim to take a closer look at the websites of the single Fraunhofer Institutes, which would add a further dimension of interdisciplinarity.

Outlook and first results

Once the indicators are calculated, we can first of all compare them with the help of a correlation analysis. This provides us with a view on the interrelation between the indicators as well as the respective data sources. In order to find out whether they measure the same or different things, we further employ a factor analysis. First results show that though some indicators are highly interrelated - especially the bibliometric indicators - there is variance what the indicators measure, implying that our additional indicators seem to deliver some explanatory power. Preliminary factor analyses show that there seem to be three different latent constructs within the concept of "interdisciplinarity", namely "cooperation behavior", "knowledge generation and distribution" (mostly citation based measures) and "discipline and field structure" (spread across disciplines and fields in terms of publications, patents and employment structure). Next steps comprise the refinement of the indicators, the extension of the indicator set to further, web-based indicators as well as the testing whether the interdisciplinarity measures are related to performance or "impact" in a broader sense.

ABSTRACT. Background

This study creates a novel inventory that characterizes factors influencing the research agendas of researchers in STEM fields – the Multi-Dimensional Research Agendas Inventory – Revised – STEM (MDRAI-R-STEM). This inventory is based on an instrument which was previously developed for the social sciences and optimizes it by reducing the number of items per dimension improving at the same time its psychometric properties, but also includes new dimensions – Academia Driven and Society-driven – that reflect the greater influence of societal and organizational structures in knowledge production and demands for research impact.

A recent quantitative study elaborated an inventory to identify factors influencing the research agendas of researchers (Horta & Santos, 2016). The inventory was designated as “Multi-Dimensional Research Agenda Inventory” (MDRAI), and while this inventory is the first of its kind to the best of our knowledge, it was designed having in mind researchers performing research in the fields of the social sciences. Our study departs from MDRAI and using a dataset of over 4,500 researchers located all over the world and from STEM fields who provided key information concerning their research agendas through participating on an online survey implemented in 2017 and 2018, will contribute to develop a novel instrument that identifies key factors influencing the research agendas of researchers in STEM fields.

Methods Using a global sample of 4826 researchers from STEM fields, Confirmatory Factor Analysis was employed to develop an instrument capable of assessing STEM researchers’ research agendas. Validity, Reliability, and Sensibility were evaluated as part of the instrument validation exercise. In terms of methodological innovation, the analyses use alternative measurements and analytical tools to cope with the large sample size. Multi-group analysis was conducted to demonstrate measurement invariance across all fields of knowledge.

Results The final instrument is comprised of 40 questions spread over 8 distinct dimensions – Scientific Ambition, Divergence, Collaboration, Mentor Influence, Tolerance to Low Funding, Discovery, Academia Driven, and Society Driven. Model fit was adjudged as very good (GFI = 0.944; CFI = 0.951; PCFI = 0.849; RMSEA = 0.037). Factorial and discriminant validity were confirmed, and convergent validity was largely demonstrated with only minor issues to report. Reliability and Sensitivity were confirmed. No notable differences were found between Exact and Natural Sciences and Engineering and Technology fields.

Discussion Eight distinct dimensions were identified by the instrument. The dimension Scientific Ambition stresses the relevance of engaging in research agendas that may provide recognition for one’s work from peers and help to achieve positions of intellectual and field authority in the field communities of interest (Latour and Woolgar, 2013; Whitley, 2000). Similarly, to Scientific ambition, Collaboration is another dimension is also of critical importance in the MDRAI-R-STEM as it underlines the understanding that collaborative agendas are necessary in all fields of knowledge, and that collaborating or not with peers is seen as a key decision when embarking in new research agendas. The dimensions Tolerance to Low Funding and Mentor Influence, also appear to be critical dimensions in influencing the research agendas of researchers. Greater scores in Tolerance to Low Funding indicates that researchers are not discouraged from lack of available research funding to start specific research agendas, meaning that they do not place an emphasis on research funding when deciding on pursuing a research agenda. A larger score in the Mentor’s influence suggests that the PhD supervisor continues to have a say or a degree of influence in a researcher research agenda, while the opposite means that the researcher embarks in research agendas without requesting the PhD supervisor guidance or opinion. The dimension Discovery in the MDRAI-R-STEM encapsulates the MDRAI dimensions Discovery and Conservative. A greater the score on the Discovery dimension suggests that a researcher is more propense to engage in research agendas that are riskier, focused on emerging and unexplored themes that have a greater potential for breakthroughs but also for failure. The dimension Divergence, maintained the same structure that it has in the MDRAI, including its sub-dimensions (branching-out and multidisciplinary) but in a similar fashion to the Discovery dimension, it also summarizes the MDRAI dimensions Divergence and Convergence, into a sole dimension having the two MDRAI independent dimensions placed into a one-dimension continuum. A higher score in the Divergence dimension means that researchers establish research agendas that link and involve knowledge from other fields of knowledge. The first new dimension of MDRAI-R-STEM is Academia Driven, which means the extent to which the research agenda is influenced by holistic, valuative and normative traits and dispositions related to the scholarly and academic environment and social structure, with whom the researcher identifies himself or herself with. The higher the score in this dimension, the more the researcher research agenda conforms to and is aligned with the questions, topics and strategic focuses that the academic environment may determine as a priority. This dimension has two sub-dimensions. The sub-dimension Field relates to the extent to which the researcher is influenced in his or her research agenda by scientific priorities that are reached by consensus in the field community as those to be prioritized (Becher and Trowler, 2001). A higher score on this sub-dimension means that the research agendas of researchers will be more influenced by the field community priority focus. The other sub-dimension, Institution, is associated to propensity to which the researcher aligns his or her research agendas to the research strategic targets of the institution where he or she is working. The higher the score on this sub-dimension the higher that propensity will tend to be, while the lower the more the research agenda of an individual researcher will be affected by institutional constrains. The second new dimension of MDRAI-R-STEM is Society driven, which measures the propensity that the research agenda is set towards solving society related challenges. The higher the score on this dimension, the greater it is the focus of the research agenda on focusing on solving such challenges, while the lower the score, the less is the focus on a society related challenge. This dimension has two sub-dimensions. The first sub-dimension is society, which shows the incidence of society related challenges in the research agenda, while the second sub-dimension, non-academics, measures the degree of auscultation and participation of laymen and non-experts in the design of the researcher research agenda. The higher the score on this sub-dimension the greater is the likelihood of engagement with non-research communities in forms of research that involve “action research community” or “participatory research”. Noticeably, the two sub-dimensions indicate that one can have a society focused research agenda without needing to collaborate with non-expert communities.

References: Becher, T., & Trowler, P. (2001). Academic tribes and territories. Buckingham: SRHE. Horta, H., & Santos, J. M. (2016). An instrument to measure individuals’ research agenda setting: the multi-dimensional research agendas inventory. Scientometrics, 108(3), 1243-1265. Kuhn, T. S. (2012). The structure of scientific revolutions. University of Chicago press. Latour, B., & Woolgar, S. (2013). Laboratory life: The construction of scientific facts. Princeton University Press. Santos, J. M., & Horta, H. (2018). The research agenda setting of higher education researchers. Higher Education, 76(4), 649-668. Whitley, R. (2000). The intellectual and social organization of the sciences. Oxford University Press on Demand.

ABSTRACT. Digital transition poses numerous challenges for decision-makers. Despite some public interventions are inevitable, it is business that will play a crucial role in transforming markets, innovation systems, and technological development. So, there is a growing need in a new market and industrial policies. This problem is especially important for the emerging economies, where existing institutions of developmental state and traditional industrial policy seem to be less relevant – if not obsolete – for the new digital realities. In this context, an “agency” approach is applied, where private and sometimes public business entities are de-facto assigned to a role of agents for development, responsible for technological breakthroughs, emergence of new innovation ecosystems and institutes, etc. In this study we focus on comparative analysis of Russian and Chinese digital policies, with a special focus on issues related to agency approach. In China we may trace the evolution of public policies toward I&IT up to the beginning of 2000s. Some scholars tend to explain early successes of Chinese I&IT by the lack of systematized regulation resulted in a de-facto “sandbox” regime. This is well illustrated by a unique situation with foreign investments in the II&T: de-jure they were restricted because of “strategic” nature of these industries and markets, but de-facto actual inflows of foreign capital through a specific instrument of Variable Interest Entities (VIEs) were ignored by the regulators – or even welcomed. In a favorable economic conditions (sharp growth of demand; synergies between Internet markets and electronics manufacturing; etc.) this spurred fast rise of Chinese I&IT markets, lead by so-called BAT group (Baidu, Alibaba, Tencent). I&IT successes spurred governmental attention. In line with previous industrial policy originally stake was made on the state-owned telecomm enterprises (SOEs) and only after they failed to lead the markets, BAT appeared on the regulator`s radar as agents for I&IT. New policies were a mix of updated traditional and new approaches. Classic industrial efforts persisted, like Internet infrastructure build-up, with archaic solutions also in place – like blocking of Facebook, Google, and other competitors since 2010, or de-facto ignoring BAT monopolism. But gradually a different approach evolved. Huge public investments were made in digital R&D and in venture capital (VC), as well as pro-business institutional and regulatory moves – all with strong bias toward BAT. BAT were engaged in and benefited from nation-wide technology and trade and investment initiatives. As a result, BAT transformed itself in a multi-billion market phenomena with colossal VC activities, notable R&D budgets ($2-3 bln each), diversified set of services and important technological portfolios (from AI to fintech) – with strong and growing innovation ecosystems. However, existing PRC policies bear challenges. Among them are monopolism; market distortion risks due to the governmental interventions; ambiguous approach toward the Internet; need to optimize enormous and fast-growing BAT ecosystems and investments; etc. For now these challenges are softened by market growth, enormous financial resources of BAT, and preferential state policy. But this will not last forever. Russian digital policies are partly similar, partly opposite to a Chinese case. Market conditions are suboptimal due to structural and institutional limitations for economic growth – sharpened by sanctions. Innovation system is fragmented, with lack of high-tech private civilian industries, especially electronics. Despite since 1990s Russian commercial IT/Internet markets and companies were on the rise, it was only in 2015-2017 when government introduced more or less systematic efforts in support of disruptive I&IT technological and market development. The resources allocated or authorized were significant, but not even close to the ones of the PRC or to the real needs of Russian economy. For example, in 2019-2024 government plans to spend about $25 bln for “Digital Economy” National Project – with additional $9 bln from the industry, mostly SOEs. Not less important, federal policies is medium-term and project-oriented with lack of strategic focus – like almost ignoring insufficient capital availability for tech-savvy SMEs. Regulatory approaches are also mixed. Most radical industrial policy instruments (like blocking of competitors) were never implemented, but “technology nationalism” (requirement to localize data centers, software import substitution, “sovereign” Internet ideas) is visible. Bureaucratic interventions are on the rise too, contradicting even with own plans of the Central Bank and some Ministries to create regulatory sandboxes for fintech and other II&T. Special attention is paid to competences (from “Quantoriums” - science parks for kids, to nation-wide coordination of educational and training activities by the Agency for Strategic Initiatives (ASI), affiliated with the President) and ecosystem building (like in National Technology Initiative). However, efforts are underfinanced and challenged with regulatory and bureaucratic drawbacks. As in PRC, agents for development play crucial role in Russian policy. However, the situation is ambivalent due to unclear lines between corporate and federal responsibilities, and federal interventionalist actions. Secondly, despite strong positions of private I&IT companies like Yandex, agent group included much broader set of actors – presumably, because of lobbying, some economic considerations, and past successes of potential agents. Along with private I&IT giants the list is dominated by SOEs – Sberbank (one of biggest and highly digitalized Russian bank, headed by I&IT enthusiast G.Gref), Rostelecom, dual-use manufacturing companies like Rosatom and Rostec, etc. On one hand, this “mix” represented objective complexity and ubiquitous nature of I&IT and need for economy-wide and industry-immune digital policies. On the other, it seem to weaken general logic, since (in accordance with C.Christensen`s theory) traditional corporations may be interested not in the disruptive innovations, but in maintaining control over the markets and economy. These SOE agents became key beneficiaries – and partly drivers – for new extensive governmental efforts, but, in turn were assigned with some additional functions, from competence-building and up to broadening VC support, which do not match their actual abilities, processes, and interests – with unclear results. This ambivalence, among other factors, lead to a surprising rise of importance of regional policies, especially ones of Moscow and Kazan, which along with municipal needs are focused on new markets, SME development, and more. Comparison of the PRC and Russian and general conceptualization of agent approach for I&IT make it possible to formulate several conclusions. Assigning developmental functions to “agents” seem to be more optimal for I&IT, than direct governmental interventions. However, selection of an agent and overall efficiency appear to be sound only if based on its market successes, and presumably has time limits. Agency does not substitute (and rather actualize) governmental efforts in institutions- and human capital formation, while requiring reassessment of public sector role, with a gradual shift to a strategic collaboration and coordination model, ideated by D.Rodrick. This appear to be a challenge for emerging economies, considering industrial policy path dependence, restraining evolution toward a more balanced system, immune from both excessive government interventions and rent-seeking / monopolistic and other market distortions on the agents` side. A separate conclusion may be formulated considering possible combination of agency and regional development logic. Regions or, more likely, megacities – leaders in II&T, with its more flexible regulatory practices, possible cluster activities/effects, market specifics, and other advantages may complement efforts of agents, driving changes from the demand side, and acting as a kind of “sandboxes” for both I&IT and industrial policy.

ABSTRACT. The emergence of data-driven innovation based on the rapid advancement in the Internet of Things (IoT) and artificial intelligence creates exciting opportunities as well as considerable challenges in promoting societal benefits while regulating the risks associated with it. As a vast amount of diverse kinds of data is increasingly available from various sources that were not previously accessible, a wide range of sectors are currently undergoing significant transformation. In energy, smart grid systems lower costs, integrate renewable energies, and balance loads. In transportation, dynamic congestion-charging systems adjust traffic flows and offer incentives to use park-and-ride schemes, depending upon real-time traffic levels and air quality, whereas car-to-car communication can manage traffic to minimize transit times and emissions and eliminate road deaths from collisions. The speed of technological advancement is accelerating, and those technologies that used to be separate are increasingly interconnected and interdependent with one another, creating a significant degree of uncertainty in their impacts and consequences. There is a widening gap between technological and institutional changes, which poses a serious challenge to policy makers in devising measures for governing innovation.

Data-driven innovation critically depends on efficient and effective collection, diffusion, and utilization of data. The development of smart cities is hence facilitated through readily availability of and accessibility to data and its mutual exchange with stakeholders in different sectors. Unlike the traditional mode of innovation, which tends to rely on closed, well-established relationships between enterprises in a specific sector, the new mode of innovation requires open, dynamic interactions with stakeholders possessing various kinds of data. Close cooperation and collaboration on data becomes crucial in the innovation process, from the development of novel technologies to deployment through field experimentation and to legitimation in society. Careful investigation is necessary to explore what kinds of policy approaches would be effective in stimulating data-driven innovation by facilitating coordination and integration of emerging technologies through open exchange and sharing of data.

The approach of regulatory sandbox has recently been proposed to stimulate innovation by allowing experimental trials of novel technologies and systems that cannot currently operate under the existing regulations by specifically designating geographical areas or sectoral domains. Regulatory sandboxes enable firms to test innovative products, services, and business models in an actual market environment, while ensuring that appropriate safeguards are in place. Potential benefits include facilitating greater data availability, accessibility, and usability for innovators and reducing the time and cost of getting innovative ideas to market by reducing regulatory constraints and ambiguities. While regulatory sandboxes have primarily been introduced to the financial sector to encourage innovation on fintech, it is only recent that the approach started to be applied in cyber-physical systems such as smart cities, which involve human health and safety. There are few empirical studies conducted to investigate to what processes the approach of regulatory sandbox is introduced, by what measures it is implemented, and what impacts and consequences are expected in the context of smart cities.

This study aims to examine policy measures to stimulate data-driven innovation for smart cities with a focus on coordination and collaboration in the context of accelerating technological progress and deepening interconnection and interdependence. Detailed case studies of Japan and China are conducted to investigate how the idea of regulatory sandboxes is introduced and applied in policy making and what impacts and consequences are made on facilitating data exchange and sharing to create innovation for smart cities. Opportunities and challenges in policy design and implementation are explored for stimulating data-driven innovation through compatibility, interoperability, and integration of data while addressing societal concerns about privacy, cyber security, and public safety.

A preliminary analysis of the Japanese case shows the characteristics of the process of creating innovation for smart cities and illuminate some of the key policy challenges in this field. In Japan, proprietary standards among competing providers initially slowed down the development of smart technologies. Eventually the standard of Open Automated Demand Response 2.0 was adopted, following feasibility, interoperability and connectivity testing through close collaboration between major players in the energy industry and the government. The adoption of the common scheme had a significant impact on driving innovation on home and building energy management systems, with smart thermostats, rooftop solar, battery energy storage, and other appliances smoothly integrated through data exchange and sharing.

Novel technologies have been emerging, however, in other sectors related to the Internet of Things, in which virtually everything is getting connected so that many technologies that were formerly operated separately can coordinate with each other efficiently. Various standards, such as ZigBee and Bluetooth Low Energy, have been under rapid advancement, leading to a serious need to promote cooperation among relevant actors working in different sectors. The traditional system of government-industry collaboration did not function effectively, as new players including start-ups and entrepreneurs have been entering the field of smart cities with innovative technologies that are often developed independently. That has posed a serious challenge to policy making, as different sectors have their specific issues and concerns and the policies and regulations to deal with them are not necessarily coordinated with each other, which effectively creates an obstacle to facilitating data exchange and sharing for innovation in integrated smart cities.

Demonstration of smart cities in various parts of the country has become critical for testing promising technologies and raising awareness among the general public. They were mainly aimed at verifying emerging advanced technologies concerning smart cities, including cogeneration, renewable energy, energy storage, electric vehicles, and energy management systems. At the same time, these projects were also aimed at developing robust business models with close collaboration among relevant stakeholders, including local communities and residents, as well as technology providers in various sectors. The way in which the demonstration projects were implemented were locally adjusted, considering the specificities of the economic and social conditions and contexts, with the effect of promoting technological integration, reliability, and learning through trial and error. Various types of new promising technologies were verified, adopted, and integrated, including facilities for renewable energy, energy storage batteries, and energy management systems, effectively leading to a decline in the prices of component technologies and the costs of operating energy systems. These projects were especially important in providing collaborative platforms in which novel technological functionalities were tried out. The tightly knit groups involved in the projects contributed to producing valuable data, facilitating the sharing of that data among the stakeholders involved.

At the same time, there still remain several challenges in facilitating data-driven innovation with the regulatory sandbox approach. While large established companies tend to have advanced technological expertise and capabilities concerning various instruments and facilities in smart cities, local governments and communities do not necessarily possess sufficient knowledge or experience of dealing with technical measures. Under the existence of the significant degree of asymmetry of data and knowledge between large technology companies on the one side, and local government and communities on the other side, we need to consider how it would be possible to secure serious and active participation of end users in an equal and equitable manner for jointly facilitating innovation. Robust business models are also currently missing, which has the effect of discouraging private companies from taking over the demonstration projects that have mainly been financed by the public sector.

ABSTRACT. In a nutshell, the project is on the adoption of technologies like robotics, expert systems, and other specialized technologies like familial DNA matching. We will estimate adoption functions for multiple smart technologies using panel data from a range of agencies in the United States.

The oncoming “second machine age”, represented through the widespread adoption and use of smart technologies such as artificial intelligence and robotics, will fundamentally change society and governments at all levels. At this point, we are especially interested in this question: “How do public organizations prepare and adopt smart work provision for future works?”

We have addressed some aspects of this coming change in a recent paper in the Journal of Public Administration Research and Theory on machine learning for public administration. In that paper, we offered a full description of a basic technology infrastructure that underlies broader applications like artificial intelligence, and then showed how machine learning and affiliated methods can change how we learn about the world of public administration.

Yet, we believe the practice of public administration is not the same as the study of the practices of governments. While we see natural applications for research, and while researchers are often on the cutting edge of adopting new learning technologies like machine learning, agencies live in different worlds that both enable and constrain the technologies they adopt and use – and that also affect the likelihood of their success.

For instance, while our focus is naturally on artificial intelligence, and while the public has focused on technologies like robotics, agencies have very specific technology needs. Those needs will include smart technologies that exist largely outside the eye of the public. One example is advanced DNA matching technologies, which some agencies will find to be more important than robotics or expert systems like artificial intelligence.

Our project addresses this competition among and across smart technologies for attention by governments and adoption by agencies. Learning about the adoption process for technologies is best done by comparing how different technologies are selected by a pool of eligible agencies. In that context, we want to understand the factors that contribute to the adoption of smart technologies over time.

Our data for this project are from two waves of panel data asking about adoption of a half-dozen different smart technologies by the same set of agencies at multiple levels of government. Specifically, we assess adoption by publicly-funded forensic crime laboratories in the United States. Our data are from 2009 and 2014 – the most recent data available (the census occurs every five years). These laboratories are managed at multiple levels of government, including the local, regional, state, and federal levels.

The advantages of this research approach are that we can assess both longitudinal and cross-sectional adoption rates. We can also address the impact of many other aspects of lab management, including certification and professionalization. Finally, we can observe the use of multiple smart technologies – including those that may be even more important to these agencies (such as familial DNA matching).

ABSTRACT. The main aim of innovation policy is to support experimentation with new technologies, products, processes, or business models, and accelerate its diffusion throughout the economy and society. Yet innovation policy per se is not very experimental. Policymakers invest billions funding many scientific and business experiments, but they rarely experiment themselves with their own programs and activities, at least in a structured way.

Innovation policymakers face a complex and continuously evolving system and have very limited evidence on how most effectively to influence it. One alternative to navigate all these unknowns and shed some light on the possible answers is to become more experimental. That is, exploring a wide range of ideas, testing out the most promising ones at small scale, learning which are likely to work better, and only then scaling them up.

This would mean turning the current model of policymaking upside down. Despite all the unknowns, governments often act as if they had all the answers, rather than recognizing that they do not. They introduce new policies without prior small-scale testing, assuming they have chosen the best design and hoping it will work.

Experimental approaches are increasingly being adopted across many policy fields, but innovation policy has been lagging. This paper reviews the case for policy experimentation in this field, describes the different types of experiments that can be undertaken, discusses some of the unique challenges to the use of experimental approaches in innovation policy, and summarizes some of the emerging lessons, with a focus on randomized experiments.

What does it mean to be experimental?

The word “experiment” is often used in many different ways, so it is useful to clarify what the meaning of an experiment actually is. In short, an experiment is a test. More specifically, the Cambridge English Dictionary defines experiment as “a test done in order to learn something or to discover if something works or is true”.

This definition captures the key characteristic of a policy experiment: learning. It is intentionally set up to learn. It has a clearly structured learning strategy, defined ex-ante rather than as an after-thought, and generates new information, evidence or data. Therefore, a government pilot “trying something new” is not a policy experiment, unless the systems and processes required to learn from it are also put in place. This includes a timeframe with clear limits or checkpoints: there is date at which you assess the results and decide whether to continue the experiment, tweak it, scale it up, or discontinue it.

Policy experiments can be used in different contexts and with different objectives. They can be divided into two groups: those that are focused on exploration and discovery (understanding how the world works), and those framed around evaluation (finding out what works). Within the first group, experiments can be used to test assumptions about the problem to be fixed, the underlying drivers of behaviors, or the solution being considered. Alternatively, experiments can also be used to explore the feasibility and potential of a new intervention. The second group of policy experiments are focused on evaluation, although from two different perspectives: impact evaluations that estimate the ultimate impact of an intervention on outcomes, and process optimization experiments that measure intermediate impacts of changes in the process.

Experiments are at the core of policy experimentation, but the process of experimentation involves other important steps. It starts with understanding the problem, creatively exploring unobvious ideas, and developing hypotheses and potential solutions that can be tested. It does not end when the results of the test become available. Instead, governments that have successfully embraced a culture of experimentation not only set up experiments, but they also make sure the resulting learning and evidence is used in decision-making, scaling-up successful ideas while continuing to iterate and experiment.

Experimentation in innovation policy

Innovation policy can itself be conceived as a continuous learning and discovery process about new technologies, the inner workings of the innovation system, and the effectiveness of programs and policies that seek to influence it.

The four types of experiments in examined in the paper play a role in this process. Innovation experiments can be used to understand how different types of innovation processes or methods actually work (Boudreau and Lakhani, 2016), which can also generate useful insights that inform the design of new programs and policies. Alternatively, experiments can also be framed around specific policy challenges, and be used to explore solutions that contribute to address them. Finally, they can also be used with a program evaluation mindset, to test whether a program works and how it can be improved.

The case for experimentation is reinforced by the complexity of the system that innovation policymakers try to influence, a very dynamic context that continuously evolves (with new challenges and opportunities regularly emerging), and high levels of uncertainty (in terms of policy levers and potential interactions, returns on investment from programs, or future scenarios among many others).

There is a range of methods that can be used to learn from policy experiments, including randomized controlled trials (RCTs), A/B testing, rapid cycle testing, ethnographic research, human-centered design or mixed methods (among many other qualitative and quantitative approaches). Ultimately, the choice of method relies on the question being asked and the context in which the experiment is taking place, which determine what is feasible and desirable.

The innovation policy questions that randomized trials can(not) address

Running randomized trials on innovation policy questions can be more difficult than in other fields for several reasons. First, the outcomes of innovation policies are not always easy to measure. A second challenge is that outcomes can take longer to become visible than in other fields. Third, innovation outcomes can be very skewed. Fourth, innovation ecosystems are complex environments. Lastly, many innovation policy challenges are multidimensional, and so is the solution space.

None of the challenges above are insurmountable. How easy it is to address them depends on the policy being considered and the aim of the experiment (i.e., impact evaluation, process optimization or mechanism experiments). In some cases, the compromises required may make the use of randomized trials unfeasible or undesirable, while in others trials can add substantial value.

The menu of innovation policies is wide, so the paper considers what type of policies and what type of questions are amenable to different types of randomized experiments, and showcases some of the emerging evidence from trials on the field. It concludes by exploring some of the key barriers to policy experimentation in this field, and some strategies to overcome them.

ABSTRACT. There has of late been renewed interest , in the field of ‘Science of Science and Innovation’ (SciSIP) or ‘meta-research’ (evidenced for example the Methods in Research on Research (MiRoR) initiative) – and some new funding opportunities for a field of science that has traditionally been somewhat of an orphan area . How meta-research is funded, supported, incentivised, published and generally communicated shapes what is available, discoverable and known by science policy practitioners to use and build upon. This panel aims to bring diverse perspectives on how knowledge is exchanged between research and practice in science policy. How do SciSIP researchers and practitioners interact? Through which channels is meta-research shared across social scientists, consultants and policy makers? To which extent expertise on either academia or practice is useful, forgotten or lost in translation? A particular aspect to be explored is whether new types of publishing platforms/scholarly infrastructure can help to make research on research more robust, publicly available, discoverable and shared. Many of the issues around the reproducibility in science and the causes of research waste apply also to science policy-related research; how can we ensure maximum value and use of research on research? While there are some well-established (but often fragmented?) academic pockets of ‘meta-research’ across the globe that actively contribute to science and innovation policy, there is a wealth of other related-research that is relevant to science policy but that is conducted in non-traditional academic settings (often conducted from within, or commissioned by, a funding agency or a research institution) or as a secondary academic focus for many researchers. Furthermore, policy-related research (often termed ‘grey literature’) has traditionally been locked away in hard to find and access reports, stored on internal organisation websites, and in formats that make the content hard to fully access and use. To be useful, such research needs to be both discoverable and, where applicable, trusted. There have been a number of recent innovations in aspects of the research delivery ecosystem, for example with some funding agency experimentation with research funding models, and particularly in how research is shared and ‘published’ for others to access and use. Initiatives like ORCID and Crossref are working to improve the availability of researcher and research-related meta-data and infrastructure and help to provide the evidence-base on which experiments work and which don’t. Artificial Intelligence (AI) is likely to greatly enhance the way we discover and access research of all types. Robust research on research might be (or already is?) a critical component of an effective science and innovation system in the future. This panel session aims to present examples of how research on research can be more integrated and linked to science policy practice from a variety of perspectives.

We propose that the session would be Chaired by Ismael Ràfols (Ingenio (CSIC-UPV, Universitat Politènica de València), visiting professor at CWTS (University of Leiden) and adjunct faculty at SPRU (Science Policy Research Unit at the University of Sussex). The speakers would include: 1. Liz Allen Director of Strategic Initiatives F1000/Visiting Senior Research Fellow, Policy Institute at King’s College London) (confirmed) – focusing on new approaches in scholarly communication/publishing that support sharing and access to research findings and insights, and beyond conventional article formats 2. Cassidy Sugimoto Professor University of Bloomington, Alabama/SciSIP programme lead at NSF (confirmed) – focusing on how funding agencies can help to bridge the divide between science research and policy and help to ensure the maximum reach and impact of the products of research (tbc) 3. Erik Arnold Chairman of the Technopolis Group /Adjunct Professor in research policy at the Royal Institute of Technology (KTH), Stockholm (confirmed) – use of science policy research in consultancy and policy (tbc) 4. Barry Bozeman. Director, Center for Organization Research and Design, and Arizona Centennial Professor of Technology Policy and Public Management.