Mapping trending topics and leading producers in innovation policy research

Introduction

Over the last few decades, public administrations and the scientific community have increased their focus on innovation research (Martin, 2012; OECD, 2015). The main reason is that research and innovation enhance competitiveness, boost growth and create jobs, whilst improving social well-being, healthcare, transport and digital services and providing countless new products and services (European Commission, 2014). Moreover, innovative economies are more productive, are more resilient, adapt better to change and are able to support higher living standards (Asheim and Moodysson, 2017).

Public administrations are investing heavily in innovation activities and are promoting firms innovation to improve the business environment through the design, implementation and evaluation of innovation policies (Mazzucato and Semieniuk, 2017). Meanwhile, the scientific literature on innovation research has grown substantially in the last few decades. Its rate of growth has outstripped that of other research areas, which suggests that academics from multiple disciplines are investigating how innovation activity and innovation processes influence economies and societies (Cancino et al., 2017). This body of innovation research includes investigations into innovation policy (e.g. Flanagan and Uyarra, 2016; Uyarra and Ramlogan, 2016; Coenen et al., 2017; Edler and Fagerberg, 2017), analyses of innovations addressing important social and global issues such as resource scarcity and climate change (e.g., Ghisellini et al., 2016; Loiseau et al., 2016), and assessments of innovation policies deployed in different regions and countries (e.g. Borrás and Jordana, 2016; Fu et al., 2016; Isaksen et al., 2017).

Innovation policy has emerged as a new field of economic policy in the last few decades. The term is now commonly used given the widespread view that policy plays a role in supporting innovation. Innovation policy attempts to influence innovation activity, often to increase economic growth (Fagerberg, 2017). The origins of the term lie in the intellectual environment that developed around the Science Policy Research Unit at the University of Sussex from the late 1960s onwards (Fagerberg et al., 2011). In particular, Professor Roy Rothwell, of the Unit, did much during the 1980s to increase interest in the topic (Rothwell, 1982). However, the real surge of interest occurred in the 1990s, when national governments and international organisations, such as the OECD, started to pay attention to this phenomenon.

In view of this background, the main objective of this article is to analyse innovation policy research using bibliometrics and focusing primarily on authors and journals. There have been several recent bibliometric studies of innovation focused on the leading countries, regions and universities (Merigó et al., 2016) or the most relevant authors (Cancino et al., 2017). However, to the best of our knowledge, no bibliometric studies have focused specifically on innovation policy research. For innovation policy scholars, the number of academic studies of this topic can obscure a general picture in the search for information. We therefore focus on two research questions:

RQ1. What are the most relevant research themes within innovation policy research?
RQ2. Which authors and journals produce the most innovation policy research?

To answer these questions, we use bibliometric methods such as performance analysis and science mapping (Cobo et al., 2011). The rest of the article is structured as follows: the bibliometric methods and data used in the paper are introduced, the results are presented, and the paper concludes with a summary of the main findings.

Bibliometric methods and data

The analysis presented in this paper uses bibliometrics combining two main procedures: performance analysis and science mapping (Cobo et al., 2011). Bibliometric performance analysis uses a wide range of indicators and techniques. These include total number of studies, total number of citations, publications by country, university or author, the h-index, and word frequency analysis (López-Rubio et al., 2020). The total number of studies is an indicator of absolute productivity. The total number of citations is also an absolute measure. It does not consider the study lifetime (i.e., publication year). The h-index combines the total number of studies and total number of citations into a single measure. If, for a given set of studies, N studies have received at least N citations, then the h-index for that set of studies will be N (Hirsch, 2005).

The h-index also has limitations. For example, a researcher with three heavily cited publications would have the same h-index as a researcher with three publications with only three citations each. Therefore, this indicator does not reflect well on researchers who have highly cited publications but moderate productivity (Egghe, 2006). Other relative bibliometric indicators include the ratio of citations per study, which favours authors with few articles but a high number of citations, and the ratio of citations per year, which accounts for the study publication date (publication lifetime). Both ratios are proxy variables of efficiency.

Overall, a higher number of publications or citations does not necessarily indicate higher research quality. Historically, there has been much discussion about which indicators should be used to accurately measure scientific production and influence. In recent years, the concept of responsible research and innovation has gained currency as a framework for research governance. Building on this, Wilsdon et al. (2015) proposes the notion of responsible metrics as a way of framing appropriate uses of quantitative indicators in the governance, management and assessment of research. Furthermore, bibliometric indicators address only one of many aspects of research. Bibliometrics should not be used for evaluation without reference to more in-depth qualitative assessments (Hicks et al., 2015). The present analysis covers all the above-mentioned indicators to provide a comprehensive, multi-faceted overview, and to overcome the limitations of each individual indicator (Mingers and Leydesdorff, 2015).

Science mapping is and interdisciplinary field originated in information science and technology. Science mapping is the development and application of computational techniques to the visualization, analysis, and modelling of a broad range of scientific and technological activities as a whole. This is an interdisciplinary field emerging from traditional library information science in the areas of bibliometrics, citation analysis, and computer sciences in the areas of visualization, visual analytics, data mining and knowledge discovery (Cobo et al., 2011). Bibliometric maps are based on the quantitative analysis of bibliographic data. Therefore, bibliometric mapping can be used to monitor a scientific field to determine its cognitive structure, evolution, and main actors and to visualize the results for specific bibliometric indicators (Noyons et al., 1999). Various software packages can perform bibliographic mapping (Cobo et al., 2011). We use VOSviewer, a free software tool to produce bibliometric maps (Van Eck and Waltman, 2010).

The most common bibliometric mappings are based on bibliographic coupling, co-citations, co-authorship and keyword co-occurrence (Merigó et al., 2016). Bibliographic coupling measures the shared intellectual background amongst documents. A strength value is calculated between each document in the sample based on the references shared by the two documents (Kessler, 1963). The more shared references there are, the stronger the theoretical foundations shared by the two documents are assumed to be. Bibliographic coupling makes it possible to link documents with a similar research focus, thereby revealing the knowledge structure of a field (Jarneving, 2007). Co-citation analysis identifies the shared background of publications in a data set. Two documents are co-cited if one or more documents cite both articles (Small, 1973). The weight of a co-citation is based on the number of articles that co-cite the two documents. It reveals a network of cited documents rather than linking the documents in the data set (Garfield, 2001). Co-authorship identifies research collaboration networks based on the number of co-authored documents (Katz and Martin, 1997). Lastly, keyword co-occurrence identifies links amongst research topics in a particular field based on the frequency of co-occurrence of keywords in documents (Callon et al., 1983).

Data

Our data were gathered from the Web of Science Core Collection database, owned by Clarivate Analytics. This database is one of the most important sources of bibliometric information. It records consistent, standardised information (Adriaanse and Rensleigh, 2013; López-Rubio et al., in press-b). According to the OECD (2018), all types of R&D carried out or paid for by business enterprises are considered by definition innovation activities of those firms. Therefore, we executed the following search in October 2020: ‘innovation policy’ OR ‘innovation policies’ OR ‘policy of innovation’ OR ‘policies of innovation’ OR ‘R&D policy’ OR ‘R&D policies’ OR ‘policy of R&D’ OR ‘policies of R&D’. We executed the search for the fields title, abstract; and keywords (both author keywords and Keywords Plus generated by the Web of Science) from 1960 to 2019.

We excluded all document types except for articles, reviews, proceedings papers and early access articles. The search returned 2,929 records. Figure 1 depicts the distribution of these documents by research area. The research areas constitute a subject categorization scheme that allows to identify, retrieve and analyse these documents. Besides, journals and books covered by Web of Science are assigned to at least one category and each category is mapped to one research area. Figure 1 shows that innovation policy generally spans one or more research streams across the disciplines of business, public governance and sustainable development.

Results

This section reports the main bibliometric results for the 2,929 selected documents up to and including the year 2019. These documents consist of 2,255 articles, 704 proceedings papers, 76 reviews and 14 early access articles. Between 1960 and 2019, the number of citations received by this set of studies was 40,560, or 13.8 citations per study.

Figure 2 depicts the total number of publications and citations per year from 1982 to 2019. Between 1960 and 1981, only one study was published on innovation policy research. This study was published in 1969, and it received its first citation in 1983. However, after 1981, documents were published on this topic every year. This trend in publications has continued upwards reaching a maximum number of annual publications of 389 in 2019. Figure 2 also shows that these publications have received a high number of citations. The maximum was achieved in 2019, with 7,768 citations. The thresholds of 1,000, 2,000 and 4,000 annual citations were surpassed in 2010, 2012 and 2016, respectively.

The most cited innovation policy studies indexed in the Web of Science Core Collection.

To identify the most cited innovation policy studies, both in absolute and relative terms, we selected the 1% of studies with the most citations and the 1% with the highest number of citations per year. The Appendix, lists the thirty most cited studies on innovation policy research based on Web of Science Core Collection data. It also shows the thirty studies with the highest number of citations per year. Because of overlap between the two groups of studies, the table contains forty-two studies, ranked by total citations.

The first paper is Regional innovation systems: Institutional and organisational dimensions (Cooke et al., 1997). This paper suggests that the most severe problems related to the scale and complexity of the national innovation system are mitigated by a subnational focus such as that of regional systems of innovation. The second study, The learning region: Institutions, innovation and regional renewal (Morgan, 1997), focuses on an interactive model of innovation for regional development. The paper analyses European Union regional policy measures and studies the regional innovation strategy in Wales. The third study, One size fits all? Towards a differentiated regional innovation policy approach (Todtling and Trippl, 2005), examines different types of regions. The aim is to show that there is no ideal innovation policy model and that the type of policy depends on regional features.

The study with the most citations a year is Constructing regional advantage: platform policies based on related variety and differentiated knowledge bases (Asheim et al., 2011). This paper presents a regional innovation policy model based on the idea of constructing regional advantage. It categorises knowledge into analytical (science based), synthetic (engineering based) and symbolic (arts based), with different requirements of virtual and real proximity mixes. The implications of this research are the ability to track evolving platform policies. These platform policies facilitate economic development within and between regions. The action lines followed are suitable to incorporate the basic principles behind related variety and differentiated knowledge bases. The second document is the study by Todling and Trippl (2005) described previously. The third study, Strategic niche management and sustainable innovation journeys: theory, findings, research agenda, and policy (Schot and Geels, 2008), focuses on the strategic niche management approach. According to this approach, sustainable journeys can be facilitated by creating technological niches. These technological niches are protected spaces that allow experimentation with the co-evolution of technology, user practices and regulatory structures. Assuming that if such niches are constructed appropriately, they can act as building blocks for a broader social shift towards sustainable development, the empirical findings show that analysis of the internal dimensions of these niches must be complemented by attention to external processes.

To produce Figure 3, we used the VOSviewer overlay visualisation and the average publication year variable to depict the keyword co-occurrence map. We considered both author keywords and Keywords Plus and a minimum of two occurrences for the forty-two studies previously identified as the most cited papers on innovation policy, resulting in a total of forty-two keywords. The Web of Science Core Collection Keywords Plus field is created by an algorithm that provides expanded terms stemming from the record’s cited references or bibliography. The item colour indicates the average publication year. Table 1 presents the twenty most common keywords with the number of occurrences and co-occurrences and the average publication year.

We excluded the keywords innovation policy innovation, R&D and policy because these terms were part of the original search query. Besides these search terms, the most common keywords and their co-occurrences indicate that the main topics for authors and journals are based on three pillars:

1. Innovation systems, also including the keywords dynamics, systems and performance.
2. Science and knowledge, also including the keywords technology, networks, spill-overs and clusters.
3. Governance and sustainability transitions, also including policy mix.

The most productive and influential authors in innovation policy research

Table 3 lists the 34 authors with more than three publications and more than 200 citations in innovation policy research based on data indexed in the WoS Core Collection. The authors are ranked by total citations. The table also shows the total number of studies, the h-index of the documents included in the analysis (not all of a given author’s publications), the citations per study and each author’s current affiliation (institution and country).

The authors are from a wide range of institutions, most of them in Europe (31 authors of 34). The UK (8), Sweden (4), the Netherlands (4), Denmark (3) and Italy (3) have the strongest presence. Two of the authors are affiliated with institutions in the USA. One is affiliated with an institution in Taiwan.

Five authors have more than 1,200 citations each: P. Cooke, B.T. Asheim, B.A. Lundvall, E. Uyarra and L. Coenen. E. Uyarra has the highest h-index (13), indicating the best combination of productivity and influence, followed by L. Coenen (10), J. Edler and L. Klerkx (9), and C. Edquist and L. Georghiou (8). According to the data indexed in the Web of Science, E. Uyarra is also the most productive author with 16 publications, followed by L. Coenen (13), L. Klerkx and M. Mazzucato (11), and J. Edler (10). Finally, B.A. Lundvall leads the ranking of citations per study (376.3), followed by P. Cooke (346.0) and K. Morgan (265.5).

According to Cole and Cole (1973), there are four categories of scholars based on two dimensions: productivity and citations. Figure 4 depicts the total number of studies (productivity) versus the total number of citations (influence) of the top authors. The relationship between the number of studies and the number of citations is a representation of the ratio citations per study, which is a proxy variable of efficiency. Therefore, Figure 4 shows the efficiency diagram for top authors in innovation policy research. The axes were calibrated according to the average values of the total number of studies (6.9) and the total number of citations (596.9), respectively, in order to classify top authors in the following four quadrants:

1. Highly prolific. Scholars belong to this quadrant if both the number of studies and the number of citations are greater than the average.
2. Perfectionists. Scholars belong to this quadrant if the number of studies is lower than the average but the number of citations is greater than the average.
3. Mass producers. Scholars belong to this quadrant if the number of studies is greater than the average but the number of citations is lower than the average.
4. Less influential. Scholars belong to this quadrant if both the number of studies and the number of citations are below the average.

Based on Figure 4, E. Uyarra, L Coenen, B. T. Asheim, L. Klerkx, J. Edler, M. Trippl, C. Edquist, L. Georghiou, K. Flanagan and D. Czarnitzki are highly prolific. P. Cooke, B. A. Lundvall, K. Morgan and A. M. Hjalager are perfectionists. M. Mazzucato, R. Capello, J. M. Zabala-Iturriagagoitia, K. M. Weber, M. P. Hekkert, H. Hellsmark and S. Kuhlmann are mass producers. The remaining authors are less influential.

Another interesting issue is that of author co-citations. Author co-citation analysis shows the structure of connections between authors who are frequently cited together (White and Griffith, 1981). Co-citation analysis considers the references cited in the documents under study. This approach broadens the focus of the analysis because the cited documents may not be indexed in the Web of Science. Figure 5 presents the results of this analysis using a threshold of more than 100 citations and the 100 most representative links. Figure 5 corroborates the relevance of Lundvall, Edquist, Cooke and Asheim. However, this map also shows other important authors, such as the European Commission, the OECD, R.R. Nelson and C. Freeman.

Co-authorship analysis allows us to analyse research collaboration networks. Figure 6 depicts co-authorship between countries with at least thirty documents. The figure is based on VOSviewer network visualisation. In a network representation, the size of a node or label is proportional to that item’s relevance, and the network connections closely identify linked items. The location and colour of an item are determined by its cluster. Only the 100 strongest links are displayed. Table 3 presents the distribution of these twenty-seven countries according to the clusters generated by VOSviewer. Within each cluster, countries are ranked by total citations.

Clusters 3, 4, 5 and 6 are formed by European countries. Cluster 3 comprises Eastern European countries and Italy, Cluster 4 comprises Central European countries plus Romania, and Clusters 5 and 6 comprise Nordic European countries and Estonia. Cluster 1 is formed by Asian countries and the United States, France and Australia. Cluster 2 comprises several European countries, Canada and Brazil. These results show that clusters tend to be formed by countries that share the same or a similar language, culture or innovation policy. Sharing a language and/or culture should be expected to facilitate communication and understanding, whilst having similar innovation strategies and common goals promotes joint efforts.

The most productive and influential journals in innovation policy research

Table 4 presents the most productive journals in innovation policy research considering all the journal Web of Science categories and the fourteen journals with more than twenty studies indexed in the Web of Science. For each journal, Table 4 includes the total number of studies, the total number of citations, the h-index, the citations per study and the journal impact factor for the year 2019.

Research Policy is the leading journal by total number of studies, total number of citations, h-index and citations per study. This journal is followed by Technological Forecasting and Social Change, Regional Studies and European Planning Studies, depending on the indicator. Technological Forecasting and Social Change has the highest impact factor for the year 2019, with a value of 5.846. This journal is followed by Research Policy and Energy Policy. All these journals are high impact journals, which highlights the increasing importance of this research field in the last decade. This fact also shows that these journals have received a high number of citations and suggests that they are influential amongst the scientific community.

Based on the approach provided by Cole and Cole (1973), Figure 7 shows the efficiency diagram for top journals. The axes were calibrated according to the average values of total number of studies (61.9) and total citations (1,975.4), which allowed us to classify top journals into four categories: highly prolific, specialists, mass producers and less influential. Based on Figure 7, Research Policy and Technological Forecasting and Social Change are highly prolific journals, whereas Regional Studies is a specialist journal. Science and Public Policy and European Planning Studies are mass producers. The other journals are less influential.

The articles published in a given research area and journal tend to influence future studies submitted to that journal. Therefore, by analysing study citers, we can identify other journals focused on innovation policy research. We identified 25,703 citing studies from the Web of Science, excluding all document types except for articles, reviews, proceedings papers and early access articles.

Table 5 presents the ten journals with the most citing studies. Interestingly, one of the journals included in this ranking, Journal of Cleaner Production, does not appear in the list of the most productive and influential journals in Table 5. Journal of Cleaner Production is ranked fourth, with 565 citing studies. Sustainability occupies the second position in this ranking, with 790 citing studies. Journal of Cleaner Production is a cross-disciplinary journal that publishes research on cleaner production and environmental and sustainability research and practice. It is aimed at helping societies to become more sustainable. Sustainability focuses on the environmental, cultural, economic and social sustainability of human beings. These results highlight the increasing influence of sustainable development and transitions within innovation policy research in recent years (Schot and Geels, 2008; Kivimaa and Kern, 2016).

Discussion and conclusions

This study provides an overview of the most cited documents and the most relevant topics in innovation policy research, as well as of the authors and journals that produce the most research in this field. We used Web of Science Core Collection data from 1960 to 2019, excluding all document types except for articles, reviews, proceedings papers and early access articles.

There are other recent bibliometrics studies that carry out deep analyses on innovation research. Merigó et al. (2016) focuses on the most productive and influential countries in innovation research between 1989 and 2013 classifying the results in periods of five years. This study also analyses the research developed in several supranational regions. The leading journals in the field are also studied individually identifying the most productive countries in each of the journals. Cancino et al. (2017) analyses the most productive and influential authors in innovation research developed between 1989 and 2013 by calculating several author-level bibliometric indicators such as the total number of publications and citations, and the h-index. The results demonstrate that the most influential authors are not necessarily the most productive researchers. The study also analyses the most productive and influential authors in the leading journals in the field.

Our article provides diverse and significant contributions to the bibliometric literature on innovation compared with the aforementioned studies. First, we specifically focused on innovation policy within the innovation research field, and considered all years up to and including 2019. The principal Web of Science research areas covered by the documents under analysis and the keyword co-occurrence in the forty-two most influential studies show that the most relevant topics for authors and journals are based on three pillars: innovation systems and business, science and knowledge, and governance and sustainability transitions.

The concept of innovation systems originated between the end of the 1980s and the middle of the 1990s. The concept emerged in the context of the European industrial economies’ transformation into knowledge-based economies. A national innovation system consists of a network of economic agents together with the institutions and policies that influence these agents’ innovation behaviour and performance (Freeman, 1987; Lundvall, 1992; Nelson 1993). The idea of applying the national innovation system framework to a smaller geographical area (regional or even local) was later proposed by Cooke (1992).

The subsequent triple helix model (Etzkowitz and Leydesdorff, 2000) complements the innovation system approach. Whereas business is the central actor for innovation in the innovation systems model, in the Triple Helix model, innovation processes are considered to arise from multiple sources, namely the relationships between three main agents: the government (public administration), universities (science) and industry (business). This model has now evolved into the quadruple helix, the quintuple helix and even the sextuple helix model (government, university, industry, knowledge society, sustainability and entrepreneurship). The quadruple helix emphasises the importance of the knowledge society and knowledge democracy for knowledge production and innovation (Campbell et al., 2015), while the quintuple helix stresses the need for the socioecological transition of society and the economy to address, for example, global warming (Carayannis et al., 2012), and the sextuple helix adds entrepreneurship as a sixth dimension (López-Rubio et al., in press-a).

Secondly, P. Cooke has the highest total number of citations, followed by B.T. Asheim, B.A. Lundvall, E. Uyarra and L. Coenen. According to the data indexed in the Web of Science, E. Uyarra is the most productive author and has the highest h-index. J. Edler, L. Klerkx, C. Edquist, L. Georghiou, M. Mazzucato, K. Morgan, M. Trippl, K. Flanagan, D. Czarnitzki, and A. M. Hjalager are also ranked highly by total number of studies, total number of citations, h-index or citations per study. According to Cancino et al. (2017), the ranking of the most productive and influential authors in innovation research is dominated by U.S. authors. However, in the particular case of innovation policy research, the most influential authors are based at institutions in European countries. This spread may result from the strong focus of the European Union and the UK on innovation policies, especially since the 2008 global economic crisis, and is in line with the fact that the EU and the UK actively promote innovation processes and activities (Bergek et al., 2008).

For example, Horizon Europe (European Commission, 2018), the new EU innovation policy framework for the period 2021 to 2027 that will replace Horizon 2020, is based on three pillars: excellent science, global challenges and European industrial competitiveness, and innovative Europe. Furthermore, the authors’ co-citations corroborate the relevance of Lundvall, Edquist, Cooke and Asheim, and also shows other important authors, such as the European Commission, the OECD, R.R. Nelson and C. Freeman. These results highlight the relevance of the pioneering researchers in national innovation systems (Lundvall, Freeman and Nelson) and regional innovation systems (Cooke), as well as other authors and institutions that have also actively promoted the innovation system approach (e.g. Edquist, the OECD and the European Commission).

Third, the mapping of co-authorship across countries based on authors’ affiliations suggests that research collaboration networks usually include scholars working at institutions in countries with the same or a similar language, culture or innovation policy. A similar language or culture aids communication and understanding, whilst having similar innovation goals encourages joint efforts.

Fourth, Research Policy is the leading journal in terms of the total number of studies, total number of citations, h-index and citations per study. Technological Forecasting and Social Change heads the journal impact factor ranking for 2019. The Journal of Cleaner Production, one of the ten journals with the most studies citing innovation policy research, is not included in the innovation policy rankings. Sustainability is the second biggest citer of innovation policy studies. Both journals focus on sustainability research. Therefore, these results reflect the major role of sustainable development and growth, renewable and clean energy, smart specialisation, and the circular economy within innovation policy in the last few years (Staffas et al., 2013; McCann and Ortega-Argilés, 2015; Loiseau et al., 2016; Geissdoerfer et al., 2017; McDowall et al., 2017).

Finally, this analysis has some limitations. First, innovation policy research documents not indexed in the Web of Science were not included in the analysis. However, our study also includes co-citation analysis of authors to help overcome this limitation. For the co-citation analysis, the cited references did not need to be indexed in the Web of Science. Second, innovation policy studies indexed in the Web of Science but not contained in the selected research areas were not included in the analysis. Third, we used the full counting method, so each author was awarded one unit, regardless of whether the article had one or multiple authors. Fourth, in relation to journals, the citations obtained from the Web of Science data are not weighted. Therefore, all citations are considered equally, regardless of journal quality. Despite these limitations, this paper identifies the most cited and relevant documents and their main topics, as well as the most cited and influential authors and journals in innovation policy research. Furthermore, this paper shows that the combination of bibliometric performance analysis and bibliometric science mapping offers a tool for evaluators to complement qualitative analyses of a research field.

Acknowledgements

Norat Roig-Tierno wishes to thank Project [GV/2019/063], funded by the Generalitat Valenciana, and the “Agencia Estatal de Investigación y el Fondo de Desarrollo Regional (FEDER)” - RTI2018- 093791-B-C22 (MCI, AEI/FEDER, UE) for supporting this research.

About the authors

Pedro López-Rubio received his PhD in Economics from the Universitat Politècnica de València, Spain. His research is focused on innovation policy and models through quantitative and qualitative methodologies. He has presented papers at international conferences on innovation, technology transfer, knowledge management and regional development. He can be contacted at: pedloru@doctor.upv.es.
Norat Roig-Tierno is a Professor in the Department of Economics and Social Sciences at the Universitat Politècnica de València, Spain. His research is focused on innovation, regional development and the application of qualitative methodologies. He has presented numerous papers at international conferences. He can be contacted at: norat.roig@upv.es.
Alicia Mas-Tur is a Professor in the Management Department at the Universitat de València, Spain. She has presented numerous papers at international conferences. She can be contacted at: alicia.mas@uv.es.

How to cite this paper

Appendix: The thirty most cited studies on innovation policy research based on Web of Science Core Collection data