2.1. Prior Research

2.1.1. Characteristics of the Fourth Industrial Revolution

Technological innovation triggered by the Fourth Industrial Revolution does not go directly from technology to the market, but is moving toward changing politics, gradually leading the economy, affecting society, and entering the market. In other words, it can be seen as a continuous process of connection of innovations that will spread in an S-shape while forming a time-series causal relationship between key elements, and change the frame of the current society (Kotler et al., 2016; see Fig. 1).

 

Fig. 1

Comparison of the technological flow of the existing technological innovation (dotted line) and the Fourth Industrial Revolution (S-shape). Source: Revised from Kotler et al. Sigma Books (2016).

 

For example, in the case of artificial intelligence (AI) and block chain technology, which are representative technologies of the Fourth Industrial Revolution, the markets have not yet been formed, so they first pass the legal regulatory framework (negative or positive), and then the national economic framework and social culture would go through the process of spreading to the market through proper adaptation. In this process, technological innovation takes place through the process of openness, cooperation, and creative convergence of ideas.

2.1.3. Multi-level Perspectives

A sustainability transition occurs when a particular socio-technical transition is directed towards sustainability. The sustainability transition is defined by Markard et al. (2012). It is described as a long-term, multi-dimensional, and fundamental transformation process in which traditional socio-technological systems are transformed into more sustainable alternatives (Asquith et al., 2017). The general characteristics of a sustainability transition include four elements.

First, it entails profound changes along multiple dimensions: technological, organizational, political, economic, behavioral, and sociocultural (Markard et al., 2012; Rotmans et al., 2001). Second, it requires interaction between multiple actors such as industry, government, users, and social groups. Third, it is a long-term process that takes several decades to unfold (Geels, 2012; Rotmans et al., 2001). Fourth, it requires the development and dissemination of a wide range of innovations, including new technologies, policies, standards, and social practices (Geels et al., 2008).

In general, the most well-known concept and theoretical idea for exploring phenomena in the field of socio-technological change is the Multi-level Perspective (MLP) (Geels, 2010). MLP provides a holistic view of the multidimensional complexity of transformation in socio-technological systems (see Fig. 2).

 

Fig. 2

Conceptual diagram of multi-level perspective. Source: Revised from Asquith et al. European Environment Agency (2017, p. 24).

 

Understanding transitions in MLP occurs as a result of coevolutionary interactions between the three levels of analysis (Geels, 2010): (1) the niche level at which innovation occurs and builds momentum; (2) the socio-technological system level at which the established structures and networks of actors, institutions, and economic practices stabilize over time; and (3) the landscape level, which refers to a broader context in which major influencers such as the Fourth Industrial Revolution or global discourses such as digital transformation are happening.

The most important level in the MLP framework is the niche level. Geels (2010) describe a niche as a ‘protected space’ in which radical innovation, pioneer projects, and learning processes occur. Accommodating development can lead to new and stable socio-technological formations at the regime level. The framework concept is understood as an analytic analysis concept. While the analysis concept of social technological systems refers to tangible and measurable factors (e.g., market shares, regulations, consumption patterns), regimes are understood as more intangible analytical entities. A socio-technical framework refers to the rules and routines that are inherent in actors and lead to specific actions.

Many scholars agree that technological innovation is influenced by interactions within innovation systems (Revilla & Kiese, 2009). Technological change and interactions with institutional and organizational structures have been explored for systems at various spatial scales. In addition to the national innovation system (Nelson, 1992), representative examples include the concept of a regional innovation system (Asheim & Isaksen, 2002; Cooke, 2001).

MLP defines transition as a niche (a trajectory for radical innovation), a social-technical framework (a trajectory of established practices and related rules that stabilize existing systems), and an extrinsic socio-technical environment. Rip and Kemp (1998) and Geels (2002, 2005) consider it a non-linear process due to the interaction of development at the three levels of analysis: (1) niche innovation builds internal momentum; (2) changes in landscape level put pressure on regimes; and (3) regime destabilization creates windows of opportunity for niche innovation. The evolving interaction can be further subdivided into several stages, such as emergence, take-off, acceleration, and stabilization (Rotmans et al., 2001). Each of these steps can be associated with a specific mechanism (Geels, 2005).

A niche market here means R&D labs, subsidized pilot projects, or users with special needs and willingness to support new innovations. As a ‘protected space’ like a niche market, R&D PIE is assumed in this study. Niche markets are important for transition because they provide the seeds for systemic change. The literature on niche innovation (Kemp et al., 1998; Schot & Geels, 2008) distinguishes three key processes of niche development. The first is the alignment of expectations or visions that aim to provide guidance for innovation activities and to attract attention and funding from external actors. Second, the establishment of social networks and the registration of more agents expands the resource base of niche innovation. Third, it is a learning and coordination process at different levels, such as technology design, market demand and user preferences, infrastructure requirements, organizational issues and business models, policy tools, and symbolic meanings (Geels, 2011).

2.2. Research Frame Composition

If the above-mentioned contents are applied to R&D PIE, the accuracy of R&D investment analysis will be improved and it will be widely accepted by many ministries. The alignment of the various R&D PIE strategies and coordination processes results in a stable composition (dominant design) at the national level. And as networks grow (particularly strong actors’ participation can deliver legitimacy and resources to niche innovations), niche markets gain momentum.

The Multi-phase Concept (MPC) and MLP each consist of four distinct phases: the predevelopment phase, take-off phase, breakthrough phase, and stabilization phase, and three interacting transition levels (niche, regime, and landscape) which aim to theorize the complex transition dynamics (Chang et al., 2017).

Strategic Niche Management (SLM), which aims to identify the characteristics of successful niche markets, and Transition Management (TM), as an innovative method to solve complex social problems, have been used as research tools, but mainly to proactively influence the sustainability transition and are used as a policy tool for management.

SNM proposes three parts (articulation of expectations, the building of social networks, and a multidimensional learning process) and a niche process (Kemp et al., 1998), and TM proposes four steps to fostering new networks for sustainability (Kemp et al., 1998).

Ultimately, in order to successfully apply MLP to society, steps such as setting a transition area, developing a transition agenda, mobilizing a resulting transition network, and coordinating a vision, agenda, and coalition are necessary (Loorbach, 2010; Rotmans & Loorbach, 2009). MLP had great difficulty operating the levels in empirical studies due to poor conceptualization of the three proposed levels (i.e., niche, system, and landscape). In particular, policy makers and institutions are generally seen as part of a system that needs to change, rather than as separate actors with the power to steer society in a long-term direction. TM creates a social movement on sustainability through new alliances and networks (Loorbach & Rotmans, 2010).

In such a rapidly changing industrial environment, government R&D investment is needed for rapid and sustainable innovative growth that reflects new perspectives and methods to link and spread the accumulated strengths of the main industries, supplement the necessary capabilities, and solve the social and industrial challenges at hand. It is a reality that requires a total innovation of the system.

In addition, soft power for strengthening social, cultural, and institutional capabilities such as creativity orientation, border crossing, and deregulation should be concurrently improved.

In order to realize this, we break away from the existing R&D budget allocation and adjustment method for each project, and integrate management and evaluation of R&D projects scattered by departments by field. An investment analysis platform that is configured and supported is necessary (see Fig. 3).

Fig. 3

Conceptual diagram of research frame composition. Source: Revised from Chang et al. Sustainable Development (2017, p. 362).

3.1. R&D PIE

In Korea, at the Economic Ministers’ Meeting held in February 2018, “Government R&D Investment Innovation Measures to Support Innovative Growth-Introduction of Package-Type R&D Investment Platform System” was adopted as the agenda. As a Korean model that can differentiate itself from the Fourth Industrial Revolution promotion policies of advanced R&D countries, “technology-industry-manpower-policy linkage (Cross-cutting Innovation),” and its supporting system of package-type R&D investment platform (R&D PIE) construction, was adopted (see Fig. 4).

Fig. 4

Matching result of R&D PIE investment technology group (example of autonomous vehicles). R&D PIE, R&D Platform for Investment and Evaluation. Source: based on OPSI (Observatory of Public Sector Innovation). https://oecd-opsi.org/innovations/rd-platform-for-investment-and-evaluation-rd-pie.

This is a cross-cutting model that is differentiated from the German model centered on manufacturing and factory (factory creates value), the US model centered on data and high-tech industries (data creates value), and the Japanese model centered on robots and human knowledge (human knowledge creates value). The key content is the evolution of the ecosystem of major industries (cross-cutting technologies create value) through the platform’s concept of ‘technology-industry-manpower-policy linkage’ (European Commission, 2014a; see Fig. 5).

Fig. 5

Example of EU cross-cutting roadmap conceptual diagram. Source: European Commission (2014a, p. 31).

The National R&D PIE, which responds to the Fourth Industrial Revolution based on big data, selected GPTs for each field through big data analysis such as the global thesis DB (50 million cases), patent DB, and corporate DB (see Table 1).

The current R&D projects and the technology classification system are mapped. It is used for discovering various fields of investment necessary for setting investment direction, business planning, and performance management.

In Korea, dynamic collaborations between relevant ministries are required to realize more successful innovative growth, and it is necessary to establish a science and technology integrated data-based innovation control platform that can quickly respond to policy adjustments. In order to discuss the step-by-step implementation of R&D innovation measures and sustainable outcomes, an approach from the perspective of the national innovation system, not led by a single ministry, is required. In order to advance the innovation model, cross-ministerial linkages and cooperation are required.

R&D PIE is a platform that provides comprehensive support for ‘technology-human resource training-system-policy’ in the form of a package, breaking away from the budget allocation method for each project. We perform R&D investment decision-making support.

This was actually systemized by benchmarking the system according to the EU’s Horizon 2020 cross-cutting roadmap.

The EU identified key supporting technologies (KETs) to promote technological innovation by utilizing the cross-cutting roadmap methodology proposed in 2012. The Horizon 2020 roadmap presented the investment priorities of European Structural and Investment Funds (ESIF) and the European Investment Bank (EIB) based on KET (see Fig. 6). With a budget dedicated to KET of nearly €6 billion, Horizon 2020 was created to provide KET with high importance and visibility to promote industrial innovation and balance R&D&I (including pilot lines and demonstrators) for projects closer to the market to promote industry, through which R&D and innovation support was readjusted. Addressing clear industry and market needs across a wide range of industries, KET is recognized as helping to identify promising areas of innovation. Also, because of its systematic relevance to Europe’s ability to innovate, modernize its industrial base, and solve social challenges, the EU considers that full use of KET will ensure a social return on investment and job creation in the EU (European Commission, 2014b).

Fig. 6

Examples of the EU’s KET roadmap (Electronics and Communication Systems). KET, key supporting technology. Source: European Commission (2014b, p. 6).

3.2. Framework of R&D PIE

In particular, a framework for enhancing R&D total factor productivity was established. In general, the rate of change of total factor productivity can be decomposed into four factors: technology, distribution, scale, and operational efficiency (Kumbhakar et al., 2000):

Existing R&D policies pursued maximization of quantitative aspects (scale efficiency) rather than qualitative aspects based on input expansion. However, in order to improve the overall efficiency of government R&D investment, improvements such as sector targeting and improvement of the operating system (technology), role differentiation and concentration (distribution) between actors, and improvement of system and execution efficiency (operation) are necessary (see Fig. 7).

Fig. 7

Conceptual diagram for improving investment efficiency of R&D PIE. R&D PIE, R&D Platform for Investment and Evaluation.

In order to improve the quality of R&D efficiency, (1) focusing R&D investment in target fields (technical efficiency), (2) re-establishing roles for each entity (distribution efficiency), and (3) improving systems and maximizing execution efficiency (operational efficiency), are required.

The R&D PIE model affects the economy for the following reasons:

3.3. Application Process of R&D PIE

The R&D PIE searches for fields requiring investment, which consists of searching for patents and theses by technology group, linking technology and industry, extracting indicators, and deriving fields requiring investment. We collected domestic and overseas publication trend information of patent and thesis information, and extracted indicators (employment inducement coefficient, value-added inducement coefficient, forward industry linkage effect, backward industry linkage effect) according to technology and industry linkage logic. In addition, the investment field was derived by considering the R&D investment amount information for the past five years (see Fig. 8).

Fig. 8

Application process of R&D PIE. R&D PIE, R&D Platform for Investment and Evaluation; GPT, General Purpose Technology; IPC, International Patent Classification.

According to the results of establishing a package-type R&D investment strategy conducted by Korea’s Science and Technology Innovation Headquarters in 2021, the employment inducement coefficient in the industry-related analysis of the 16 R&D PIE fields was 8.0 to 11.3 per billion won. The value-added inducement coefficient was analyzed to be between 0.965 and 0.985. The induced effect was higher than the manufacturing average (employment inducement coefficient 2.285, value added inducement coefficient 0.581), and it was analyzed that this was due to an economic spill-over effect. As well, in the front-to-back linkage effect analysis that reflects the inter-industry linkage characteristics of the technology, the average back-industry linkage effect of the manufacturing industry was 1.125, while the average of the R&D PIE technology group was 1.134, which was slightly higher. The manufacturing average was 1.161, which is slightly higher than the average of the R&D PIE technology group, 1.01. In conclusion, the R&D PIE technology group showed strong characteristics of the downstream industry, which is interpreted as high in the nature of basic technology (see Fig. 9).

Fig. 9

Economic index results by investment field of R&D PIE. R&D PIE, R&D Platform for Investment and Evaluation.

In case of budget allocation and adjustment in 2021, 47 new projects were to be discovered and KRW 2.134.2 trillion invested in ten major fields (see Table 2).

4.1. Government R&D Investment and Innovative Growth Causal Map

In general, innovation refers to education and R&D. In most cases, however, innovation is the result of various learning processes embedded in everyday economic activity. Therefore, interactions that increase the efficiency of producers and interactions between users and producers are key to innovation. While traditional innovation research is concerned with the resources put into the R&D system, the innovation system takes a holistic approach. The innovation system includes institutional, organizational, social, and policy factors as well as economic factors that influence innovation. In this aspect, the approach of the innovation system must first identify the elements and analyze the relationship between the elements, and this analysis method constitutes the causal map in the system dynamics. In other words, the causal map is a schema for synthesizing causal relationships between various variables by focusing on the feedback structure, and has been used as an analysis tool in the stage before system dynamics modeling.

First, the basic development for such feedback composition in the national innovation system is as follows. First, R&D activities including national R&D investment will affect knowledge accumulation through national R&D projects, and this can lead to technological innovation or gradual productivity improvement through knowledge accumulation. Second, in general, knowledge accumulation brings technological progress and has a direct effect on growth along with labor and capital; and since knowledge accumulation and technological progress increase the return on investment of human and physical capital, it can have an indirect effect on growth by increasing the efficiency of other production factors such as labor and capital. Third, growth expansion through knowledge accumulation acts as an incentive to expand R&D investment, while allowing companies and information to secure R&D investment funds equivalent to a certain portion of GDP, thereby acting as a driving force to achieve another technological innovation. In other words, the high economic growth achieved by R&D investment leads to an increase in R&D investment in the private and public sectors again through an increase in corporate profits and an increase in the government budget. Lundvall (1992) called this phenomenon a ‘cumulative causation’ between technology and growth. He argued that R&D and technological innovation enhance a country’s technological capability and bring economic growth through capital accumulation, which in turn becomes an investment resource for advanced technology and an incentive (Pianta, 1995).

On the other hand, technological progress can be further promoted through institutional factors, particularly the feedback process of regulatory reform. Technologically advanced companies will erode the market share of existing monopolies, which will lead to relaxation of monopoly regulations. In addition, deregulation will stimulate technological innovation and bring about a reduction in product prices and diffusion of core technologies through rapid improvement in productivity.

The current economic crisis in Korea is a link that leads to ‘innovative growth & Fourth Industrial Revolution enhancement → scientific and technological innovation → innovative ideas → activation of technology convergence (innovative knowledge accumulation) → technological innovation’ as shown in Fig. 10. This is due to the fact that the government is implementing the lagging indicators such as deregulation and job creation in a ‘weak or missing state.’ In the new national science and technology innovation system (application of R&D PIE) that responds to digital transformation, ‘R&D investment → knowledge accumulation → innovative growth and enhancement of the Fourth Industrial Revolution → technological innovation and convergence activation → improvement of total factor productivity.’ The virtuous cycle mechanism that leads to job creation → economic growth → new R&D investment is working. Here, innovation and job creation based on the Fourth Industrial Revolution play the role of an intermediate chain leading to the final result, the creation of national wealth. The framework that normally connects this missing link is the concept of “Cross-cutting Innovation” applying R&D PIE.

Fig. 10

Causal map of the national science and technology innovation system. R&D PIE, R&D Platform for Investment and Evaluation.

4.2. Analysis of Simulation Results

In the simulation model, ‘National R&D Investment,’ ‘Ecosystem Value Chain Linkage Level,’ ‘4th Industrial Revolution Level,’ ‘Technology Convergence, ‘Technology Innovation,’ ‘Job Creation,’ ‘National Welfare,’ ‘National Competitiveness,’ etc. of the variables appear as stock variables, and the remaining variables were regarded as flow or auxiliary variables (see Appendix).

In particular, the four index values, ‘Index of Package,’ ‘New Idea Effect,’ ‘Employment Creation Effect,’ and ‘Promoting Commercialization,’ which can be seen as effects by applying R&D PIE, were set to the highest (0.9) and lowest (0.1) values. Sensitivity analysis of the R&D conversion effect according to successful application of R&D PIE was conducted (see Fig. 11).

Fig. 11

Simulation model of the national science and technology innovation system.

It is the result of computer simulation when all the initial and exponential values of the stock level in the model are 0.1 and 0.9. In the simulation results, the trend of national competitiveness and job creation, which are low-volume (level variables), is expressed over time. This is because national competitiveness and job creation indicators are representative indices that determine the remaining stock level. The values of each storage variable in Fig. 11 do not mean values in the real world. In the world of equalized units, it refers to the value calculated by simulating on the computer when each variable is converted to the basic relation equal unit modeling method with a qualitative scale ranging from 0 to 1 from a general point of view. In the case of the simulation of the effect of applying R&D PIE, assuming an index value of 0.9 as a result of the simulation, the job creation effect and driving national competitiveness exceeded 0.8 around 28 months, and national competitiveness exceeded 0.8 after 56 months. On the other hand, in the case where the R&D PIE, which assumed an index value of 0.1, was not applied in the simulation, job creation drives national competitiveness during the first 62 months, and after that, job creation slows down and national competitiveness exceeds job creation. All of them could not rise anymore and appeared to be in a state of equilibrium (see Fig. 12).

Fig. 12

Sensitivity analysis of R&D PIE application. R&D PIE, R&D Platform for Investment and Evaluation.