Journal of Evolutionary Economics (2023) 33:773–796

https://doi.org/10.1007/s00191-023-00829-3

REGULAR ARTICLE

A conceptual framework for modeling heterogeneous

actors’ behavior in national innovation systems

Apostolos Vetsikas1 · Yeoryios Stamboulis1

Accepted: 3 July 2023 / Published online: 17 July 2023

© The Author(s) 2023

Abstract
Various analytical frameworks, such as the National Innovation System (NISs) and
N-tuple innovation helices, have been developed to address technological change at a
spatial or sectoral-technological level. Several quantitative methodological approaches
have been used to capture the effectiveness, efficiency, and overall performance of inno-
vation at the national level. Reviewing these approaches, we highlight important aspects
of the innovation process, such as actor heterogeneity, the intensity of interactions, and
evolutionary dynamics within and between innovation subsystems that are often under-
estimated. We conceive NISs consisting of five interacting helices: government, aca-
demia, industry, society, and finance. Actors belonging to these helices develop their
behavior – in terms of resource commitment/allocation – in the context of interdepend-
encies and interactions that condition the effectiveness and efficiency of their actions.
As a result, their expectations are formed from their perception of how other actors
and the system behave. We develop a conceptual framework that goes beyond the static
illustration of ‘innovation scoreboards’ and linear models. It illustrates how individ-
ual parameter changes – in one helix of the system – may generate non-linear effects
throughout. We use a causal loop diagram (CLD) to depict the intricacies of the interac-
tions amongst various elements in NISs, and a stock-and-flow diagram (SFD), which
forces more detailed specification of causal mechanisms. Our framework facilitates
helix-based actor heterogeneity and highlights the key causal mechanisms and feedback
loops – set in motion from actor interactions – that govern NIS’s evolution and perfor-
mance without losing oneself in immense detailed complexity.

Keywords National innovation system (NIS),·Innovation helices · System dynamics

(SD) · Heterogeneity

JEL Classification C63 · O30

* Apostolos Vetsikas
[email protected]
Yeoryios Stamboulis
[email protected]
1
Laboratory of Management, Department of Economics, University of Thessaly, Volos, Greece

13
Vol.:(0123456789)
774 A. Vetsikas, Y. Stamboulis

1 Introduction

The concepts of Innovation Systems (ISs) and N-tuple innovation helices emerged
in the 1980s and 1990s to respond to the criticism of linear thinking of innovation
and of efforts to link demand-pull and technological advances in a holistic frame-
work. Both approaches consider the innovation process as a complex set of interac-
tions across many parts of the system.
The NISs concept was developed initially by Freeman (1987), Lundvall (1992),
and Nelson (1993), emphasizing the social, institutional, and evolutionary character
of innovation. The Triple Helix model (later developed in ‘N-tuple innovation heli-
ces’) was originated by Leydesdorff and Etzkowitz (1996), arguing that interactions
among academia, industry, and government are crucial factors determining condi-
tions for the innovation process (Etzkowitz and Leydesdorff 2000). Both approaches
(ISs and N-tuple innovation helices) are conceptual frameworks that conceive inno-
vation as activities and interdependencies and interactions between heterogeneous
actors that co-operate and co-evolve.
During the last two decades, increasing interest in the quantitative measurement
of innovation at the macro level has been explored through a variety of instruments
(benchmarking analysis/scoreboards) and empirical methods such as cluster analy-
sis, econometric analysis and modeling, data envelopment analysis (DEA), and sys-
tem dynamics (SD) (Chaminade et al. 2018). Although these methods have been
employed to capture the effectiveness, efficiency, and overall performance of NISs
in a broad range of countries (developed, developing, emerging, in transition), criti-
cal issues and questions remain, including:

• How can the evolutionary path of NISs be captured in dynamic models that
examine the specificities, behavior, and interactions of their constituents/compo-
nents?
• Through which indicators can the dynamic patterns of change and the evolution
of NISs over time be captured?
• What are the drivers of change and points of leverage within NISs for policy
effectiveness?
• What are the coordination mechanisms amongst the multitude of heterogeneous
actors (co-opetitive dynamics, feedback mechanisms, etc.)?

Considering these questions, we conceive NISs consisting of five interacting hel-

ices: government, academia, industry, society, and finance. We consider the heli-
ces as interacting ‘subsystems-actors’ and define the NIS as a system of interacting
helices. These intrinsically heterogeneous actors (heterogeneity in terms of goals,
intentionality, and original endowments such as learning capabilities, size, location,
experiences, etc.) develop their behavior – in terms of resource commitment/alloca-
tion – in the context of interdependencies and interactions that condition the effec-
tiveness and efficiency of their actions (Stamboulis 2007; Muñoz and Encinar 2014).
We develop a conceptual framework for modeling actors’ behavior in NISs. It
enables helix-based actor heterogeneity and highlights the key causal mechanisms
and feedback loops that govern the evolution and performance of NISs, all without

13
A conceptual framework for modeling heterogeneous actors’… 775

getting overwhelmed by excessive complexity. We use a CLD to depict the intrica-

cies of the interactions amongst various elements in NISs, and a SFD that forces
more detailed specification of causal mechanisms. We adopt a broad NIS concept
embracing not only institutions that are directly and explicitly involved in research
and development (R&D) but also the social, economic, and political arenas of inno-
vation (Fagerberg 2003).
Our main contribution is twofold. First, we provide a comprehensive overview
of policy instruments and empirical methods for the quantitative measurement of
NISs. To the best of our knowledge, no previous study makes a critical review of the
diverse quantitative methods for the analysis of NISs. Second, we propose a rela-
tively novel method for assessing NISs, which could lead to better policymaking
in the science, technology, and innovation domain and help countries better design
their innovation policies. The mainstream approaches lack a ‘system view’ of the
NIS and cannot model how the strategy of NIS emerges dynamically in a complex
social system from the decisions made by the agents involved in them (Dawid 2015).
We contribute to the modeling of NISs by developing a conceptual framework,
which constitutes a more holistic approach in comparison to relative studies (e.g.,
Grobbelaar and Buys 2005; Samara et al. 2012; Castellacci and Dizyee 2019). It
incorporates neglected elements in existing modeling such as innovative mindset
and strategic behavior of main actors, institutional and regulatory framework, finan-
cial support, and interactive learning (user–producer, inter-firm collaboration, etc.).
Section 2 provides a review of the instruments and empirical methods used hith-
erto for the quantitative study and assessment of innovation performance at the
national level, highlighting their main strengths and weaknesses. Section 3 describes
the role of helices in the context of NISs and how SD modeling facilitates the depic-
tion of interaction and evolution. Section 4 presents the proposed conceptual frame-
work using a CLD and a SFD. Section 5 concludes with a summary and the main
expected results from the developed framework.

2 Quantitative measurement of innovation performance

at the national level: A review of instruments and empirical
methods

Measuring innovation is paramount for action by managers and policymakers, but

the development information and indicators will always reflect the prevailing con-
ceptualization of reality (Chaminade et al. 2018).
Several instruments using benchmarking analysis rank and compare NISs in
scoreboards. Well-known examples are the European Innovation Scoreboard, the
Global Innovation Index, the Global Competitiveness Index, the EBRD Knowledge
Economy Index, and the Global Knowledge Index (Monteiro and Carayannis 2017;
Chaminade et al. 2018). Scoreboards or ‘league-tables’ consist of dimensions-pillars
and indicators contributing to measuring and evaluating technological capabilities
and innovation at the macro level. Through their annual reports, scoreboards sug-
gest policies that depend on the performance of countries both on individual and
composite (weighted) indicators. The indicators examined are based on various

13
776 A. Vetsikas, Y. Stamboulis

statistical sources to capture the multidimensional nature of technological change

(Archibugi et al. 2009). However, there is a lack of both a theoretical model (‘meas-
urement without theory’) and relationships between the indicators of the pillars-
dimensions (Cirillo et al. 2019). In addition, the selection criteria for indicators have
even been called ‘eclectic, even somewhat arbitrary’ (Schibany and Streicher 2008;
Adam 2014). The indicators used are technology-based (or science-based), cannot
capture tacit knowledge, and underestimate the DUI mode;1 the rankings tend to
focus on science-based learning rather than on learning by doing, using and interact-
ing (Chaminade et al. 2018). Composite indicators, which present a typical problem
of ‘aggregation between apples and oranges’ (e.g., inputs and outputs), are charac-
terized as biased and practically useless for policy. A low rank does not identify the
area(s) necessitating the most urgent policy actions, and a high rank does not neces-
sarily reflect a satisfactory performance (Archibugi et al. 2009; Adam 2014; Grupp
and Schubert 2010).
A slightly more sophisticated analysis than benchmarking of individual countries
is the one performed by some innovation scholars who attempt to group or cluster
countries; a priori (by specific common characteristics) or a posteriori (similar lev-
els of technological capacity or performance around technology clubs) (Chaminade
et al. 2018). Cluster analysis studies (e.g., Radosevic 2004; Balzat and Pyka 2006;
Filippetti and Peyrache 2011; Pinto and Santos Pereira 2013; Mahroum and Al-Saleh
2013; Alnafrah and Zeno 2020) use a wide range of indicators, covering a significant
number of countries and present the importance of some aspects of NISs for growth
and development. However, the studies use short periods of analysis, and the results
(rankings) are similar to scoreboards (i.e., expected cluster groups). Another signifi-
cant ascertainment is the absence of mechanisms linking the different elements of the
system (Chaminade et al. 2018).
Econometric analysis and modeling are also used to shed more light on the rela-
tionship between different dimensions of NISs and growth in a global comparative
perspective (e.g., Furman et al. 2002; Fagerberg and Srholec 2008; Krammer 2009;
Filippetti and Archibugi 2011; Kravtsova and Radosevic 2012; Bartels et al. 2012;
Castellacci and Natera 2013; Duarte and Carvalho 2020). Advanced econometric
methods such as structural equation modeling (SEM), cointegration analysis, error
correction models (ECMs), and causality tests are employed to investigate the main
determinants of innovation performance at the national level. However, panel data
analysis (group of countries with different characteristics) neglects the matters of
‘path dependence’ and ‘historical context’ (Chang 2011). Further, literature stud-
ies usually focus on developed countries and do not consider the DUI mode and the
heterogeneity of actors (e.g., financial system, civil society, academia). Also, main-
stream statistical and econometric approaches often reduce the complexity of NISs
by not decomposing them into subsystems (Paredes-Frigolett et al. 2021).
DEA has been widely used in different cross-country studies on measuring the
efficiency of NISs (Chaminade et al. 2018). Models such as constant returns to scale

1
The DUI mode is an experienced-based mode of learning based on Doing, Using and Interacting, and
refers to know-how and know-who, which is tacit and often highly localized (Jensen et al. 2007).

13
A conceptual framework for modeling heterogeneous actors’… 777

(CRS) or/and variable returns to scale (VRS) are used to measure the efficiency of
NISs and rank them due to their performance (e.g., Nasierowski and Arcelus 2003;
Guan and Chen 2012; Kou et al. 2016; Carayannis et al. 2016; Edquist et al. 2018;
Dobrzanski 2020; Barbero et al. 2021). These studies span a variety of countries
using a broad set of technology-based indicators. Indicators are categorized into
inputs (e.g., R&D expenditure, number of researchers, and population with tertiary
education) and outputs (e.g., number of patents, number of scientific publications,
medium-tech or high-tech exports, and GDP). DEA assumes the innovation process
as linear and does not capture the actors’ role and what happens inside the ‘black
box’ of the innovation process. DEA studies neither employ ‘soft’ variables nor pro-
vide in-depth analysis of ‘small’ countries in the list of efficient.
Recent experience has highlighted the value of simulation to examine complex
social and economic phenomena, such as innovation (Galanakis 2006; Malerba
et al. 2008; Safarzyńska and van den Bergh 2010; Uriona and Grobbelaar 2019).
A growing number of empirical studies employ SD modeling and simulation to
explore NISs dynamics (e.g., Lee and von Tunzelmann 2005; Grobbelaar and Buys
2005; Samara et al. 2012; Castellacci and Dizyee 2019). CLDs are used to formulate
dynamic hypotheses and SFDs to illustrate the behavior of ISs at the national level
(Uriona and Grobbelaar 2019). SD method can capture dynamic complexity (inter-
actions amongst the actors over time), feedback loops, and time delays (Santos et al.
2018). However, SD models concentrate on specific parts of an IS, losing sight of
its big picture and underestimating the DUI mode (Uriona and Grobbelaar 2019). In
addition, existing studies focus on peripheral countries (e.g., Taiwan, South Africa,
Brazil, and Cuba) without providing country comparisons.
The instruments and quantitative methods mentioned above capture specific
innovation matters (e.g., efficiency, efficacy, time lags) and provide some insightful
findings for innovation policies at the national level. However, the methodologies
used to model ISs are still dominated by mainstream statistical analyses, which have
focused on the performance rather than the strategy of NISs (Paredes-Frigolett et al.
2021). Too much of the policy focus has been on optimal performance, where there
is great variety in both system characteristics and the ‘wider setting’ in which the
system is operating (Ricard 2015). Research issues such as path dependencies, DUI
mode, actor heterogeneity, the intensity of interactions, and evolutionary dynamics
within and between innovation subsystems should be examined to capture the evolu-
tionary patterns of NISs’ change.
NIS actors constitute enablers of structural change, and this requires addressing
their decisions, activities, and behaviors (in terms of resource commitment, collabo-
ration, and competitive responses). A plethora of feedback loops emerge through
the interactions of modular subsystems affecting the overall system performance and
its success (Freeman 2002). More sophisticated conceptual frameworks considering
the commitment of resources, the innovative mindset of heterogeneous actors (e.g.,
firms, universities, society, financial system, public authorities), and their mutual
interactions can contribute to our understanding about diverse trajectories of NISs.
In the following sections, we conceive and elaborate on the evolution of NISs as a
process involving both the allocation of actors’ resources and the cognitive and insti-
tutional endowment of the IS (Stamboulis 2007).

13
778 A. Vetsikas, Y. Stamboulis

3 Innovation helices as ‘interacting actors’ and the SD approach

We combine the NISs and ‘N-tuple innovation helices’ frameworks and employ SD
to illustrate and examine the interdependencies and interactions between intrinsi-
cally heterogeneous subsystems. In Section 3.1, we analyze how the various inno-
vation actors interact in the context of NISs, affecting the process of technologi-
cal change. Section 3.2 provides a brief presentation of SD and its central concepts,
ideas, and building blocks.

3.1 Heterogeneous innovation actors within a NIS

NISs are ensembles of organizations and institutions operating at the national level,
which jointly contribute to the development and diffusion of innovations (Adamides
2018). NISs rest on complementarities and synergies operating between different sub-
systems, notably (firms, universities, industrial institutes, infrastructure, finance, voca-
tional training, etc.) (Radosevic 2022). The systemic approach to innovation addresses
the non-linear and heterogeneous nature of innovation processes. Instead of focusing
solely on the number of new product and process innovations, it also encloses R&D
efforts by business firms and public actors as well as the determinants of innovation
such as learning processes, incentive mechanisms, or the availability of skilled labor, as
well as the interplay between organizations and institutions (Balzat and Hanusch 2004).
Amongst the actors involved in the innovation process, the industry is the most
important, as it is the main field of innovation realization (Chung 2002; Freeman
and Soete 1997). Firms play the most crucial role in NISs innovating in interaction
with other firms and knowledge infrastructure (Lundvall 2016). As repositories of
knowledge, firms are the main engines for technological development and wealth
creation (Ranga and Etzkowitz 2013).
National governments affect the innovation process, by strengthening the science
and technology infrastructure (Teubal et al. 1996), via procurement and policy initi-
atives (Mazzucato 2016), and by improving the regulatory framework for innovation
and reduced market uncertainty (Freeman and Soete 1997). Governments establish,
maintain, and adjust institutions such as the legal system, patent system, and tax
system (Liu and White 2001). Their role is affected by the economy’s position in the
world technological frontier (Mahmood and Ruffin 2005).
In a knowledge-based economy, universities – academia in general – are a stra-
tegic actor as a human capital provider, knowledge creators, and seed-bed of new
firms (Etzkowitz et al. 2000; Deiaco et al. 2012). Besides teaching (1st mission) and
research (2nd mission), the ‘entrepreneurial university’ encompasses a ‘third-mis-
sion’ of economic development by fostering links with knowledge users and facili-
tating technology transfer (Perkmann et al. 2013; Feola et al. 2021; Compagnucci
and Spigarelli 2020). The modern university is not only the cradle of talent cultiva-
tion but also the laboratory of scientific research, especially basic research, and the
birthplace of significant scientific and technological achievements. Universities, like
all organizations, are products of their history and can act as catalysts – through
their three core missions – in NISs (Deiaco et al. 2012).

13
A conceptual framework for modeling heterogeneous actors’… 779

In the existing Quintuple Helix Innovation Model, the civil society is associated with
‘media’, ‘creative industries’, ‘culture’, ‘values’, ‘art’, ‘creative class’, and bottom-up
actions while the natural environment has been added as environmental and ecological
issues affect the direction of knowledge production and innovation (Carayannis et al.
2012; Park 2014). Nevertheless, environmental factors are not under any in-system con-
trol and should not be mistaken as system components. Because they do not interact
with the system, environmental factors dependent on the system should not be consid-
ered as part of it (Hughes 1987). In contrast, the financial system constitutes a signifi-
cant individual structural element of NISs due to the varying degrees of appropriability
and uncertainty of innovative efforts (Christensen 2010; Wonglimpiyarat 2011).
A financial system function is necessary to support innovative entrepreneurs
from the early stages of their start-up process with different forms of financing.
Established innovative firms should have efficient access to equity and credit mar-
kets (Tylecote 2007). The foundation and survival of vigorous new firms depend
on a sophisticated private financial system that can support new firms during their
infancy (Mowery and Rosenberg 1993). Different financing forms are required for
every stage of the innovation process; venture capital (VC) is proposed for the R&D
process, stock markets for financing the business development phase, while banking
and bond debts seem appropriate for the period of innovation dissemination in the
form of gross fixed capital formation (Guilhon and Montchaud 2006). Although the
financial sector is represented as a subsystem of NISs in several studies (Lee and
von Tunzelmann 2005; Lee 2006; Samara et al. 2012) is not included in the Quintu-
ple Helix Innovation Model, as a distinct helix. In our approach, the financial sub-
system is conceived as a fifth helix instead of ‘natural environment’.
These subsystems helices are relatively autonomous, and catching up, forging
ahead, or falling behind can be understood through the interactions amongst these
subsystems (Radosevic 2022).

3.2 System dynamics (SD) method

SD is a computer simulation modeling technique used to holistically analyze com-

plex dynamic feedback systems to generate insight for the emergence of nonlinear
behavior and design policies that will improve system performance. SD facilitates
the depiction of emergent behavior2 and system evolution and the examination of
structural differences and changes in a quantifiable manner. Complex systems are
composed of boundedly rational agents, feedback loops, and time delays (Sterman
2000; Uriona and Grobbelaar 2019).
SD assumes that system models are structural accounts of the real-world system
and aims to create endogenous explanations of (non-linear) system behavior and
the challenges that they face (Richardson 2011). The link between system struc-
ture and dynamic behavior is one of the defining elements of SD. In a sense, a

2
‘Emergence’ is a term of scale (Cohen and Harel 2007) and expresses the appearance of novelty or
something previously absent or unprecedented (Lawson 2015). Behavior emerges – in a transcendental
mode (Lawson 1997) – from the interaction of the elements of the system.

13
780 A. Vetsikas, Y. Stamboulis

simulation model can be viewed as an explicit and consistent theory of the behav-
ior it exhibits (Kampmann and Oliva 2009). SD is established as an appropriate
method for developing dynamic models to deal with the structural and functional
complexity of the innovation processes at a national level, as it provides policy-
makers with powerful tools for sense-making, forecasting, and decision-making
(Ricciardi et al. 2020; Allen 2014).
In SD, systems are represented in three ways: (i) CLDs, which highlight the feed-
back structure of the system; (ii) SFDs, which serve to understand the dynamic behav-
ior of the system through computer simulations; and (iii) the mathematical notation
underlying the SFDs, which are partial differential equations (Sterman 2000).
From a SD perspective, a system’s structure consists of stocks, flows, and feed-
back loops (reinforcing and balancing). Stock variables (levels) are the accumula-
tions within the system, while the flow variables (rates) represent the activities in the
system, which result from the decision-making process. Feedback (interaction) is
what makes systems dynamic; without such feedback, systems are static.
Nowadays, several schools of heterodox economics (e.g., evolutionary, institu-
tional, behavioral, post-Keynesian) utilize SD models to explain systemic and emer-
gent phenomena in complex systems (Radzicki 2020).
The combination of N-tuple innovation helices and SD facilitates a meaningful
way to explore and understand the functional dynamics that emerge amongst hetero-
geneous actors, meeting the need for a representation of real-world dynamics that
appropriately describes NIS trajectories so that future trends can be estimated (Phil-
lips and Linstone 2016).

4 A helix‑based framework for modeling the dynamics of NISs

Although most countries are striving to develop a more holistic innovation policy,
linearity still prevails due to the emphasis on the ‘commercialization of science’
through a simplistic technology-push model (Edquist 2014; Diercks et al. 2019).
Linear approaches and ‘innovation scoreboards’ dominate the discourse, underplay-
ing the endogenous and evolutionary dynamics of NISs.
We develop a conceptual framework for modeling heterogeneous actors’ behav-
ior in NISs. The framework goes beyond the static illustration of scoreboards and
linear models and illustrates how individual parameter changes – in one helix of the
system – may generate non-linear effects throughout. Thus, we build a helix-based
framework of NISs representing interacting helices and their corresponding struc-
ture and interactions. We expect to identify the salients (the most efficient or effec-
tive components of the system) and the reverse salients (components out of phase
or falling behind in terms of efficiency) that emerge through the evolution of NISs.
In Section 4.1, we classify an indicative set of innovation-related indicators that
contribute to the construction of our conceptual framework. In Section 4.2, we
develop our dynamic hypothesis using a CLD that illustrates the reinforcing and bal-
ancing loops of interaction, cooperation, and coordination amongst and within the
five subsystems helices. In Section 4.3, we present a SFD that depicts the mutual
interactions between the various subsystems.

13
A conceptual framework for modeling heterogeneous actors’… 781

4.1 Classification of innovation‑related indicators into five helices

Recent studies have introduced preliminary taxonomies of innovation-related indica-

tors in helices to examine the role and performance of subjects within NISs (Cirillo
et al. 2019; Maruccia et al. 2020; Barcellos-Paula et al. 2021; Paredes-Frigolett et al.
2021; Vetsikas 2023).
Similarly, we classify a broad range of indicators from various databases (World
Bank, OECD, Eurostat, etc.) into helices to develop a concise and complete dash-
board of innovation system behavior and performance. These indicators cover the
key dimensions of various subsystems’ resources, activities, framework conditions,
institutions, and outcomes.
In Table 1, we highlight the most important indicators of the parameters used in the
analysis that follows in Section 4.2 and 4.3. Indicators capturing interactions between
different actors (e.g., university–industry collaborations, tertiary education attainment,
government expenditure on tertiary education) are also classified along the five helices.
Next, we develop a CLD to depict the interactions and interdependencies amongst
various elements in NISs and a SFD to offer a slightly sounder basis for inference of
behavior.

4.2 Dynamic hypothesis

CLDs capture the structure of a system in SD methodology. A CLD3 represents the

key feedback mechanisms. These mechanisms are either negative feedback (balanc-
ing) loops or positive feedback (reinforcing) loops.4
Figure 1 illustrates our dynamic hypothesis of the mechanics of NISs in the form
of a CLD that illustrates the interactions within and amongst the helices of industry,
academia, government, society, and finance. We notice the most important reinforc-
ing and balancing feedback loops.
The first reinforcing loop (R1) shows that more investment in business R&D
(BER&D) attracts personnel in R&D activities that enhance accumulated knowledge
in firms. Increased knowledge leads to more innovation outcomes (new products
and/or processes), which lead to returns on innovation activities (e.g., revenues and
business profits) which, eventually, confirm the behavior of resource commitment to
innovation, leading to further business R&D investments (Galanakis 2006; Lee and
von Tunzelmann 2005; Lee 2006; Samara et al. 2012; Choi et al. 2016).
Furthermore, returns on innovation can lead to an increase in the flow of VC that
amplifies the growth of new innovative firms (e.g., startups). The higher number of
firms engaged in innovation increases business R&D investments, creating a second
reinforcing loop (R2) besides the first one (R1).

3
CLDs allow to map systems, identify loops, determine their polarity (+ and R or – and B for reinforc-
ing and balancing loops respectively), trace them, and communicate about specific loops or sets of loops
(Sterman 2000; Pruyt 2013).
4
Reinforcing or positive feedback loops may be perceived as a trend reinforcing mechanism which may
lead to escalation and eventually loss of control, while a balancing or negative feedback loop should be
understood as counterbalancing one, often interpreted as a force pulling towards a target (Pruyt 2013).

13
782 A. Vetsikas, Y. Stamboulis

Table 1  Summary of indicators classified in helices

Helices Indicators

Industry Business R&D expenditure, Business R&D personnel, SMEs with product or process
innovation, GDP per hour worked, Medium and high-tech exports, Sales of new-to-mar-
ket/new-to-firm innovations, Global value chains, SMEs collaborating with others, Pat-
ents, Industrial designs, Charges for the use of intellectual property, University–industry
collaboration in R&D, Intensity of competition, Distance from technology frontier, Birth
rate of firms, Death rate of firms
Academia Higher education R&D expenditure, New entrants (Bachelor, Master, PhD), Students
enrolled in tertiary education (Bachelor, Master, PhD), Tertiary graduates (Bachelor,
Master, PhD), Academic staff, Scientific articles, Citations, Number of projects, Quality
of research institutions, Quality of the education system
Government Government expenditure on tertiary education, Percentage of Business R&D expenditure
financed by government, Government effectiveness, Rule of law, Regulatory quality,
Political stability and absence of violence, Control of corruption, Protection of property
rights, Government procurement of advanced tech products
Society Employment with tertiary education, Unemployment with tertiary education, Immigration
& Emigration (flows), Buyer sophistication, Lifelong learning, Population completed
tertiary education, Individuals with above basic overall digital skills, Disposable income
Finance Venture capital (seed, start-up, later stage)

Similar to R1 and R2 loops, a third reinforcing loop (R3) is identified through the
causal links ‘number of innovative firms-BER&D-Business R&D personnel-BERD
activity-Accumulated knowledge in firms-innovation outcomes-returns on innova-
tion-number of innovative firms’. As the number of innovative firms increases, there
is an expected increase of investment in (commitment to) R&D. An increase in busi-
ness R&D investment leads to an increase in R&D personnel in firms. Increased per-
sonnel engaged in R&D activity enhances accumulated knowledge in firms, which
further increases innovation outcomes. An increase in innovation results leads to an
increase in returns on innovation (conditional to complementary investment). As
these returns are increasing, there is a greater stimulus for innovative firms.
The fourth reinforcing loop (R4) illustrates that more investment in higher edu-
cation R&D (HER&D) leads to an increase in academic personnel (researchers),
and their increased number enhances research activity (e.g., new scientific articles,
participation in projects) that boosts new research results. Increased research results
enhance the quality of the academic sector, which further attracts new higher educa-
tion investments for research (Rodríguez and Navarro-Chávez 2015; Rad et al. 2015).
The fifth reinforcing loop (R5) shows that an increased level of tertiary education
(population with tertiary education) attainment would lead to greater demand for a
better quality of institutions (Marginson 2016; Botero et al. 2013). Improved qual-
ity of institutions decreases the trend of skilled workers (Ngoma and Ismail 2013;
Cooray and Schneider 2016; Dimant et al. 2013) to migrate and thus reduces ‘brain
drain’ which – if unimpeded – dwindles the pool of the population with tertiary
education (Tritah 2008).
A similar reinforcing loop (R6) describes the causal links between ‘popu-
lation with tertiary education-pressure for a better quality of institutions-
quality of institutions-number of innovative firms- BER&D – business R&D

13
A conceptual framework for modeling heterogeneous actors’… 783

Fig. 1  Causal loop diagram (CLD) of the interactions between the helices of a NIS (The CLD has been
designed on the Vensim system dynamics software. We use different colors corresponding to the five heli-
ces to facilitate their distinction: orange for industry, yellow for academia, light blue for government,
green for society, and purple for finance. Bold characters denote decision/policy variables (e.g., public
expenditure on tertiary education and R&D, subsidies to BER&D).)

personnel- unemployment-brain drain- population with tertiary education’. As

the increasing education level of individuals leads to an increase in the quality
of institutions, new firms are established in an environment with more favorable
conditions (Cieślik and Goczek 2018; Farla et al. 2016; Alam et al. 2019). New
and existing innovative firms would increase their investment in R&D – in terms
of resource commitment – attracting new business R&D personnel. An increase
in business R&D personnel in firms leads to a decrease in unemployment and
a consequent decrease in the migration of high-skilled workers. Therefore, a
decrease in ‘brain drain’ leads to an increase in the country’s tertiary graduates
(key knowledge stock).
The seventh reinforcing feedback loop (R7) explains that an increase in innova-
tion outcomes leads to a decrease in the distance of a country from the global tech-
nological frontier and more developed NISs (i.e., successful catch-up). As this dis-
tance is increasing, a country presents fewer opportunities to participate in global
value chains (GVCs) (Ye et al. 2020). However, intensive participation in interna-
tional networks and value chains enhances collaborations and learning by interac-
tions from abroad (external sources of knowledge) which further boosts the stock
of knowledge in firms (Goñi and Maloney 2017; Tajoli and Felice 2018; Kergroach
2019). As aforementioned, increasing knowledge promotes innovation outcomes.

13
784 A. Vetsikas, Y. Stamboulis

The eighth reinforcing feedback loop (R8) constitutes an extension of the first rein-
forcing loop (R1). It includes the connection of the population with tertiary educa-
tion and the sophistication of users. In CLD, it is described by the following linkages
‘population with tertiary education-sophistication of users-interactive learning-accu-
mulated knowledge in firms-innovation outcomes-returns on innovation-BER&D-busi-
ness R&D personnel-unemployment-brain drain-population with tertiary education’.
We should emphasize that all the above reinforcing loops may work in both direc-
tions leading to an increase or decrease of the relevant behaviors and results.
Regarding the balancing feedback loops, the first one (B1) describes how more
innovation outcomes decrease the distance of a country from the global technologi-
cal frontier while an increased distance from the frontier would decrease the absorp-
tion of technology and exports (Özak 2018). Increased medium and high technology
exports further enhance the exposure to more intense competition, and intensifying
competition between firms decreases the returns on innovation.
A second balancing feedback loop (B2) is identified through the relationship of
intensive competition, returns on innovation, and the number of innovative firms.
An increase of the latter may consequently lead to increased competition (Amendola
et al. 2000), and increased competition decreases returns on innovation. As afore-
mentioned, increased returns on innovation (as ‘windows of opportunity’) boost the
creation of new innovative firms.
In Section 4.3, we further develop our dynamic hypothesis using a SFD, as CLDs
do not always explain well how flows influence stocks and how time delays may
emerge. Furthermore, CLDs may lead to mislabeling of loops and do not provide a
sound basis for the rigorous deduction of behavior (Lane 2000; Pruyt 2013).

4.3 Development of SFD for the interactions of heterogeneous actors

SFDs are used to illustrate the mathematical model and the differential equations (Oli-
vares-Aguila and ElMaraghy 2021). In SFDs, stocks are symbolized by rectangles,
flow variables by valves, converters (intermediate variables) by circles, and constants
(model parameters) by rhombuses (Sterman 2000; Lane 2000). In Fig. 2, we present
the SFD where the five subsystems-helices operate and interact within a NIS. We need
to consider that each actor or subsystem is made up of diverse elements and that this
micro-diversity changes over time because of successes and failures (Allen 2014).
Adopting the evolutionary logic of variation-selection-retention (VSR), we consider
that actors decide each period on the allocation of resources (continuous trial-and-
error learning) amongst a variety of choices (Safarzyńska and van den Bergh 2013).
Innovation actors (much like biological organisms) try various choices (responses) in
a competitive situation until the one found is sufficiently successful in achieving their
goal. The commitment of resources and the corresponding choices may lead either to
success (e.g., increasing returns on innovation, new products or processes, increasing
research outcomes) or failure (e.g., increasing distance from world technological fron-
tier, decreasing participation in projects and global value chains).
Within an IS, a percentage of firms engage in innovation activities and com-
mit the required resources (financial and human). The R&D is the key activity that

13
A conceptual framework for modeling heterogeneous actors’… 785

Fig. 2  Stock-and-flow diagram (SFD) of the interactions between the helices of a NIS (The SFD has
been designed on the Powersim system dynamics simulation software. We use orange for industry, yel-
low for academia, light blue for government, green for society and purple for finance. Decision/policy
variables (e.g., public expenditure on tertiary education, subsidies to BER&D) are represented by rhom-
buses. The Bordeaux-colored rhombus depicts the exogenous nature of technical change.)

creates know-how. An organization accumulates knowledge through R&D. Firms

usually do not innovate in isolation but interact with other organizations through
complex relations, often characterized by several loops of reciprocity and feedback
mechanisms (Edquist 2011). Both product and process knowledge result from net-
working and interactive learning with other actors such as universities, government,
firms, and end-users (Malerba 1992; Nooteboom 1999; Edquist 2011).
We illustrate accumulated knowledge (product and process related) as an array
stock variable while learning is an array flow variable adding more knowledge to the
stock (see Fig. 2). In order to access any kind of knowledge, firms should improve their
ability to learn, often referred to as ‘appropriability’ (Cohen and Levinthal 1990). This
ability is achieved by accumulating knowledge via different learning modes (Malerba
1992; Muscio 2007). We show that accumulated knowledge enhances the absorptive
capacity of firms. The outflow of the array stock variable (accumulated knowledge)
represents the accumulated knowledge depreciation because of exogenous technical
change (Grobbelaar and Buys 2005). As firms increase their accumulated knowledge,
they can reduce their distance from the global technological frontier and participate
in GVCs increasing the rate of absorption of external knowledge and increasing their
domestic sales and exports (medium and high tech).
In the array flow variable ‘BERD projects’, the dimension of process knowledge
affects new processes related to innovation (such as changes in techniques, equip-
ment, and software). These processes are associated with increased productivity
and lower production costs. Firms achieve efficiency of committed resources and
increased returns that enhance further business R&D investment. Firms seek to
minimize costs through process innovation to maximize the efficiency of resources
employed in production (Choi et al. 2016). In contrast to process innovation, the

13
786 A. Vetsikas, Y. Stamboulis

dimension of product innovation – in BERD projects – breeds new differentiated

products. New products with upgrading quality and brand name further increase the
domestic sales and the medium and high technology exports. In both cases, new
sales and technology exports are associated with BERD projects through accumu-
lated knowledge and decreasing distance from the global technological frontier. The
distinction (product and process innovation) is widely illustrated in several SD stud-
ies (Samara et al. 2012; Lee and von Tunzelmann 2005; Lee 2006; Choi et al. 2016).
Previous SD models do not include intellectual property (e.g., patents, industrial
designs) of R&D activities in their analysis. As innovation activity (BERD projects)
increases and depending on their strategy (e.g., offensive, defensive, opportunistic)
firms have greater incentives to protect their innovation activity results. The stock
of ‘IPs’ (array variable: patents and industrial designs) generates royalties that pro-
duce returns on innovation (Kondo 1999). Furthermore, as the firm generates and
encloses more knowledge, it will be more attractive to VC (Niosi 2003).
As mentioned in the second balancing feedback loop (B2), increasing returns on
innovation encourage the creation of innovative firms. In mathematical terms, the total
number of active innovative firms (stock variable) is the integral of the birth rate of new
innovative firms (inflow) and the death rate of existing ones (outflow). In a sense, we
aspire to capture a part of the ‘creative destruction’ process employing these two flow
variables. The average rate of firm creation and the rate of firm destruction is the most
used measure in the literature on firm and employment dynamics (Aghion et al. 2021).
In the helix of academia, universities and research institutes are central actors
in education, research, and ‘third mission’ (Sánchez-Barrioluengo 2014). The mis-
sion of teaching is associated with the development of human capital in the form
of skilled labor (Bayuo et al. 2020). Increases in government expenditure on ter-
tiary education stimulate the entrants of potential graduates from academia. In each
stage, tertiary graduates leave the subsystem and constitute potential knowledge
workers (in the helix of society). Many of them may continue their careers inside
the academic sector and across different stages (postgraduate, PhD, and new aca-
demic staff) in a resource chain process (Frølich et al. 2010; Dangerfield 2014). The
process of new entrants in different levels of tertiary education and new graduates
(as potential knowledge workers) is represented by the connection of array flows
(inflows and outflows) with the two array stock variables entitled ‘tertiary students
in different levels’ and ‘potential knowledge workers’.
Regarding the research mission, we consider research activity and experience
(measured by the number of publications and research projects) as research outcomes.
Increasing participation in international projects enhances the absorption of universi-
ties’ external knowledge, contributing to new scientific publications and consequent
citations. The increasing number of citations improves the quality and reputation of
the academic sector (Bejan et al. 2020). Increasing research activity (e.g., participa-
tion in projects and knowledge networks, new publications) stimulates accumulated
knowledge in academia and its quality, which further increases collaboration (usu-
ally in R&D) between firms and academia. These collaborations facilitate interactive
learning, which, in turn, is the basis for innovation (Edquist 2011).
In recent years, universities have been urged to expand their scope from teaching
and research to a ‘third mission’, described as ‘contribution to society’ (Compagnucci

13
A conceptual framework for modeling heterogeneous actors’… 787

and Spigarelli 2020). Research outcomes usually a result of completed projects can
spur – amongst others – the creation of new innovative firms (e.g., start-ups and
spinoffs), accelerating the process of creative destruction. Furthermore, universities
engaged in patenting activity may increase their revenues via royalties. This means
that as the stock of academic patents increases, universities enhance their income
from academic licensing, and then they may re-invest in basic research.
Regarding the role of government, its increasing ability to formulate and imple-
ment policies and regulations would increase the perception of the extent to which
agents have confidence in and abide by the rules of society (Maruccia et al. 2020).
An improvement in a country’s institutional environment (e.g., control of corrup-
tion, bureaucratic quality, political stability) seems to impact positively both on the
likelihood of establishing high tech firms (Pereira and Temouri 2018; Pereira et al.
2020) and on attractiveness to VC investment (Bustamante et al. 2021). Moreover,
the higher quality of (hard) institutions would also reduce the drive for brain drain
and lead to the retention of those with high levels of educational attainment (Cooray
and Schneider 2016; Dimant et al. 2013). For instance, the high political corruption
and instability, and the lack of transparency seem to act as catalysts for brain drain in
several countries (Panagiotakopoulos 2020; Mlambo and Adetiba 2019).
Demand-side activities such as innovative public procurement can contribute to inter-
active learning, the formation of new product markets, and the articulation of quality
requirements (Edquist 2011; Radicic 2019). From the supply side, ‘traditional’ innova-
tion policy instruments such as fiscal incentives for R&D and direct subsidies to business
R&D aim to the creation of new knowledge and innovation (Edler and Fagerberg 2017).
In Fig. 2, we include these two sides of governments’ innovation activities using the var-
iable ‘public technology procurement’, as well the variable ‘subsidies to BER&D’.
Regarding the society helix, tertiary education graduates constitute a central stock
of knowledge (potential knowledge workers) in a NIS and one of the main mecha-
nisms of knowledge spillovers from academia. Citizens with higher digital literacy
seem to be more sophisticated users (rather than passive buyers) and develop their
skills and capabilities via lifelong learning. In Fig. 2, we illustrate digital literacy and
lifelong learning as factors influencing buyers’ sophistication. We assume that per-
sons with higher digital literacy and lifelong learning have higher capabilities to par-
ticipate in interactive user–producer learning (depicted in the industry helix) (Von
Hippel 1986; Lundvall 1992). In the SFD, we consider that higher sophistication
of buyers amplifies DUI mode learning. Also, individuals with higher educational
attainment are significantly more likely to lodge complaints against their government
about general services, police abuse, and corruption. For instance, NISs with higher
educational attainment (such as Australia, New Zealand, Norway, Finland, Denmark,
etc.) have higher control of corruption, political stability, and government effective-
ness than countries with low education level like Pakistan, Senegal, Sri Lanka, and
Sudan (Botero et al. 2013). Hence, in our analysis, the variable of the population with
tertiary education is associated with the variable of quality of institutions.
Financial capital is indispensable in the early stages of business development, and
the financial helix is essential for technology development and commercialization
(Wonglimpiyarat 2011). We concentrate on VC and its contribution to the creation
of new innovative firms. The VC pool (represented by a stock variable in Fig. 2)

13
788 A. Vetsikas, Y. Stamboulis

refers to available capital that may be invested in new innovative firms in exchange
for equity. Increasing returns on innovation enhance venture capitalists’ value and
income and their ‘appetite’ for new investments.
At any moment in the system’s evolution, some feedback loops may be highly influ-
ential, and others may be inactive. As the system enters new states (surpassing criti-
cal mass in accumulations),5 latent loops with limited impact on the system’s behavior
until then may suddenly become dominant, causing qualitative structural changes in
the mode of system behavior (Richardson 1995). This needs to be validated in future
research, with the use of simulation analysis, which is above and beyond the scope of
this paper. Simulation analysis can allow policymakers to better understand the results
of various choices (via continuous experimentation), the selection amongst policy alter-
natives, and the retention of successful resource commitment strategies. For example,
the exploration of various combinations of innovation subsidy policies for business and
investment in HER&D will produce different results depending on the scenarios for
‘external’ – to the S&T system – conditions (e.g., distance from technological frontier,
rate of technical change, sophistication of users) or actor behavior (e.g., the resources
committed to R&D from business as a response to subsidies, competition, etc.).
In comparison to other SD studies, which focus on specific parts of ISs (such as
R&D policies, innovation diffusion policies, and science and technology policies)
(Uriona and Grobbelaar 2019), our approach offers a broader view of NISs incorpo-
rating important issues such as the ‘creative destruction’ process, different sources
of learning (domestic and global) for firms, the ‘openness’ of a country and its par-
ticipation in GVCs, the quality of institutions, the ‘three missions’ of academia, the
active involvement of sophisticated users, and the VC role.
There are weak or lagged parts between and within helices that operate as obsta-
cles for the virtuous cycle of knowledge affecting the evolutionary path of NISs.
These parts may concern knowledge creation and exploitation, R&D activity, inter-
active learning, participation in GVCs, commercialization, regulation, financial sup-
port, sophisticated demand, etc. The crucial issue is to understand the whole system
behavior through the dominant and weak feedback loops.
Thus, improvements in innovation policies based on identified weak/lagged parts
and leverage points of the system may contribute substantially to the overall system
performance. What-if analysis, through alternative scenarios, can lead to the shift
from ‘static innovation scoreboards’ to ‘dynamic innovation dashboards’ for innova-
tion policies and new tools for innovation management.

5 Conclusions

The quantitative analysis and measurement of innovation remain a crucial but highly
arduous effort for researchers and scholars. In this paper, we reviewed the main
approaches and empirical methods that examine the performance (effectiveness,

5
‘Critical mass’ refers to a sufficient level (size of firms, number of scientists, number of R&D per-
sonnel, funding, etc.) that allows a NIS to be able to take up and effectively harness new technological
knowledge (Archibugi et al. 1999).

13
A conceptual framework for modeling heterogeneous actors’… 789

efficiency, efficacy, etc.) of NISs, highlighting their main strengths and weaknesses.
Despite the aspiration of holistic approaches, the dominant empirical methods fail
to address crucial aspects of innovation such as actor heterogeneity, the intensity of
interactions, complexity, time delays, and endogenous dynamics in NISs.
To overcome these shortcomings, we developed a generic conceptual framework for
modeling heterogeneous actors’ behavior in NISs. We considered five helices as ‘aggre-
gate actors-subsystems’ that evolve and interact asymmetrically in a co-opetitive game,
in the production, diffusion, and use of new knowledge. Different strategies and activi-
ties of actors take place based on the resources they have committed to them (Pisano
1997). Actors’ resource commitment depends on expectations and previous experience,
while strategy involves making choices on resource commitments (Stamboulis 2007).
Within a NIS, the absence of synergies between actors creates lagged or weak
elements (areas) of the system affecting its overall performance and success. We
consider that the intensity of spatio-temporal interactions within and between sub-
systems and the level of cooperation and coordination amongst them can explain the
evolutionary trajectories of NISs.
Each NIS has its own history, path dependence, institutions, routines that deter-
mine its evolution. Mapping the dominant loops – in SD terminology – may contrib-
ute to understanding the varying evolutionary patterns of NISs’ change. We expect
to identify the leverage points and suggest system-oriented policies for overall sys-
tem performance improvement through what-if analysis.
Our proposed conceptual framework allows us to develop concise models and
understand the impact of policy choices and actor behaviors. We can identify system
failures (e.g., infrastructural, institutional, interaction, and capabilities’) and design pol-
icy measures accordingly (Woolthuis et al. 2005). In comparison to existing SD mod-
eling attempts (either conceptual or with formal validation), our integrated framework
introduces new modeling elements in NISs such as innovative mindset and strategic
behavior in terms of resource commitment, institutional and regulatory framework (i.e.,
collaboration culture, quality of institutions, IPRs), financial support, policy effects,
user-producer interaction, and interactive learning, exogenous factors, and global tech-
nical change. It enables quantification for comparative analysis, scenario (what-if) pro-
cesses, and ‘cost-benefit’ calculations by investigating different possible impacts of
specific interventions. Through scenario analysis, we may identify successful policy
interventions (what works and what does not) in terms of resource commitment.
In further research, we aim to validate the structure of the model with histori-
cal data, examine diverse scenarios for selected NISs, and provide some compara-
tive analysis (e.g., North vs. South, innovation leaders vs. emerging innovators).
Moreover, we could examine the diversity6 (considering the properties of variety,
balance, and disparity) of structural elements (e.g., firms, and universities) within
NISs for capturing the evolutionary patterns over time. The framework may also be
enriched by detailing further the financial helix and incorporating factors related to

6
For instance, Stirling (2007; 2010) suggests a general quantitative non-parametric diversity heuristic
that allows exploration of trade-offs between diversity and performance including consideration of sys-
tem constraints and interactions.

13
790 A. Vetsikas, Y. Stamboulis

the country’s economic growth (GDP), complementary assets (e.g., consultancy and
technical services – KIBS), demand-side activities (e.g., quality requirements, envi-
ronmental-social awareness), and supporting functions for innovation and entrepre-
neurship (e.g., incubation activities, pattern formation). At a different level of analy-
sis, depending on the availability of relevant data, the proposed framework may be
adapted at the regional or sectoral level.
Acknowledgements The authors would like to thank the two anonymous reviewers for their very helpful
critical feedback in an earlier version of the paper. The authors would also like to thank the participants
of the 18th International Schumpeter Society Conference, July 8–10, 2021, that took place in the Luiss
Guido Carli University, for sharing freely their thoughts, remarks, and suggestions, for an earlier ver-
sion of this paper that has been presented at the conference. Any remaining errors are solely the authors’
responsibility.

Funding Open access funding provided by HEAL-Link Greece. This work was partially supported by
the Hellenic Foundation for Research and Innovation (HFRI) under the HFRI PhD Fellowship grant 455.

Data availability The present paper did not require any use, generation, and analysis of data.

Declarations The present paper did not require any interaction or experimentation that demands ethical
compliance.

Conflict of interest The authors declare that they have no conflicts of interest.

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License,
which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as
you give appropriate credit to the original author(s) and the source, provide a link to the Creative Com-
mons licence, and indicate if changes were made. The images or other third party material in this article
are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the
material. If material is not included in the article’s Creative Commons licence and your intended use is
not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission
directly from the copyright holder. To view a copy of this licence, visit http://​creat​iveco​mmons.​org/​licen​
ses/​by/4.​0/.

References
Adam F (2014) Measuring national innovation performance: the Innovation Union Scoreboard revisited.
Springer, Ljubljana
Adamides E (2018) Critical realism in the analysis of national innovation systems. In: Vliamos S, Zoumbou-
lakis M (eds) Institutionalist Perspectives on Development. Palgrave Macmillan, Cham, pp 105–124
Aghion P, Antonin C, Bunel S (2021) The power of creative destruction. Harvard University Press,
Cambridge
Alam A, Uddin M, Yazdifar H (2019) Institutional determinants of R&D investment: Evidence from
emerging markets. Technol Forecast Soc Change 138:34–44. https://​doi.​org/​10.​1016/j.​techf​ore.​
2018.​08.​007
Allen PM (2014) Evolution: complexity, uncertainty and innovation. J Evol Econ 24:265–289. https://​doi.​
org/​10.​1007/​s00191-​014-​0340-1
Alnafrah I, Zeno B (2020) A new comparative model for national innovation systems based on machine
learning classification techniques. Innov Dev 10(1):45–66. https://​doi.​org/​10.​1080/​21579​30X.​
2018.​15641​24
Amendola M, Gaffard JL, Musso P (2000) Competition, innovation and increasing returns. Econ Innov
New Technol 9(2):149–181. https://​doi.​org/​10.​1080/​10438​59000​00000​07
Archibugi D, Howells J, Michie J (1999) Innovation systems in a global economy. Technol Anal Strat
Manag 11(4):527–539. https://​doi.​org/​10.​1080/​09537​32991​07311

13
A conceptual framework for modeling heterogeneous actors’… 791

Archibugi D, Denni M, Filippetti A (2009) The technological capabilities of nations: The state of the
art of synthetic indicators. Technol Forecast Soc Change 76(7):917–931. https://​doi.​org/​10.​1016/j.​
techf​ore.​2009.​01.​002
Balzat M, Hanusch H (2004) Recent trends in the research on national innovation systems. J Evol Econ
14(2):197–210. https://​doi.​org/​10.​1007/​s00191-​004-​0187-y
Balzat M, Pyka A (2006) Mapping national innovation systems in the OECD area. Int J Technol Glob
2(1–2):158–176. https://​doi.​org/​10.​1504/​IJTG.​2006.​009132
Barbero J, Zabala-Iturriagagoitia JM, Zofío JL (2021) Is more always better? On the relevance of decreas-
ing returns to scale on innovation. Technovation 107:102314. https://​doi.​org/​10.​1016/j.​techn​ovati​
on.​2021.​102314
Barcellos-Paula L, De la Vega I, Gil-Lafuente AM (2021) The Quintuple Helix of Innovation Model
and the SDGs: Latin-American Countries’ Case and Its Forgotten Effects. Mathematics 9(4):416.
https://​doi.​org/​10.​3390/​math9​040416
Bartels FL, Voss H, Lederer S, Bachtrog C (2012) Determinants of National Innovation Systems: Policy
implications for developing countries. Innovation 14(1):2–18. https://​doi.​org/​10.​5172/​impp.​2012.​
14.1.2
Bayuo BB, Chaminade C, Göransson B (2020) Unpacking the role of universities in the emergence,
development and impact of social innovations–A systematic review of the literature. Technol Fore-
cast Soc Change 155:120030. https://​doi.​org/​10.​1016/j.​techf​ore.​2020.​120030
Bejan A, Gunes U, Sahin B (2020) University rankings: Quality, size and permanence. Eur Rev
28(4):537–558. https://​doi.​org/​10.​1017/​S1062​79872​00000​6X
Botero J, Ponce A, Shleifer A (2013) Education, complaints, and accountability. J Law Econ 56(4):959–
996. https://​doi.​org/​10.​1086/​674133
Bustamante CV, Mingo S, Matusik SF (2021) Institutions and venture capital market creation: The case
of an emerging market. J Bus Res 127:1–12. https://​doi.​org/​10.​1016/j.​jbusr​es.​2021.​01.​008
Carayannis EG, Barth TD, Campbell DF (2012) The Quintuple Helix innovation model: global warm-
ing as a challenge and driver for innovation. J Innov Entrep 1(1):1–12. https://​doi.​org/​10.​1186/​
2192-​5372-1-2
Carayannis EG, Grigoroudis E, Goletsis Y (2016) A multilevel and multistage efficiency evaluation of
innovation systems: A multiobjective DEA approach. Expert Syst Applic 62:63–80. https://​doi.​org/​
10.​1016/j.​eswa.​2016.​06.​017
Castellacci F, Dizyee K (2019) Policy strategies for economic development in Cuba: A simulation model
analysis. Dev Policy Rev 37(6):769–788. https://​doi.​org/​10.​1111/​dpr.​12419
Castellacci F, Natera JM (2013) The dynamics of national innovation systems: A panel cointegration
analysis of the coevolution between innovative capability and absorptive capacity. Res Policy
42(3):579–594. https://​doi.​org/​10.​1016/j.​respol.​2012.​10.​006
Chaminade C, Lundvall BÅ, Haneef S (2018) Advanced Introduction to National Innovation Systems.
Edward Elgar Publishing, Cheltenham
Chang HJ (2011) Institutions and economic development: theory, policy and history. J Inst Econ
7(4):473–498. https://​doi.​org/​10.​1017/​S1744​13741​00003​78
Choi K, Narasimhan R, Kim SW (2016) Opening the technological innovation black box: The case of the elec-
tronics industry in Korea. Eur J Oper Res 250(1):192–203. https://​doi.​org/​10.​1016/j.​ejor.​2015.​08.​054
Christensen JL (2010) The role of finance in national systems of innovation. In: Lundvall BÅ (ed)
National systems of innovation: Toward a theory of innovation and interactive learning, 2nd edn.
Anthem Press, London, pp 151–172
Chung S (2002) Building a national innovation system through regional innovation systems. Technova-
tion 22(8):485–491. https://​doi.​org/​10.​1016/​S0166-​4972(01)​00035-9
Cieślik A, Goczek Ł (2018) Control of corruption, international investment, and economic growth–Evi-
dence from panel data. World Dev 103:323–335. https://​doi.​org/​10.​1016/j.​world​dev.​2017.​10.​028
Cirillo V, Martinelli A, Nuvolari A, Tranchero M (2019) Only one way to skin a cat? Heterogeneity and
equifinality in European national innovation systems. Res Policy 48(4):905–922. https://​doi.​org/​10.​
1016/j.​respol.​2018.​10.​012
Cohen I, Harel D (2007) Explaining a complex living system: dynamics, multi-scaling and emergence. J
R Soc Interface 4(13):175–182. https://​doi.​org/​10.​1098/​rsif.​2006.​0173
Cohen W, Levinthal D (1990) Absorptive Capacity: A New Perspective on Learning and Innovation.
Adm Sci Q 35:128–152. https://​doi.​org/​10.​2307/​23935​53

13
792 A. Vetsikas, Y. Stamboulis

Compagnucci L, Spigarelli F (2020) The Third Mission of the university: A systematic literature review
on potentials and constraints. Technol Forecast Soc Change 161:120284. https://​doi.​org/​10.​1016/j.​
techf​ore.​2020.​120284
Cooray A, Schneider F (2016) Does corruption promote emigration? An empirical examination. J Popul
Econ 29(1):293–310. https://​doi.​org/​10.​1007/​s00148-​015-​0563-y
Dangerfield B (2014) Systems thinking and system dynamics: A primer. In: Brailsford S, Churilov L,
Dangerfield B (eds) Discrete-event simulation and system dynamics for management decision
making. Wiley, Chichester, pp 26–51
Dawid H (2015) Modeling the economy as a complex system. In: Alves-Furtado B, Sakowski P, Tovolli
M (eds) Modeling Complex Systems for Public Policies. IPEA, Brasilia, pp 191–216
Deiaco E, Hughes A, McKelvey M (2012) Universities as strategic actors in the knowledge economy.
Cambridge J Econ 36(3):525–541. https://​doi.​org/​10.​1093/​cje/​bes024
Diercks G, Larsen H, Steward F (2019) Transformative innovation policy: Addressing variety in an
emerging policy paradigm. Res Policy 48(4):880–894. https://​doi.​org/​10.​1016/j.​respol.​2018.​10.​028
Dimant E, Krieger T, Meierrieks D (2013) The effect of corruption on migration, 1985–2000. Appl Econ
Lett 20(13):1270–1274. https://​doi.​org/​10.​1080/​13504​851.​2013.​806776
Dobrzanski P (2020) The efficiency of spending on R&D in Latin America region. Appl Econ
52(46):5020–5034. https://​doi.​org/​10.​1080/​00036​846.​2020.​17529​00
Duarte MP, Carvalho FM (2020) The drivers of national system of innovation in portugal: A panel data
analysis. J Inform Syst Eng Manag 5(2):em0114. https://​doi.​org/​10.​29333/​jisem/​8248
Edler J, Fagerberg J (2017) Innovation policy: what, why, and how. Oxford Rev Econ Policy 33(1):2–23.
https://​doi.​org/​10.​1093/​oxrep/​grx001
Edquist C (2011) Design of innovation policy through diagnostic analysis: identification of systemic
problems (or failures). Ind Corp Change 20(6):1725–1753. https://​doi.​org/​10.​1093/​icc/​dtr060
Edquist C (2014) Striving towards a holistic innovation policy in European countries-but linearity still
prevails! STI Policy Rev 5(2):1–19. https://​doi.​org/​10.​22675/​STIPR.​2014.5.​2.​001
Edquist C, Zabala-Iturriagagoitia JM, Barbero J, Zofío JL (2018) On the meaning of innovation perfor-
mance: Is the synthetic indicator of the Innovation Union Scoreboard flawed? Res Eval 27(3):196–
211. https://​doi.​org/​10.​1093/​resev​al/​rvy011
Etzkowitz H, Leydesdorff L (2000) The dynamics of innovation: from National Systems and “Mode 2” to
a Triple Helix of university–industry–government relations. Res Policy 29(2):109–123. https://​doi.​
org/​10.​1016/​S0048-​7333(99)​00055-4
Etzkowitz H, Webster A, Gebhardt C, Terra BR (2000) The future of the university and the university of
the future: evolution of ivory tower to entrepreneurial paradigm. Res Policy 29(2):313–330. https://​
doi.​org/​10.​1016/​S0048-​7333(99)​00069-4
Fagerberg J (2003) Schumpeter and the revival of evolutionary economics: an appraisal of the literature. J
Evol Econ 13(2):125–159. https://​doi.​org/​10.​1007/​s00191-​003-​0144-1
Fagerberg J, Srholec M (2008) National innovation systems, capabilities and economic development. Res
Policy 37(9):1417–1435. https://​doi.​org/​10.​1016/j.​respol.​2008.​06.​003
Farla K, De Crombrugghe D, Verspagen B (2016) Institutions, foreign direct investment, and domestic
investment: crowding out or crowding in? World Dev 88:1–9. https://​doi.​org/​10.​1016/j.​world​dev.​
2014.​04.​008
Feola R, Parente R, Cucino V (2021) The Entrepreneurial University: How to Develop the Entre-
preneurial Orientation of Academia. J Knowl Econ 12:1787–1808. https://​doi.​org/​10.​1007/​
s13132-​020-​00675-9
Filippetti A, Archibugi D (2011) Innovation in times of crisis: National Systems of Innovation, structure,
and demand. Res Policy 40(2):179–192. https://​doi.​org/​10.​1016/j.​respol.​2010.​09.​001
Filippetti A, Peyrache A (2011) The patterns of technological capabilities of countries: a dual approach
using composite indicators and data envelopment analysis. World Dev 39(7):1108–1121. https://​
doi.​org/​10.​1016/j.​world​dev.​2010.​12.​009
Freeman C (1987) Technology and economic performance: lessons from Japan. Pinter, London
Freeman C (2002) Continental, national and sub-national innovation systems—complementarity and eco-
nomic growth. Res Policy 31(2):191–211. https://​doi.​org/​10.​1016/​S0048-​7333(01)​00136-6
Freeman C, Soete L (1997) The economics of industrial innovation. Pinter, London
Frølich N, Schmidt EK, Rosa MJ (2010) Funding systems for higher education and their impacts on
institutional strategies and academia. Int J Educ Manag 24(1):7–21. https://​doi.​org/​10.​1108/​09513​
54101​10130​15

13
A conceptual framework for modeling heterogeneous actors’… 793

Furman JL, Porter ME, Stern S (2002) The determinants of national innovative capacity. Res Policy
31(6):899–933. https://​doi.​org/​10.​1016/​S0048-​7333(01)​00152-4
Galanakis K (2006) Innovation process make sense using systems thinking. Technovation 26(11):1222–
1232. https://​doi.​org/​10.​1016/j.​techn​ovati​on.​2005.​07.​002
Goñi E, Maloney WF (2017) Why don’t poor countries do R&D? Varying rates of factor returns across the
development process. Eur Econ Rev 94:126–147. https://​doi.​org/​10.​1016/j.​euroe​corev.​2017.​01.​008
Grobbelaar SS, Buys AJ (2005) A conceptual systems dynamics model of research and development
activities in South Africa. South Afr J Ind Eng 16(2):103–121. https://​doi.​org/​10.​7166/​16-2-​169
Grupp H, Schubert T (2010) Review and new evidence on composite innovation indicators for evaluating
national performance. Res Policy 39(1):67–78. https://​doi.​org/​10.​1016/j.​respol.​2009.​10.​002
Guan J, Chen K (2012) Modeling the relative efficiency of national innovation systems. Res Policy
41(1):102–115. https://​doi.​org/​10.​1016/j.​respol.​2011.​07.​001
Guilhon B, Montchaud S (2006) The difficulties of financing innovation in Europe. International Journal
of Entrepreneurship and Innovation Management 6(4–5):383–394. https://​doi.​org/​10.​1504/​IJEIM.​
2006.​010372
Hughes TP (1987) The evolution of large technological systems. In: Bijker W, Hughes T, Pinch T (eds)
The social construction of technological systems: New directions in the sociology and history of
technology. The MIT Press, Cambridge, pp 51–82
Jensen MB, Johnson B, Lorenz E, Lundvall BÅ (2007) Forms of knowledge and modes of innovation.
Res Policy 36(5):680–693. https://​doi.​org/​10.​1016/j.​respol.​2007.​01.​006
Kampmann CE, Oliva R (2009) Analytical methods for structural dominance analysis in system dynam-
ics. In: Meyers R (ed) Encyclopedia of Complexity and Systems Science. Springer, New York, pp
8948–8967
Kergroach S (2019) National innovation policies for technology upgrading through GVCs: A cross-coun-
try comparison. Technol Forecast Soc Change 145:258–272. https://​doi.​org/​10.​1016/j.​techf​ore.​
2018.​04.​033
Kondo M (1999) R&D dynamics of creating patents in the Japanese industry. Res Policy 28(6):587–600.
https://​doi.​org/​10.​1016/​S0048-​7333(98)​00129-2
Kou M, Chen K, Wang S, Shao Y (2016) Measuring efficiencies of multi-period and multi-division sys-
tems associated with DEA: an application to OECD countries’ national innovation systems. Expert
Syst Applic 46:494–510. https://​doi.​org/​10.​1016/j.​eswa.​2015.​10.​032
Krammer SM (2009) Drivers of national innovation in transition: Evidence from a panel of Eastern Euro-
pean countries. Res Policy 38(5):845–860. https://​doi.​org/​10.​1016/j.​respol.​2009.​01.​022
Kravtsova V, Radosevic S (2012) Are systems of innovation in Eastern Europe efficient? Econ Syst
36(1):109–126. https://​doi.​org/​10.​1016/j.​ecosys.​2011.​04.​005
Lane DC (2000) Diagramming conventions in system dynamics. J Oper Res Soc 51(2):241–245. https://​
doi.​org/​10.​1057/​palgr​ave.​jors.​26008​64
Lawson T (1997) Economics and reality. Routledge, London
Lawson T (2015) Essays on the Nature and State of Modern Economics. Routledge, London
Lee TL (2006) An alternative approach to technology policy assessment: dynamic simulation analysis of
Taiwan’s IC industry. Int J Technol Policy Manag 6(2):121–153. https://​doi.​org/​10.​1504/​IJTPM.​
2006.​010907
Lee TL, Von Tunzelmann N (2005) A dynamic analytic approach to national innovation systems: The IC
industry in Taiwan. Res Policy 34(4):425–440. https://​doi.​org/​10.​1016/j.​respol.​2005.​01.​009
Leydesdorff L, Etzkowitz H (1996) Emergence of a Triple Helix of university—industry—government
relations. Sci Public Policy 23(5):279–286. https://​doi.​org/​10.​1093/​spp/​23.5.​279
Liu X, White S (2001) Comparing innovation systems: a framework and application to China’s transi-
tional context. Res Policy 30(7):1091–1114. https://​doi.​org/​10.​1016/​S0048-​7333(00)​00132-3
Lundvall BÅ (1992) National systems of innovation: towards a theory of innovation and interactive learn-
ing. Pinter Publishers, London
Lundvall BÅ (2016) The learning economy and the economics of hope. Anthem Press, London
Mahmood IP, Rufin C (2005) Government’s dilemma: The role of government in imitation and innova-
tion. Acad Manag Rev 30(2):338–360. https://​doi.​org/​10.​5465/​amr.​2005.​16387​891
Mahroum S, Al-Saleh Y (2013) Towards a functional framework for measuring national innovation effi-
cacy. Technovation 33(10–11):320–332. https://​doi.​org/​10.​1016/j.​techn​ovati​on.​2013.​03.​013
Malerba F (1992) Learning by Firms and Incremental Technical Change. Econ J 102:845–859. https://​
doi.​org/​10.​2307/​22345​81

13
794 A. Vetsikas, Y. Stamboulis

Malerba F, Nelson R, Orsenigo L, Winter S (2008) Public policies and changing boundaries of firms in a
“history-friendly” model of the co-evolution of the computer and semiconductor industries. J Econ
Behav Organ 67(2):355–380. https://​doi.​org/​10.​1016/j.​jebo.​2006.​10.​010
Marginson S (2016) The worldwide trend to high participation higher education: Dynamics of
social stratification in inclusive systems. Higher Educ 72(4):413–434. https://​doi.​org/​10.​1007/​
s10734-​016-​0016-x
Maruccia Y, Solazzo G, Del Vecchio P, Passiante G (2020) Evidence from Network Analysis application
to Innovation Systems and Quintuple Helix. Technol Forecast Soc Change 161:120306. https://​doi.​
org/​10.​1016/j.​techf​ore.​2020.​120306
Mazzucato M (2016) From market fixing to market-creating: a new framework for innovation policy. Ind
Innov 23(2):140–156. https://​doi.​org/​10.​1080/​13662​716.​2016.​11461​24
Mlambo VH, Adetiba TC (2019) Brain drain and South Africa’s socioeconomic development: The waves
and its effects. J Public Aff 19(4):1942. https://​doi.​org/​10.​1002/​pa.​1942
Monteiro S, Carayannis E (2017) The quadruple innovation helix nexus: A smart growth model, quanti-
tative empirical validation and operationalization for OECD countries. Palgrave Macmillan, New
York
Mowery DC, Rosenberg N (1993) The US national innovation system. In: Nelson R (ed) National innova-
tion systems: A comparative analysis. Oxford University Press, New York, pp 29–75
Muñoz FF, Encinar MI (2014) Intentionality and the emergence of complexity: an analytical approach. J
Evol Econ 24(2):317–334. https://​doi.​org/​10.​1007/​s00191-​014-​0342-z
Muscio A (2007) The impact of absorptive capacity on SMEs’ collaboration. Econ Innov New Technol
16(8):653–668. https://​doi.​org/​10.​1080/​10438​59060​09839​94
Nasierowski W, Arcelus FJ (2003) On the efficiency of national innovation systems. Socio-Econ Plann
Sci 37(3):215–234. https://​doi.​org/​10.​1016/​S0038-​0121(02)​00046-0
Nelson RR (1993) National Systems of Innovation. Oxford University Press, New York/Oxford
Ngoma AL, Ismail NW (2013) The determinants of brain drain in developing countries. Int J Soc Econ
40(8):744–754. https://​doi.​org/​10.​1108/​IJSE-​05-​2013-​0109
Niosi J (2003) Alliances are not enough explaining rapid growth in biotechnology firms. Res Policy
32(5):737–750. https://​doi.​org/​10.​1016/​S0048-​7333(02)​00083-5
Nooteboom B (1999) Innovation and inter-firm linkages: new implications for policy. Res Policy
28(8):793–805. https://​doi.​org/​10.​1016/​S0048-​7333(99)​00022-0
Olivares-Aguila J, ElMaraghy W (2021) System dynamics modelling for supply chain disruptions. Int J
Prod Res 59(6):1757–1775. https://​doi.​org/​10.​1080/​00207​543.​2020.​17251​71
Özak Ö (2018) Distance to the pre-industrial technological frontier and economic development. J Econ
Growth 23(2):175–221. https://​doi.​org/​10.​1007/​s10887-​018-​9154-6
Panagiotakopoulos A (2020) Investigating the factors affecting brain drain in Greece: looking beyond
the obvious. World J Entrep Manag Sustain Dev 16(3):207–218. https://​doi.​org/​10.​1108/​
WJEMSD-​10-​2019-​0076
Paredes-Frigolett H, Pyka A, Leoneti AB (2021) On the performance and strategy of innovation sys-
tems: A multicriteria group decision analysis approach. Technol Soc 67:101632. https://​doi.​org/​10.​
1016/j.​techs​oc.​2021.​101632
Park HW (2014) Transition from the triple helix to N-tuple helices? An interview with Elias G
Carayannis and David FJ Campbell. Scientometrics 99(1):203–207. https://​doi.​org/​10.​1007/​
s11192-​013-​1124-3
Pereira V, Temouri Y (2018) Impact of institutions on emerging European high-growth firms. Manag
Decis 56(1):175–187. https://​doi.​org/​10.​1108/​MD-​03-​2017-​0279
Pereira V, Corradini C, Temouri Y, Mellahi K (2020) Investigating institutional, economic and social
determinants of European regions for firm growth through employment generation. Br J Manag
31(1):162–183. https://​doi.​org/​10.​1111/​1467-​8551.​12383
Perkmann M, Tartari V, McKelvey M et al (2013) Academic engagement and commercialisation: A
review of the literature on university–industry relations. Res Policy 42(2):423–442. https://​doi.​org/​
10.​1016/j.​respol.​2012.​09.​007
Phillips F, Linstone H (2016) Key ideas from a 25-year collaboration at technological forecasting &
social change. Technol Forecast Soc Change 105:158–166. https://​doi.​org/​10.​1016/j.​techf​ore.​2016.​
01.​007
Pinto H, Santos Pereira T (2013) Efficiency of innovation systems in Europe: an institutional approach to
the diversity of national profiles. Eur Plann Stud 21(6):755–779. https://​doi.​org/​10.​1080/​09654​313.​
2012.​665033

13
A conceptual framework for modeling heterogeneous actors’… 795

Pisano GP (1997) The development factory: unlocking the potential of process innovation. Harvard Busi-
ness Press, Boston
Pruyt E (2013) Small system dynamics models for big issues: Triple jump towards real-world complexity.
TU Delft Library, Delft
Rad MF, Seyedesfahani MM, Jalilvand MR (2015) An effective collaboration model between industry
and university based on the theory of self-organization: A system dynamics model. J Sci Technol
Policy Manag 6(1):2–24. https://​doi.​org/​10.​1108/​JSTPM-​08-​2014-​0035
Radicic D (2019) Effectiveness of public procurement of innovation versus supply-side innovation meas-
ures in manufacturing and service sectors. Sci Public Policy 46(5):732–746. https://​doi.​org/​10.​
1093/​scipol/​scz026
Radosevic S (2004) A two-tier or multi-tier Europe? Assessing the innovation capacities of Central
and East European countries in the enlarged EU. JCMS: J Common Market Stud 42(3):641–666.
https://​doi.​org/​10.​1111/j.​0021-​9886.​2004.​00522.x
Radosevic S (2022) Techno-economic transformation in Eastern Europe and the former Soviet Union–
A neo-Schumpeterian perspective. Res Policy 51(1):104397. https://​doi.​org/​10.​1016/j.​respol.​2021.​
104397
Radzicki MJ (2020) System dynamics and its contribution to economics and economic modeling. In:
Dangerfield B (ed) System dynamics encyclopedia of complexity and systems science series.
Springer, New York, pp 401–415
Ranga M, Etzkowitz H (2013) Triple Helix systems: an analytical framework for innovation policy and
practice in the knowledge society. Ind Higher Educ 27(4):237–262. https://​doi.​org/​10.​5367/​ihe.​
2013.​0165
Ricard LM (2015) Coping with system failure: Why connectivity matters to innovation policy. In: Pyka
A, Foster J (eds) The evolution of economic and innovation systems. Springer, Cham, pp 251–276
Ricciardi F, De Bernardi P, Cantino V (2020) System dynamics modeling as a circular process: The smart
commons approach to impact management. Technol Forecast Soc Change 151:119799. https://​doi.​
org/​10.​1016/j.​techf​ore.​2019.​119799
Richardson GP (1995) Loop polarity, loop dominance, and the concept of dominant polarity (1984). Syst
Dyn Rev 11(1):67–88. https://​doi.​org/​10.​1002/​sdr.​42601​10106
Richardson GP (2011) Reflections on the foundations of system dynamics. Syst Dyn Rev 27(3):219–243.
https://​doi.​org/​10.​1002/​sdr.​462
Rodríguez JC, Navarro-Chávez CL (2015) A system dynamics model of science, technology and innova-
tion policy to sustain regional innovation systems in emerging economies. Int J Innov Reg Dev
6(1):7–30. https://​doi.​org/​10.​1504/​IJIRD.​2015.​067649
Safarzyńska K, van den Bergh JC (2010) Evolutionary models in economics: a survey of methods and
building blocks. J Evol Econ 20(3):329–373. https://​doi.​org/​10.​1007/​s00191-​009-​0153-9
Safarzyńska K, van den Bergh JC (2013) An evolutionary model of energy transitions with inter-
active innovation-selection dynamics. J Evol Econ 23(2):271–293. https://​doi.​org/​10.​1007/​
s00191-​012-​0298-9
Samara E, Georgiadis P, Bakouros I (2012) The impact of innovation policies on the performance of
national innovation systems: A system dynamics analysis. Technovation 32(11):624–638. https://​
doi.​org/​10.​1016/j.​techn​ovati​on.​2012.​06.​002
Sánchez-Barrioluengo M (2014) Articulating the ‘three-missions’ in Spanish universities. Res Policy
43(10):1760–1773. https://​doi.​org/​10.​1016/j.​respol.​2014.​06.​001
Santos SP, Belton V, Howick S, Pilkington M (2018) Measuring organisational performance using a mix
of OR methods. Technol Forecast Soc Change 131:18–30. https://​doi.​org/​10.​1016/j.​techf​ore.​2017.​
07.​028
Schibany A, Streicher G (2008) The European innovation scoreboard: Drowning by numbers? Sci Public
Policy 35(10):717–732. https://​doi.​org/​10.​3152/​03023​4208X​398512
Stamboulis YA (2007) Towards a systems approach to innovation systems and policy. Int J Technol Glob
3(1):42–55. https://​doi.​org/​10.​1504/​IJTG.​2007.​012359
Sterman JD (2000) Business Dynamics: Systems Thinking and Modeling for a Complex World. Mc
Graw-Hill Education, Boston
Stirling A (2007) A general framework for analysing diversity in science, technology and society. J R Soc
Interface 4(15):707–719. https://​doi.​org/​10.​1098/​rsif.​2007.​0213
Stirling A (2010) Multicriteria diversity analysis: A novel heuristic framework for appraising energy port-
folios. Energy Policy 38(4):1622–1634. https://​doi.​org/​10.​1016/j.​enpol.​2009.​02.​023

13
796 A. Vetsikas, Y. Stamboulis

Tajoli L, Felice G (2018) Global value chains participation and knowledge spillovers in developed and
developing countries: An empirical investigation. Eur J Dev Res 30(3):505–532. https://​doi.​org/​10.​
1057/​s41287-​017-​0127-y
Teubal M, Foray D, Justman M, Zuscovitch E (1996) Technological Infrastructure Policy: An Interna-
tional Perspective. Kluwer Academic Publishers, Dordrecht
Tritah A (2008) The brain drain between knowledge-based economies: the European human capital out-
flow to the US. Écon Int 3:65–107. https://​doi.​org/​10.​3917/​ecoi.​115.​0065
Tylecote A (2007) The role of finance and corporate governance in national systems of innovation. Organ
Stud 28(10):1461–1481. https://​doi.​org/​10.​1177/​01708​40607​075676
Uriona M, Grobbelaar SS (2019) Innovation system policy analysis through system dynamics modelling:
A systematic review. Sci Public Policy 46(1):28–44. https://​doi.​org/​10.​1093/​scipol/​scy034
Vetsikas A (2023) Assessing the Performance of National Innovation Systems with a Helix-Based Com-
posite Indicator: Evidence from 24 European Countries. Eastern Eur Econ 1-32. https://​doi.​org/​10.​
1080/​00128​775.​2023.​21677​22
Von Hippel E (1986) Lead users: a source of novel product concepts. Manag Sci 32(7):791–805. https://​
doi.​org/​10.​1287/​mnsc.​32.7.​791
Wonglimpiyarat J (2011) The dynamics of financial innovation system. J High Technol Manag Res
22(1):36–46. https://​doi.​org/​10.​1016/j.​hitech.​2011.​03.​003
Woolthuis RK, Lankhuizen M, Gilsing V (2005) A system failure framework for innovation policy
design. Technovation 25(6):609–619. https://​doi.​org/​10.​1016/j.​techn​ovati​on.​2003.​11.​002
Ye C, Ye Q, Shi X, Sun Y (2020) Technology gap, global value chain and carbon intensity: evidence from
global manufacturing industries. Energy Policy 137:111094. https://​doi.​org/​10.​1016/j.​enpol.​2019.​
111094

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps
and institutional affiliations.

13