Diffusion of innovation, innovation studies, and innovation policies: cross-fertilisation to improve smart specialisation strategies (S3)

Abstract

The paper reviews the diffusion research approach to the analysis of innovations and its impact on innovation studies. This social approach brought a new understanding of the innovation process by focusing on analysing the factors that influence the spread of innovations instead of focusing on its technical basis. We also review the different approaches to the study of the spread of policies – policy diffusion, policy transfer and policy mobility. We review the reception of diffusion research in innovation studies to recover some insights and to integrate them into current innovation policy debate. We use the innovation policy of Smart Specialisation Strategies (S3) as an example of how this integration could be done. After reviewing the main challenges of S3, we show the importance of improving the implementation phase of innovation policies through different mechanisms. For example, we indicate the need of paying more attention to mechanisms of policy diffusion, such as learning and, specially, socialisation as these could lead to more stable and long-term policy changes.

Keywords:

• Diffusion of innovation
• innovation policies
• science policy and innovation studies
• smart specialisation strategies

JEL CODES:

• O30
• O32
• O38
• Q52

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1 Also known as ‘innovation studies’, as we indicate in section that analyses the impact of diffusion research on the field, different reviews of the field (e.g. Martin 2012; Fagerberg, Fosaas, and Sapprasert 2012; Bhupatiraju et al. 2012) differ in their consideration as part of the field of some approaches (e.g. ‘science and technology studies’).

2 The second edition was published in collaboration with Floyd Shoemaker and it was entitled Communication of Innovation: A cross-Cultural Approach [1971]. Subsequent editions reverted to the original title – third edition in 1983, fourth in 1995 and fifth in 2003.

3 Criticisms of diffusion research include: ‘pro-innovation’ bias that assumes that innovation should be diffused and adopted; ‘individual-blame’ bias makes individuals responsible for her problems instead of considering system-blame; ‘recall problem’ measuring the time of adoption; and ‘equality’ that ignores the consequences and the distribution of benefits of innovation.

4 For example, Schumpeter indicates that

the function of entrepreneurs is to reform or to revolutionize the pattern of production by exploiting an invention or, more generally, an untried technological possibility for producing a new commodity or producing an old one in a new way, by opening up a new source of supply of materials or a new outlet for products, by reorganizing an industry and so on. (Schumpeter 1942 [2003], 132)

5 Nelson and Winter (1977) defines innovation as ‘a portmanteau to cover the wide range of variegated process by with man’s technologies evolve over time’ (p. 37).

6 ‘Viewed sociologically, the process of diffusion may be characterised as the (1) acceptance, (2) over time, (3) of some specific item – an idea or practice, (4) by individuals, groups or other adopting units, linked to (5) specific channels of communication, (6) to a social structure, and (7) to a given system of values, or culture’ (Katz, Levin, and Hamilton 1963, 240).

7 This process was conceptualised initially by Ryan and Gross (1943) and defined into five-step decision making process by Rogers: knowledge, persuasion, decision, implementation, and confirmation.

8 These are: anthropology, early sociology; rural sociology; education; public health and medical sociology; communication; marketing and management; geography; general sociology; and other.

9 Here, we consider analysis of ‘innovation studies’ based on recent bibliometric approach in order to provide a more accurate picture of the influence sociological work of diffusion of innovation. Previous studies based on handbooks are not considered. Similarly, bibliometric reviews focusing on economic contributions are generally not considered (e.g. Dachs et al. 2001). Other work, less recent or based on surveys (e.g. Fagerberg and Verspage 2009) is used to complement the bibliometric picture of the field done in this section. Despite the limitations of using citations as a proxy for the influence of any academic work (e.g. English-language bias; North American journal bias; citations from books not usually considered; citations not explicit), citations have the advantage of being an objective measure, they are widely used, and they can be compared across subject areas (Martin 2012, 1222). We endorse the limitations of using bibliometric indicators as a proxy for academic quality (e.g. San Francisco Declaration on Research Assessment – DORA 2015 and Leiden Manifesto (Hicks et al. 2015). However, as indicated, citation analysis has advantages for identifying highly influential academic contributions.

10 The high impact of Granovetter’s (1985) paper about ‘embededness’ in innovation studies could be related to the specific economic and sociologic dialog displayed in the article that argues ‘that most behaviour is closely embedded in networks of interpersonal relation’ (p. 504). It could be said that this line of argument could hardly be questioned by any diffusion research scholar in sociology. Other socio-political relevant concepts, such as, ‘social capital’ has received important vicarious attention in innovation studies through the work of, for example, Janine Nahapiet and Ghoshal (1998). Granovetter’s sociological work, similarly, to other contributions of ‘neo institutionalist’ approach from sociology (e.g. DiMaggio and Powell 1983) or political science (e.g. March and Simon 1958) could be considered out of innovation studies, but had an important impact on it. It is recognised the comparative low impact of political science literature on innovation studies (Martin 2012, 1235). It could be said that the influence of the socio-political ‘neo institutionalist’ approach in the innovation studies came mainly through the organisational and management studies literature.

11 Rogers’s work (1962, 1983, Rogers and Shoemaker 1971) rank first in the top-20 key references in the first period (1957–1979); second in the 1971–1985 period; third in the 1986–2000 period; and disappears in the 2001–2006 period rank (Rossetto et al. 2018).

12 Among the different theories of the policy process (e.g. Weible and Sabatier 2018), we choose policy diffusion due to the shared key concepts (e.g. S-shaped curved and typology of adopters) that help us to integrate insights in understanding innovation within the policy process. In addition to diffusion research, there are other approaches that address a similar question, such as ‘neo institutionalism’ (e.g. DiMaggio and Powell 1983) and organisational learning (e.g. Levitt and March 1988). However, they focus on institutional and organisational change rather than policy change.

13 Gray (1973) analysed 12 laws in education, welfare and civil rights and indicates the importance of considering different policy areas and individual policy measures separately.

14 See Walker (1973) for similarities and differences in both approaches.

15 ‘The reasons for political scientists' neglect of diffusion theory, of course, lie not in the peculiarities of their subject matter, but in the prevailing division of labour among scholarly disciplines in American universities’.

16 Some authors in policy diffusion choose ’emulation’ or ‘imitation’ over ‘socialization’ (e.g. Simmons and Elkins 2004; Shipan and Volden, 2012). Differently to their approach, we do not consider that ‘imitation’ or ‘emulation’ is a combination of socialization and learning.

17 Gilardi and Wasserfallen (2019) distinguish policy diffusion from policy transfer literature and recognise the difficulties in combining both approaches (p. 2). Berry and Berry (2018) signal the problems of trying to integrate different approaches in understanding government innovation as they rely in different views and methodologies. However, we do not aim at portraying a unified model of policy innovation. We think that it is important to complement different approaches as they seed light on different aspects of the policy innovation process.

18 It could be noted that the work of Dearing and Rogers on agenda setting (Dearing and Rogers 1996) also did not have an impact on innovation policy studies.

19 The definition and legal base established by The Regulation (EU) 1301/2013 of the European Parliament and of the Council of 17 December 2013.

20 K4G was established to strengthen the Lisbon Strategy by the Commissioner Janez Potočnik in 2005,

a group of prominent economists in the field of Knowledge for Growth in other to provide him with high-level advice on the following aspects: Contribution that knowledge can make to sustainable growth and prosperity; Optimum mix of policies needed to promote the creation, dissemination and use of knowledge; and Role that the various actors can play in stimulating a knowledge society and how to enhance the dialogue among these actors. [http://ec.europa.eu/invest-in-research/monitoring/knowledge_en.htm ]

21 Communication ‘Regional Policy Contributing to Smart Growth in Europe 2020’ (COM(2010) 553 final).

22 https://s3platform.jrc.ec.europa.eu/. It advises MS and regional authorities on how to design and implement their smart specialisation strategies, providing mutual learning, data gathering, analysis, and networking opportunities for more than 200 European regions.

23 The second ex ante conditionality on research and innovation referred to ‘Research and Innovation infrastructure. The existence of a multiannual plan for budgeting and prioritisation of investments’. (1.2)

24 ERDF funds represented 41.7% of total budget of the European Structural and Investment Funds (ESIF) for the 2014–2020 programming period (733 billion euro).

25 https://cohesiondata.ec.europa.eu/themes/1 [data retrieved: 30.04.2022]

26 These challenges refer to the translation of economic growth literature concepts, such as, ‘embeddedness’, ‘relatedness’ and ‘connectedness’ from sectoral to regional policy. Regions should increase ‘embeddedness’ of labour and industries to avoid outward human capital geographical mobility from increasing human capital regional capabilities (Granovetter 1985). They should apply a specialised diversification across related technologies (Frenken and Boschma 2007) to benefit from ‘relatedness’ and avoid a type of specialisation that do not allow regional diversification. ‘Connectedness’ should become ‘connectivity’ (Sassen 2002), a broader and spatially located concept, to avoid the Krugman shadow-effect (Krugman and Venables 1995) that indicates that policies aimed at reducing special transaction costs create a centrifugal force. See McCann and Ortega-Argilés (2015).

27 The EC have identified the following causes ‘as responsible for a difficult fulfilment of ex-ante conditionality in the area of RIS3: – A lack of political commitment to long-term strategies, and difficulties devising a bottom-up process of smart specialisation as a way to prioritise investments; – The split of responsibilities between different vertical (national, regional, municipal) and horizontal (between ministers and departments) levels, resulting in the fragmentation of powers and failed coordination; – A lack of effective, competent and adequately skilled staff in administration exacerbated by frequent legislative and institutional changes; – Weak collaboration between the public and private sector and thus little commercialisation of public research; – An absence of continuity in the entrepreneurial discovery process and a lack of key milestones, roadmaps, monitoring mechanism; – Tenuous links between financed projects and RIS3 priority areas. (EC 2017, 16).

28 ‘Since 2013, Fraunhofer ISI has conducted annual surveys of policy maker’s perception of the European Commission’s RIS3 agenda, starting from general assessments (2013) and views on options for implementation (2014) on to concrete questions on obstacles, monitoring and the need for interregional collaboration (2015)’ Kroll (2015).

29 Marinelli, Bertamino, and Fernandez-Zubieta (2019) has used this framework of policy dimensions to compare the S3 governance structures of Puglia and Extremadura, exploring the opportunities to integrate case studies results with this theoretical framework.

30 Here we do not think about constructing regional clusters of similar policies that have been considering ‘outdated’ (e.g. Shipan and Volden 2012). We recognise that in a globalised world regional policy makers or actors not only learn from experiences of nearby regions. However, ‘innovative indexes’ are necessary to categorise adopters and to know which regions or organisations are open to innovative practices.

31 According to Evans (2004), reputational indexes can’t capture the complexity of organisational structures. Jutting (2003) indicate these indexes are subject to biases as opinions are influenced by current economic performance of the country. Op. Cit. (Portes and Smith 2010, 586).

32 For example, Portes and Smith (2012) study five strategic organisations in five Latin American countries, showing that:

developmental institution, as a combination of at least some measures of internal quality with active engagement with strategic actors in their respective spheres of activity […] it is not lack of internal cohesion or meritocratic recruitment that represents the most signal shortcomings of Latin institutions, but their tendency to remain insulated, failing to engage other key actors in their environment. (Portes and Smith (2010, 615–616)

Additional information

Notes on contributors

Ana Fernández-Zubieta

Ana Fernández-Zubieta has degrees in Sociology and a PhD. in Humanities from the Carlos III University. She has a MSc in Public Policies for Science Technology and Innovation from SPRU, where she won the Geoff Oldham Prize for the best dissertation. Her main research topics are public policies for research and innovation, internationalisation of human resources for science and technology. She has several publications in international journals and books on these research topics. She was researcher at the Science and Technology Policy Research SPRU-University of Sussex, the Institute for Prospective Technological Studies (IPTS)-Joint Research Centre – European Commission (EC) and the Spanish National Research Council (CSIC). She is national expert on Research and Innovation public policies for Spain (ERAWATCH-EC and Research and Innovation Observatory (RIO-EC)). She is currently working at the University Complutense of Madrid, being the director of the ‘Economy and Innovation Policy’ research group at the Complutense Institute of International Research.