Mapping heterogeneity in a research system: The emergence of a ‘hybrid’ organizational field between science and industry

Abstract

This article analyses the diversification of research systems by focusing on the role of the new population of centres created by universities, public research organizations, and firms for the purpose of knowledge transfer and exchange. It departs from the mix of policies and the structural conditions in Spain which resulted in an organizational field frequently unnoticed in research evaluation exercises. An original methodology combining a map of centres and a survey has been used to observe their profiles, activities, and internal organization. This study highlights the role of hybrid research organizations in the system and the importance of institutionalization processes to shape their portfolio of activities and outcomes. In the conclusions, methodological and practical implications are provided for an understanding of the heterogeneity of research systems as a result of collaboration between academic science and industry.

1. Introduction

Research organizations are major components of the research system and undertake most R&D and innovation-related activities (Crow and Bozeman 2005; Jongbloed and Lepori 2015; Giannopoulou et al. 2019). In many countries they accumulate a strategic part of the knowledge reservoir, hold key infrastructures and employ a significant proportion of the workforce skilled in science, technology, and innovation. Because of their strategic position and specialized nature, they are subject to a range of influences. Public policies seek to influence their orientation in order to fulfil different economic and social purposes (Lepori et al. 2007; Cruz-Castro and Sanz-Menéndez 2015; Veletanlić and Sá 2018). Economic and policy agents make new demands when they need to solve specific problems for which science and innovation are necessary assets (Gulbrandsen 2011; Bruno et al. 2017). In particular, a trend that contributes to change and diversification of higher education institutions (HEIs) and public research organizations (PROs) is the push to increment the innovation capacities of the industrial base, as well as the demands from firms seeking infrastructures, personnel, and use-inspired and applied research.

New research organizations are emerging as the result of traditional HEIs and PROs adapting to getting involved in new arrangements, including institutional entrepreneurship, new spin-off centres, and joint ventures with other public or private bodies (Etzkowitz 2008). Others are new types of structures specifically oriented to fostering cross-sector research collaboration between academic science and industry. There are important issues to consider when dealing with these organizations in order to know their characteristics and evaluate their performance. For innovation policy purposes, it is difficult to outline the developments of existing centres and the emergence of new ones when identifying the relevant actors in a given research system. For research evaluation exercises, they are difficult to observe because of their heterogeneous mix of research fields, constituencies, and administrative forms. In particular, an important gap in current research is to disentangle the distinctive features of the new organizations, including the activities they perform, their strengths, and the major forces that shape their results.

This study is motivated by the importance of research organizations oriented to knowledge transfer to industry and formed by different initiatives of policy actors, including traditional HEIs and PROs, with the participation of the private sector. In some research systems, the emergence of new research organizations is a result of the ineffectiveness of existing structures to fulfil extended missions. For instance, bureaucratic arrangements in public research are a major cause of heterogeneity (Bozeman 1993; Bozeman and Bretschneider 1994). A lack of policy initiatives combined with bureaucratic rigidity in the presence of new demands from several actors may lead to the emergence of diverse organizations in a bottom-up process. New research centres are often the result of ways of operation that are difficult to implement in bureaucratized research systems because transforming and consolidating centres is a slow process.

Empirically, we focus on the emergence of hybrid research centres (HRCs) in Spain. Certain features of the Spanish research system are useful for studying this phenomenon and for comparative purposes. Most universities and research centres are public bureaucracies that struggle to accommodate private participation and a greater orientation to strategic research. A generation of new centres has emerged under the rationale of flexibility in order to meet requirements for research excellence or industry service,¹ though their significance is unclear because they are difficult to observe. Our research questions are as follows: what is the real role of these organizations in the research system? What are their characteristics and orientations? What factors contribute to shaping their activities and outcomes?

This article makes both methodological and analytical contributions. From a methodological perspective, we use a set of detailed procedures that can be useful for studying similar developments in other contexts. Our empirical approach combines a survey and an original map specifically designed to observe the population of centres. The map enables these organizations to be identified and classified according to their main constituencies. The survey to centre directors is suitable for observing the distinctive features of these centres, including funders, human resources, and organizational structure. From an analytical perspective, it offers a comprehensive portrait of a new organizational field emerging from previous organizational structures and policy frameworks. Moreover, it shows how institutionalization processes shape the activities and outcomes of these centres and generate heterogeneity in a research system.

The structure of this article is as follows. After this introduction, Section 2 discusses organizational change in research systems by focusing on hybridization and institutionalization processes. Section 3 briefly describes the policy and organizational context in Spain that has driven the emergence of a new organizational field. Section 4 explains the methodology for observing a new population of actors. Section 5 provides a detailed account of the organizational models and the institutionalization features through principal component and cluster analyses. The conclusions put forward implications for research evaluation, as well as for identifying relevant actors in the academic science–industry interface.

2. Organizational change in research systems

2.1 Heterogeneity and hybridization in research organizations

Research practice has dramatically changed in recent decades. Research communities and their organizations, both HEIs and PROs, have gradually adopted more complex management tools, such as Team Science, and new forms of funding and evaluations (Geuna 2001; Lepori 2011). Increased collaboration often entails new infrastructure and organizations for facilitating knowledge transfer and innovation, such as science and technology parks, liaison offices, firm incubators, university spin-offs, and HRCs. In this context, academic research organizations are also undergoing an important evolution.

Several processes entangled with the transformation of the R&D domain can be interpreted as drivers of change. A first driver of change comes from new policy initiatives that push organizations to produce relevant economic and societal results. Governments create new forms of funding and introduce management changes to obtain additional value from public investments. On occasion, they create new types of organizations that may be better equipped to deal with the new functions (Etzkowitz 2008; Boardman and Gray 2010; Smith et al. 2016). A second driver comes from the new demands emerging from non-research actors, including firms and public administrations. Interactions with non-research actors may create both economic and symbolic incentives, since research organizations either need constant funding or legitimation in a context of wider expectations (Gulbrandsen 2011; Bruno et al. 2017). The third driver of change comes from within in the form of organizational agency. Research organizations must face new ways of doing research, broader demands, and increased scientific and technological competition. Policy reforms may give PROs and HEIs more autonomy to use their own capacities and unleash their potential. Organizations are therefore in constant flux to foresee new goals and to find effective ways to meet the goals assigned.

Heterogeneous research systems are a result of this trend. Heterogeneity has become an important issue of study in research systems in order to understand the nature and system level performance required to achieve diverse policy goals (Huisman et al. 2015). Hybridization is a special manifestation of heterogeneity. In the field of R&D, hybridization is usually defined as the participation of actors from different institutional domains within the same organization (Emmert and Crow 1988; Borys and Jemison 1989; Crow and Bozeman 2005). It usually entails the adoption of a mix of practices that were previously spread over different organizational domains with different goals, activities, and legitimized procedures. Research organizations have several forms of hybridization in terms of ownership, funding, missions, regulations, and human resources. In this article, we refer to public–private partnerships or to specific centres that incorporate a diversified portfolio of activities to meet different demands, different profiles of human resources, and a mix of funding in order to promote cross-sector research collaboration between science and industry (Boardman and Gray 2010; Gulbrandsen 2011).

Research and Technology Organizations (RTOs) are important examples. The specific feature highlighted by scholars is that RTOs are intermediary actors between science and business, and aim to improve innovation and firm competitiveness (Hales 2001; Arnold et al. 2010; Giannoupolou et al. 2019). Although many of them were originally designed to provide innovation services, they are often required to increase their capacities and evolve into centres of research excellence. More recent examples are Cooperative Research Centres (CRC), defined as stable formal structures for R&D, whose mission is to promote cross-sector collaboration, knowledge and technology transfer, and ultimately, innovation (Boardman and Gray 2010: 450; Gray et al. 2013: 10). Many are designed to fulfil strategic R&D in state-of-the-art stages oriented to innovation. In some countries, they are also known as Competence Centres (CREST 2009). Finally, many examples of hybrid centres with industry participation and varying degrees of formalization can be found in the academic domain (Etzkowitz 2008; Turpin and Fernández-Esquinas 2011; Gray et al. 2013).

Some implications of this trend are worth highlighting. The literature on hybrid research organizations pinpoints the activities that private companies, research actors, and governments cannot easily address separately (Boardman and Gray 2010). The most observable benefit is often the impact on small and medium-sized enterprises (SMEs) and specialized firms collaborating with these centres, and the effects on local innovation systems in the form of knowledge-intensive business services. On the academic science side, they incentivize research partners to undertake more applied lines of research and facilitate the use of science and technology to tackle social and economic challenges.

Some unforeseen effects of heterogeneity and the emergence of hybrid forms also have important implications for policy and evaluation. First, the assimilation of entrepreneurial practices in academic science domains has an impact on incentives and rewards. The rewards to solutions that are useful for specific clients are sometimes detached from contributions to public knowledge, leading to conflicts between competing goals of organizations, role strain, and precarious scientific careers (Garrett-Jones et al. 2010). Second, the system becomes increasingly complex when an array of organizations emerges with different legal frameworks, structures, constituencies, personnel, forms of funding, and management tools. They are sometimes difficult to manage and understand from the outside when they are required to be subject to public accountability. Competition also increases when different actors pursue the very same activities and purposes. One implication is the difficulty of performing evaluation and foresight exercises to assess the composition of the research system.

Important research questions that emerge include the extent to which new centres fulfil the mission of knowledge transfer and whether they are in fact different organizations that facilitate collaborative research with industry and societal actors. In sum, what type of organizational heterogeneity is needed in the system in order to deliver different activities and missions in combination with traditional research organizations? It is therefore necessary to empirically specify the nature of the process by examining both diversification and institutionalization.

2.2 Institutionalization

Institutionalization refers to the enactment and recognition of the role of new actors in the research system. It is a necessary process for organizations to accumulate capacities and produce results, generate a critical mass, and achieve specific presence and acknowledgement in the system as important actors. The institutionalization process usually refers to both organizational stability and social embedding. One dimension of institutionalization has to do with the creation and diffusion of new organizations enacted with hierarchical structures, jobs, and resources that eventually give them a stable, formal structure. Other dimensions are bound up with symbolic elements: the emergence of values, norms, and practices that are eventually taken for granted, encompassing the legitimation of practices that contribute to visibility, stability, and access to resources (Colyvas and Jonsson 2011). The process also encompasses an organizational field understood as the array of organizations working in a recognized area of institutional life, together with the policies, rules, regulations, and routines that are applied and acknowledged in their activities (DiMaggio and Powell 1983).

Contributions from economic sociology suggest observing the interplay between institutions and organizations, because the socialization of values and norms does not happen in a vacuum but within (or between) tangible organizations (Portes 2010). The dynamics of such organizations are conditioned by the social structure of positions between social actors. Values, norms, and roles are not enough to explain institutionalization. Scholars working in university–industry collaboration show that institutionalization involves a gradual process of formalization, recognition, and legitimation, leading to more stable organizational structures (Colyvas and Powell 2006; Youtie et al. 2006).

In this study, we assume that the composition and transformation of a research system is the result of a combination of innovation policy and adaptation. The emergence of new organizations is shaped by previous organizational structures, policy frameworks, and competing actors. How a system develops should be studied by not simply focusing on formal policies and organizations identified in formal registries. The real evolution of the system, however, should be studied by observing the adaptation and emergence of old and new organizations, respectively. Institutionalization is therefore an important process for studying heterogeneity. Attention should be paid to organizational stability and social embedding as processes that allow a new population of centres to fulfil new missions and to be acknowledged as distinctive agents of the research system.

Several dimensions that allow new organizations to emerge and to become visible as differentiated actors of a research system are considered. Structuration refers to the accumulation of available resources that increase the organization’s stability and provide visibility by external actors. Complexity is associated with internal organization, diversified human resources, activities, and links to the external environment. Stabilization refers to rules, norms, and procedures within the organization that contribute to its continuity over time. Finally, Formalization consists in the process of making norms, rules, and procedures explicit through standards or other formal mechanisms easily recognized by external observers. These four features are closely interrelated and contribute to the enactment of a new organizational field. We use them to guide our empirical analysis.

3. Cross-sector research collaboration in Spain

3.1 Overview

According to the European Commission, for many years, Spain has been considered a ‘moderately’ innovative country (European Commission 2010a,b), ranked in an intermediate position among the R&D systems of developed countries. It has an extensive university sector, several networks of PROs, and several agencies for funding, evaluation, and accreditation.² In the last 30 years, higher education and scientific capacities have grown considerably alongside the economic development of the country. Public expenditure and the numbers of universities, academic researchers, and scientific publications have increased continuously. This process of growth has been subject to important drops in public expending during periods of crisis, especially between 1992 and 1997 and after 2010. Overall investment in R&D, scientific and technological productivity, and innovation performance figures are still below average for EU and OECD member countries (COTEC 2015).

Important causes of this lag are due to the gap between academic science and industry. The private sector’s contribution to R&D expenditures in Spain is much lower than in other EU and OECD countries (FECYT 2015). Of R&D personnel, 35% work for the industry sector, while 64% are employed by universities and public administrations. Without industry figures, the profile of the research system shows a more convergent pattern, especially when the results are compared with the level of investment in R&D (ICONO 2016).

Structural features of the research system contribute to this situation. Important limitations include the predominance of SMEs, the importance of low-technology economic sectors, the dearth of transnational corporations in leading high-tech sectors, the concentration of R&D in the academic sector, the complexity of the multilevel policy framework, and difficulties for cross-sectoral mobility of researchers. Several reports have provided descriptive overviews of the interface mechanisms and public–private partnerships, describing an ‘implementation gap’ between science and innovation (OECD 2005; FECYT 2006). However, still lacking are explanations of the institutional determinants that hinder cross-sector research collaboration, as well as the rationale for emerging trends in the organizational field or R&D between academic science and industry.

3.2 The institutional setup of R&D and innovation policy

Universities and PROs are the main actors in the system. They are public bureaucracies governed by the laws and regulations of central and regional governments. Laws and regulations are important features of system governance because they affect funding, expenditure, human resources, and key management issues. Governance in HEIs and PROs is shaped by rules and political decisions set up mainly at state level. Evaluation procedures are carried out by external agencies also dependent on central and regional governments. On the other hand, strategic decisions and activities of HEIs and PROs follow a bottom-up process. Most executive staff appointed at universities, and departments and research units of PROs are elected by academic constituencies. The relevant organizational units for carrying out research are research groups formed by professors and researchers. Most decisions for research programmes and projects are made at research group level.

Human resources, recruitment, and mobility of personnel are subject to rigid administrative procedures. Tenured researchers and professors are public servants. Teaching and research personnel in academic tenure tracks are subject to regulations and external evaluation of performance-based mostly on academic performance. Part-time appointments of researchers are highly restricted, and temporary cross-sector mobility for professors usually means the loss of career advantages (Andújar et al. 2015).

Public expenditure on R&D is subject to political negotiations when distributing the annual general state budget. Direct funding of PROs and the main R&D agencies depends on central government, while funding for universities and most RTOs depends on regional governments. An important issue for research management is the accountability of budgeted expenditure. In HEIs and PROs, controls for expenditure are made ex-ante, an important restriction for management when compared with similar organizations in other research systems in Europe and beyond. Contract research between sectors is also highly regulated, especially the exchange of economic resources in both directions. Until very recently, universities were not allowed to invest in innovative business or to be official partners in new joint ventures. These arrangements are overseen and approved by several external bodies and pose management constraints on creating new organizational forms.

Another feature of the Spanish innovation system is its multilevel nature. The devolution process resulted in the emergence of the regions as active agents with their own science, research, and innovation policies.³ Regional governments’ policies have been growing steadily. Initially, priorities were focused primarily on academic funding, but later moved towards innovation and development. Regional governments were also given responsibility for the funding and management of universities and research facilities in hospitals. However, the pace and direction of change have differed considerably across regions. With few exceptions, such as in the Basque Country, the early trend in the 1980s and 1990s was to support academic science because of the need to strengthen the scientific capacities of an expanding university sector. Later developments have resulted in heterogeneous regional systems in terms of the specific combination of resources.

Currently, R&D expenditure by regional governments is about 50% of total public expenditure, although the main regional programmes are concentrated in the bigger regions: Madrid, Catalonia, Andalusia, the Basque Country, and Valencia combined account for more than 80% of R&D (ICONO 2016). At the same time, European agencies have also started to play an important role, both as a source of funding for academic research and for the development of firm-based infrastructures, especially in regions and suburban areas with a lower Gross Domestic Produc (GDP). This situation has given rise to a complex system. Spain has the dynamics of a federal state with active policies carried out by several ministries, regional governments, and some autonomous research and innovation agencies, in addition to universities. The result is a fragmented policy framework with little coordination or distribution of policy responsibilities.

3.3 The emergence of hybrid research organizations

Since the 1980s, various governments have implemented policies to promote knowledge transfer to firms following a supply-side model. A wide range of tools has been launched, including grants, loans for collaborative projects, and several interface organizations. The main policies for promoting cross-sector research collaboration and knowledge transfer have resulted in the predominance of short-term projects and networks subject to policy changes and funding availability.⁴ Policymakers and officials responsible for R&D have received few incentives to adopt policy innovations applied to traditional universities and PROs that overcome rigid legal frameworks, especially those that are long-term and costly, such as collaborative research centres. New organizational forms for collaboration have emerged as the result of stakeholders’ reactions to system restrictions. Consequently, there is no single national scheme or unique model, or a federal network of RTOs, collaborative research centres or other bodies that perform collaborative research between science and industry. The variety of structures basically depends on the agreements and capabilities of the partners, as well as the availability of funding, according to the portfolio of legal forms available.

The financial and economic crisis affecting southern European countries has driven this trend due to economic restrictions and state reforms. After 2010 in particular, the downturn resulted not only in cuts to funding, but also in an increase in administrative restrictions for expending and receiving money from other actors, especially other public administrations and private bodies. These restrictions have been applied to universities and PROs in the same way as to other public administrations. The Spanish government’s application of EU austerity policies consisted in limited funding, restrictions to expenditure and regulations that hindered seeking funding from external sources. The mix of restriction and bureaucratic controls has affected the management capacities of HEIs and PROs when recruiting new research personnel, purchasing new equipment, attending to research contracts and consultancy from industry, and providing specialized training (Fernández-Esquinas 2015). These measures have resulted in a greater need for funding mixed with increased competition for international funding, and increased demands from firms and public administrations.

Many policy initiatives for linking academic science and industry do not stem from central government, but from research communities, firms, and administrations attempting to adapt to the new situation. Due to a lack of policy reforms, increased bureaucratic controls and reduced funding many agents have reacted by using their accumulated resources and capacity to act. Regional governments, some ministries, hospitals, and public bodies specializing in a variety of sectors [transport, engineering, food regulations, environment, social services, Innovation and technology centres (ITCs), etc.], together with HEIs and PROs, have actively sought models to produce knowledge more adapted to practical demands in collaboration with private actors. They have used their autonomy and the current legal framework to create alternatives to public bureaucracies, such as private foundations, associations, consortia, and for-profit joint ventures with the participation of firms. Because they are legally independent bodies, they can receive funding, recruit human resources, and provide contract research and services oriented to the demands of industry partners and public administration beyond the short-term projects and consultancy provided by research groups inside universities and PROs.⁵ One of the rationales of the emergence of these centres is to facilitate the participation of SMEs, which lack the capacity to engage in fluid collaboration with academic organizations; many are innovative but not R&D-intensive firms.

Firms participate in these new arrangements in different ways and can be part of the ownership. In some centres (e.g. foundations), they do not invest important sums, but can participate in steering committees by appointing representatives. In many cases, a considerable chunk of resources comes from the public sector in the form of direct funding, subsidies, and grants, but also in the form of infrastructure and personnel paid directly by universities and PROs. The new arrangements are therefore not completely independent of the public research sector. Frequently, public bodies, including governments, HEIs, and PROs, are part of the constituencies of new organizations. Research staff usually include a mix of newly recruited researchers and part-time professors, whose salaries are paid by the university. In sum, new centres consist in environments to carry out R&D oriented to the demands of the industrial sector, with management tools specifically designed for their missions.

The result has been the emergence of a new population of organizations specifically oriented to collaborative R&D that overlaps with existing actors. They are a diversified array of centres differing in size, legal status, participation of firms, formal missions, activities, and operational procedures. Some are RTOs created mostly by regional governments. In their early stages, many provided technology upgrading, training, and services in line with the needs of participating firms, although the evolution of industry sectors has resulted in a wide range of centres with higher competences (Cruz-Castro et al. 2012). Lately, some have been specially designed to create critical mass and an appropriate environment around a strategic technology sector, while others aim for basic science targeting commercialization (Giachi 2018).

Without examining the specific configuration and activities of these centres, evaluating their competences is difficult. But they have one important feature in common: they are a different species from traditional organizations. They blend various practices from institutional domains previously separated because of the historic institutional system setup and they correspond to the usual definition of hybrid centres outlined above. Nevertheless, some may generate redundancy when competing with traditional organizations for the same resources. They also generate complexity when seeking to identify the strengths of the system and the most suitable actors to distribute resources. The following sections develop detailed procedures in order to make them visible and to observe their orientation and activities.

4. Methodology

4.1 Towards an operational definition of HRCs

In this study, we combine the definition of CRC provided by Boardman and Gray (2010: 450) with the current definition of RTO (Gulbrandsen 2011; Zacharewicz et al. 2017) as used by the European Association of RTOs (see Section 2). We define HRCs as an analytical category that is applied to all research organizations that (1) have a stable organizational structure and are legally recognized, (2) perform R&D activities, and (3) include both public and private actors in the ownership or the management of the organization.

Therefore, we distinguish HRCs from arrangements that (1) do not have a stable organizational form and a formal status (e.g. public–private research networks, research projects, departments or units that are not independent from the organization to which they belong), (2) do not perform R&D (e.g. technology transfer offices, technology parks, firm incubators), and (3) are not hybrid in nature (e.g. PROs with no private involvement, HEIs, and private laboratories).

4.2 Mapping HRCs in Spain

There is no official register of semi-public or public–private research centres in Spain, nor a complete directory. Due to the diversity of policy actions at different administrative levels—national and regional—it is difficult to identify new actors and their constituencies.⁶ Therefore, we mapped the existing hybrid research organizations in Spain through a systematic review of secondary sources of data and web search. Following the operational definition above, we adapted the categories to the administrative and policy context of Spain. We used legal criteria to consider an organization as an autonomous entity, the involvement of private actors in public administrations, and the capacity of public administrations to co-own private entities.⁷ We selected organizations where the following characteristics are explicitly identified:

• They have a formal structure and are recognized as legal entities.

• They perform R&D activities.

• They have at least one public and one private actor among their partners.

It is important to note that we excluded collaborative agreements for R&D and innovation such as inter-institutional collaborative research networks that have no stable structure, units with private funding within existing HEIs or PROs but with no legal status, and private R&D organizations with no public partner, in addition to innovation interfaces that do not carry out R&D. We restricted our search to organizations that can be acknowledged as legal actors in the system, although the practices of private funding in academic science and public and private partnerships are wider (Fernández-Esquinas and Ramos-Vielba 2011). Our search procedure therefore allows us to focus on the emergence of a specific type of organization different from existing ones.

We applied the following three-stage search procedure:

1. We reviewed the set of public national and regional R&D programmes and plans. We used a word search engine for a list of keywords (‘public–private cooperation’, ‘collaborative R&D’, ‘research centres’, etc.) to find existing policies supporting collaborative organizations and public–private research centres.

2. With this information we performed a systematic review of a set of web directories containing extensive lists of the most important R&D centres, obtaining a long list of possible HRCs in Spain (around 300). We registered their name, location, and website.

3. We then systematically checked correspondence with the elements of operational definition. We reviewed the webpages and institutional reports containing corporate information, including the history, mission, organizational structures, partners, and activities of the centres. This search reduced the number of possible HRCs to 234 cases. After reviewing the information provided by the survey (see information about the survey below), we excluded another 18 centres for not accomplishing the three criteria. The population of HRCs identified in Spain is composed of 216 cases.

Table 1 shows the distribution of the 216 HRCs based on the above definition. We have identified two levels of categories for classifying the centres that are meaningful according to institutional background and specific features. First, we consider the general categories (boldface in Table 1). For each, we identify the specific denomination of institutional affiliation when possible:

• ITCs: ITCs constitute the largest category with almost two-thirds (64.4% of the population). Their common characteristic is that they denominate themselves as research and technologies organizations whose primary mission is the provision of direct services to industry. In this category, we also included other centres with different denominations but common goals: 43 (19.9%) are affiliated to the Spanish Federation of Innovation and Technology Entities (FEDIT); 81 (37.5%) are not part of the FEDIT or any other sector association or network; nine are associated with one of the networks of technology centres in the Basque Country, IK4; three are associated with the other network of technology centres, Tecnalia, also in the Basque Country; and three are Microsoft innovation centres working in collaboration with regional governments.

• Cooperative research and excellence networks: They make up 12.5% of the population and their distinctive characteristic is that they are created through public programmes for promoting collaborative research. The largest group is made up of nine Biomedical Research Networking Centres (CIBER, Spanish acronym for Centros de Investigación Biomédica En Red), followed by seven Institutes of Advanced Studies in Madrid (IMDEA), and in the Basque Country, seven CRCs and four Basque Excellence Research Centres (BERC). Private participation and the degree of firm involvement in the governing boards of these centres can be very diverse but are encapsulated in their policy rationale. Big corporations operating in different sectors, including high-tech (Airbus, H&P, Abengoa, Biosearch life, etc.), energy (Iberdrola, Repsol), finance [Banco de Bilbao, Vizcaya y Argentaria (BBVA)], telecommunications (Telefonica), dairy products (Danone, Pascual), and construction (Sacyr), are involved with IMDEA institutes. By contrast, private participation in Basque BERCs and Centros de Investigación Cooperativa/Copperative Research Centres (CICs) more often involves private associations, cooperatives, and local companies, but also a few big high-tech companies, including Spanish corporations of the Ibex-35 group. Private partners, such as hospitals, foundations, and associations, are linked with CIBER centres. Despite the lack of industry firms, we included CIBER in the universe because they are participated by private institutions and because of their networked structure. This fact reinforces the evidence for institutional diversity in Spanish HRCs.

• Ad hoc R&D institutes: They constitute 23.1% of the population. This category is formed by centres that are not part of a network or a programme but are created by different arrangements for specific purposes of cross-sector research collaboration, usually aligned with a strategic research line or a productive sector: 23 are semi-public centres formed by different arrangements of private entities and industries (10.6%); 11 are University Institutes for Research (Institutos Universitarios de Investigación) with public–private participation (5.1%); eight are collaborative institutes supported by the Institute of Higher Business Education (IESE, Instituto de Estudios Superiores de la Empresa), and the others can be classified as semi-private philanthropic institutes.

The distribution of the population of centres by age is quite heterogeneous and concentrated at both extremes of distribution (Table 2). There is a significant number of young centres: 51.9% were no more than 10 years old at the time of data collection (2013)—they were created after the year 2000; 19% are over 20 years old. There are fewer centres between 11 and 20 years old. Although the 1990s and early noughties seem to be the period when the creation of these organizations was less intense, interestingly, it is the period of increasing public and private expenditure for R&D. ITCs are usually older, while networks created by public programmes are more recent and concentrated in periods when these programmes received government support. In general, fewer new centres have proliferated in the period of affluent resources during the growth of the system, while more ad hoc centres have been created more recently following the stagnation of R&D funding, the introduction of administrative restrictions to public expending, and the regulations applied to the management of human resources and funding of traditional HEIs and PROs.

4.3 Survey

A structured questionnaire was designed for qualified staff members who were able to provide relevant information. The survey mainly targeted centre directors, although according to the criteria of the director at each centre; some respondents had other job profiles (Fernández-Zubieta et al. 2016). The questionnaire included sections for goals, collaborators, funding, human resources, management, activities, outcomes, and estimations of impacts.

It was sent to all 216 centres identified through the mapping process. We used a postal/web mixed method for surveying (Diment and Garrett-Jones 2007) that included an online questionnaire, an invitation through postal letters, and email and telephone reminders, using the Computer Assisted Telephonic Interview (CATI) system. Online access to the questionnaire was open from July to November 2013. We finally sent six email and three postal reminders to the centres. The survey response rate was 59.3% (128 cases). The sample shows no bias according to variables such as institutional background, organizational age, or region of location (see Appendix). Results for centres’ outcomes were obtained through a self-reporting process for the respondents to the questionnaire.⁸

We have selected some items of the questionnaire as indicators of institutionalization (Table 3). Some indicators of financial and human resources refer to the structuration and stabilization dimensions of institutionalization. Indicators of financial turnover and total workforce reflect centre size and the total organizational capital available. The number of internal administrative workers reflects the management capital (Boardman and Ponomariov 2014), while external R&D funding is a measure of capacity and external recognition, as well as organizational autonomy (i.e. when the centre does not rely on public subsidies or private fees). Other indicators are related to the complexity and formalization dimensions of institutionalization, such as the existence of formal research groups, an external advisory committee and the presence of scientific or technical evaluation practices.

According to the descriptive results of the survey, the presence of HRCs in the system is significant in Spain. The average centre has a financial turnover of millions of euro, 94 full-time workers, 78 R&D projects executed in the last 3 years through external funding (public or private), and 8 or 9 administrative employees directly recruited by the centre, in addition to other staff provided by other organizations. At the same time, more than half the centres are structured through formal research groups (55.5%) and are subject to an external technical evaluation (56.3%), while around half are subject to external scientific evaluation (44.1%) and have an external advisory committee (49.6%). Centres frequently publish their research outcomes in international (59.5%) or national (64.7%) journal articles, and communications at international conferences (75.9%). Centres also make patent applications (55.9% since startup) and supervise doctoral theses (48.3%). However, they are less likely to participate in the creation of a new R&D or innovation-related business or firm as owners (only 17.6% did so).

In sum, these data show the existence of an important group of actors, which are not fully reflected in official directories. Although many are individually acknowledged by their constituencies in their sector of activity, they remain hidden as a distinctive category of the research system. The main reasons for their lack of official visibility include either the participation of different governments (national and regional), with little coordination between them, or the initiative of PROs and HEIs that use their autonomy to create distinct formal organizations in collaboration with private actors, sometimes in collaboration with regional governments or ministries. Others are led by private actors, with the support of different levels of public administrations, and have an independent status regarding national or regional innovation policy schemes.

The map may not reflect all the centres that can be included under the umbrella of the operational definition because the search procedure cannot guarantee complete coverage. In addition, other centres or organizational units were set aside because we were unable to prove the real involvement of private entities in their ownership and operation centre. Therefore, these data should be considered as the minimum amount detected in the research system with the method applied. Overall, we estimate that these centres employ more than 20,000 workers, incorporate about 7,500 different firms and private entities, develop around 5,600 projects and mobilize 865 million euros each year (Giachi 2018). These figures suggest that HRCs are an important component of the Spanish research system.

The following analyses explore centres’ main institutional features and activities. Our aim is to detect the patterns of their organizational characteristics guided by our assumptions about institutionalization as a necessary component of the creation of distinctive types of organizations. The methods reduce the list of indicators into summative indexes using Principal Component Analysis (PCA) and Categorical Principal Component Analysis (CATPCA). We then used the resulting indexes for a hierarchical cluster analysis that classifies centres by their degree of institutionalization. After classification, we describe the profile of each class of centre using variables reflecting background, age, activities, and outcomes.

5. The emerging field of hybrid research organizations

5.1 Institutionalization and organizational models of the centres

In a first step, we applied a PCA to the four indicators related to the structuration and stabilization dimensions of institutionalization, as suggested in Section 2 (financial turnover, external R&D funding, administrative personnel, and workforce). We extracted a single component (eigenvalue higher than 1) that explains 72% of the variance. Specifically, contributions to the component are higher for the number of people working at the centre (especially administrative personnel explicitly recruited by the centre), and the number of R&D projects funded through competitive calls and contracts (Table 4). We labelled this component capacity to gather resources.

In a second step, we applied a CATPCA to the four categorical variables related to the dimensions of institutionalization considered in Section 2 as complexity and formalization (formal research groups, advisory committee, scientific and technical evaluations). Again we extracted only one significant component (Cronbach’s alpha = 0.617, eigenvalue = 1.862) which explains 46.5% of variance. The second extracted component did not show good quality measures (Cronbach’s alpha = −0.083, eigenvalue = 0.942). It explained only another 23.5% of variance, and its interpretation was not straightforward. Therefore, we decided to retain only the first component and label it as Organizational Complexity (Figure 1). Contributions of the original variables to this component are the following: external scientific evaluation (0.787); organization by research groups (0.727); external technical evaluation (0.613); and presence of an external advisory committee (0.582).

We used the two indexes of institutionalization as classificatory variables for a cluster analysis. An optimal classification in correspondence to four groups was obtained, one of them formed by only one case (Figure 2). We eliminated the outlier case (it has extremely high values of institutionalization, especially for resources) and retained the other three groups for building a categorical variable to classify HRCs according to their degree of institutionalization. To interpret the profile of each group, we performed an analysis of variance (ANOVA) on the typology of centres by degree of institutionalization. The ANOVA was highly significant and discriminatory between groups, according to the mean of classificatory variables. In a (reduced) sample of 116 cases finally included in the cluster analysis (plus the subsequently case excluded), the centres fall into three categories:

• Lower degree of institutionalization includes most of the sample (n = 66; 56.9%). Centres are characterized by low scores in both indexes resulting from the above analyses. These cases have low numbers of financial turnover, workforce, and of contracted administrative personnel and external funds for R&D, in particular. They have no internal research groups, advisory committee, or received external evaluations. In other words, they are less stable, formalized, and complex.

• Medium degree of institutionalization (n = 39; 33.6%). Centres are characterized by higher values of the variables mentioned above. They have formal research groups, advisory committees, and receive external evaluations.

• Higher degree of institutionalization (n = 11; 9.5%). Centres form a smaller group whose values in the above variables are higher, reflecting bigger and more stable structures. Box 1 gives examples of the typology.

5.2. Analysis of degree of institutionalization by types of centres

The degree of institutionalization is related to the background of the centres (Box 1 and Table 5). ITCs are distributed more equally between degrees of institutionalization. While almost all the centres showing a higher degree of institutionalization are ITCs, most cooperative research and excellence networks centres fall into the Medium degree of Institutionalization group. Many ITCs and most ad hoc R&D institutes have a Lower degree of institutionalization. It seems that centres created through direct public support become institutionalized more easily or faster than the others, while the existence of a small group of highly institutionalized ITCs could be explained by the presence of very old and long-standing initiatives, e.g. those deriving from former Industry Research Associations (Table 6).

The degree of institutionalization is also related to organizational age, though this relation must be interpreted carefully (Table 6). Centres with a higher degree of institutionalization are older. Centres with Medium degree of institutionalization show a more homogenous distribution, but with more presence of centres older than 15 years. Many centres with a Lower degree of Institutionalization are younger, but this is especially true for centres launched between 11 and 15 years prior to the survey: centres created in the period 1996–2001 are less likely to become institutionalized. In sum, these data show that the creation of new centres is slow-paced. In relative terms, however, these centres are much younger structures than most of the traditional PROs and HEIs in Spain.

According to Tables 5 and 6, institutionalization is only partially related to the history of the centres, suggesting that other structural facts can explain diversity. For instance, centres’ outcomes show different profiles. For analysing how the production of science and technology by centres varies according to their degree of institutionalization, we use six binary indicators of production, including publications, dissertations, patents, and new ventures (Table 3). Each indicator shows whether the centre obtained the outcome during the previous year, in the case of publications and dissertations, and since the creation of the centre, for patents and participation in spin-off firms (Figure 3).

The results show that institutionalization is directly related to performing more types of activities and achieving different types of outcomes (Figure 3). There is a positive relation between degree of institutionalization and the diversification of S&T outcomes. The small group of Highly Institutionalized centres usually obtains every type of outcome in over 80% of the cases. The only exception is the participation in new ventures (45.5%), though this outcome is not common to all centres. Centres with a medium degree of Institutionalization obtain each type of outcome much more frequently than those with lower Institutionalization, with the only exception of new ventures (same percentage). They also surpass Highly Institutionalized centres for international outcomes.⁹

In short, our results show that more institutionalized centres obtain different types of S&T outcomes more frequently, which suggests that they have more capacity to produce outcomes related to research (publications, patents, etc.) and that they tend to diversify their portfolio of activities leading to different outcomes. It is also possible that centres with a lower level of institutionalization specialize in activities not measured by our indicators, which is a matter for further research.

6. Conclusions

This study outlines the processes leading to heterogeneous research organizations in a national research system. We have explained the emergence of a population of distinct research centres formed by constituencies from academia and industry. We have focused on HRCs in Spain. We have shown how the lack of policy initiatives combined with bureaucratic rigidity and new demands from diverse actors—especially firms—generate heterogeneity and the structuring of a new population of research centres through a bottom-up process. The emerging population of organizations is heterogeneous due to the unplanned nature of the process and the diversity of participating actors, including governments at national and regional levels, associations of SMEs, private corporations, and universities attempting to overcome the administrative barriers of public bureaucracies.

We have found that the policy context and the nature of these organizations make them difficult to observe. They are usually overlooked by policy initiatives and evaluation exercises aimed at recognizing specific policy targets for promoting cross-sector research collaboration and innovation. Therefore, the study of heterogeneity in research systems in such circumstances is only possible through specific research designed to capture the emergence and composition of new actors. In order to make them visible, we have carried out an original methodology based on an operational definition, a systematic review of secondary data sources and a mapping process that allows us to identify a population of new centres. It has been combined with a survey to centre directors focusing on internal characteristics and outcomes. The results show the important role of these organizations, in addition to universities and public research centres. They also show different degrees of institutionalization according to organizational variables. We have identified three groups of centres oriented to different science of technology outcomes. We have found that institutionalization shapes orientation to industrial science, academic science or a mixture of both, as well as the diversification of activities.

We consider two potential implications of our findings. Regarding research evaluation, there is robust evidence to support the positive relation between different dimensions of institutionalization and a diverse portfolio of activities and outcomes, whether academic, in the form of publications and theses, or in terms of knowledge transfer to firms in the form of consultancy, applied research and training. In particular, the evidence shows that centres tend to evolve towards a more complex portfolio of research activities, becoming multipurpose. Some are subject to the funding and evaluation criteria of academic science, while others pursue knowledge transfer activities that require specific evaluation observations. An important issue is the balance between the criteria of research excellence and services oriented to fulfil the demands of their original constituencies, mainly SMEs. We can expect the diversification of multipurpose research centres to require specific efforts for research evaluation adapted to different rationales.

Regarding the implications at system level, our findings suggest that institutionalization is a key feature of an emerging organizational field. Different paces of institutionalization are therefore a major cause of heterogeneity in a research system. Dimensions related to structure, stability, formalization, and complexity are important issues to consider, because they shape the capacities and the activities of the centres. Research policy programmes for promoting organizational change may contemplate these dimensions of institutionalization and their significance. Finally, it is important to mention that the conclusions of this article have empirical constraints. The mapping process combined with a survey to centres has limitations for observing long-standing processes of institutionalization and indicators of activities. This methodology would benefit from a triangulation with qualitative observations and quantitative accounts of results. The lack of available data also suggests that new organizations should be considered in official statistical reporting in order to make them visible.

Footnotes

Acknowledgements

The authors acknowledge Juan Antonio Dominguez Álvarez, Ana Fernández-Zubieta, and Inés Andújar-Nagore for their support in collecting the data for the analysis.

We acknowledge support of the publication fee by the CSIC Open Access Publication Support Initiative through its Unit of Information Resources for Research (URICI).

Funding

This work was supported by the former Ministry of Economy and Competitiveness through the Spanish R&D Plan, grant numbers: BES-2011-047258 to S.G. and CSO-2010-14880-SOCI.

References

Appendix