Model conceptualisation phase assists a modeller in structuring the systems problem and assigning the model boundaries. This stage is particularly important for our research given the multi-actors nature of the system under study and lack of data involved. In order to achieve a higher level of innovation activity within the construction sector it is fundamental to identify how the cause-and-effect relationships among the variables of the innovation system can be combined into a complex model. The use of the conceptual model allows for a comprehensive assessment of the impact of various policies and deeper investigation on system’s behaviour under various scenarios.
4.3.1. Structural Analysis
Structural analysis using the MICMAC technique was performed to identify key factors and their influences on each other. The analysis enables a modeller to underline the variables that are essential to the system’s evolution. A unique feature of the IPSM approach presented herein is that is uses the empirically conducted structural analysis procedure to aid CLD formulation.
Once the list of the most relevant variables within the construction innovation system was confirmed, a cross-impact analysis was performed in order to identify the role of the variables. By doing so, an opinion survey with the stakeholders was conducted. The experts quantified potential relationships between the system’s elements based on four rates of direct influence of the variables on each other: no influence (0); weak influence (1); medium influence (2); and strong influence (3).
Figure 3 shows an aggregated structural analysis matrix linking the constitutive variables [
28]. The next step aimed to indicate the influence and dependency level of each variable by calculating the sums of each row and column. The MICMAC software [
24] was used to cumulate the rows and the columns for each element. As a result, the variables were ranked based on their dependency and influence level (
Figure 4).
As previously mentioned, taking into account the complexity and inherent dynamics of the problem under study, various stakeholder views should be systematically interpreted in light of several essential pathways to a rational decision-making process. Consequently, the variables were ranked for three groups of the system’s actors separately, according to their influence on other variables and the system as a whole (
Figure 5 and
Figure 6). The ‘Industry’ category was represented by construction firms’ employees and design engineers (
Table SM-1.1). The ‘Government’ participants were experts in the construction and innovation policies area. The ‘Academia’ participants were researchers and academics working in engineering and construction management fields.
According to the conducted structural analysis outcomes, there is a high level of agreement among the three groups of interviewed experts on identification of the role of various elements in relation to the construction innovation system. For instance, there is a prevailing focus on public strategies and collaboration metrics such as level of R&D activity, R&D collaboration and applied research. Nevertheless, some of the factors seem to be more essential for the industry and universities than for the government and vice versa. Thus, construction firm representatives and researchers tend to believe that the public sector plays a primary influential role in the innovation diffusion process as the level of administrative barriers, government incentives and regulation are seen to be impacting the level of innovation performance in the sector the most. On the other hand, public servants prefer the industry itself to take more initiatives in boosting the innovative activity, increasing private R&D expenditure and undertaking more research activities along with the research institutions. As mentioned above, construction firms and academia are the actors that generate innovation while government actions are the most significant drivers to innovation in Russia. Consequently, it is fundamental to understand the different and common interests of all actors within the construction innovation system in order to improve the industry’s innovative capabilities and innovation performance. In terms of the dependent variables, all the groups agreed that the innovation implementation outcomes (e.g., quality of construction projects, client satisfaction and profit maximisation) are highly dependent.
One of the main goals of the structural analysis using impact matrix cross-reference multiplication applied to a classification (MICMAC) technique is to understand individual roles of the elements within the system under investigation which in turn assists a modeller with recognising the key variables associated with strategies and policies, formulated and enforced by decision makers. Those roles could be identified from direct or indirect influence/dependence maps that are generated by MICMAC software (
Figure 7 and
Figure 8).
Every variable could be classified into the following categories depending on the quadrant they belong to:
Influential variables act as input variables that exert strong influence on other elements and the system as a whole when they change. On the other hand, those factors are not dependent on the others. This group of variables must have a priority for decision makers when considering strategic actions and policy design under different scenarios.
Dependent variables represent output variables that have low influence but are the most impacted by other variables and the system.
Relay variables are both highly influential and dependent. These variables describe the system and condition of its dynamics as they are the most unstable and could change to be input or output variables.
Autonomous variables are neither influential nor dependent and have low potential to affect the system. In other words, these variables exist within the system but are not controlled by the dynamics of the model.
As can be seen, the key influential and relay variables again confirm the importance of the public sector’s role within the construction industry in Russia (
Figure 7).
The comparison between the maps for direct and indirect influences reveal hidden key variables or influences. Some of the factors have higher potential to change the dynamics of the system. For instance, the profit maximisation variable became more influential, proving to be not only an important result of successful innovation implementation but also a significant motivating factor while considering investing in innovative solutions. The level of tax incentives variable became less influential while the government incentives variable shifted to become more influential. Hence, government support in the form of additional grants, funds and subsidies is a more significant driver that encourages construction firms to innovate.
4.3.2. CLD Development
After the role of each variable was identified, it was necessary to look at the system as a whole in order to understand how the variables are interrelated. Therefore, as the next step, influence graphs were created using the MICMAC software [
24] to highlight the networks of elements that influence one another. Interconnections among the variables were indicated by arrows that represent different levels of impact of the variables on each other from weak to strong. Within the context of systems modelling, these arrows illustrate the dynamic behaviour of the system while the influence diagrams are associated with a cause-and-effect diagram (
Figure 9). It is clear that MICMAC influence graphs are not user-friendly, particularly for the stakeholder engagement purposes. However, as mentioned previously, the graphs work as an initial reference for the logical building of a CLD. The process of a qualitative model construction is always subjective. Nevertheless, the transformation of the generated influence diagrams into the systems conceptual model in a form of a CLD is based on a comprehensive analysis of interconnections among dynamic variables. Furthermore, a solid theoretical foundation along with previously conducted exploratory study followed by expert participation sessions corroborate the modelling process. The visualization of conversion of the influence graph into a CLD is illustrated in
Figure 9.
As can be seen in
Figure 9, only the strongest connections among the key variables were observed. The analysis investigates how the elements affect each other and how their actions can be transmitted throughout the system. It also should be taken into account that relations between some of the variables can occur through other variables. Additionally, a modeller applies the accurate knowledge when identifying positive and negative causal relationships by answering the question: ‘What are the impacts of variable
i on variable
j at the present?’ The polarity is ‘+’ when two elements change in the same direction (i.e., increase or decrease together). The polarity is ‘−’ when one variable increases while the other decreases and vice versa.
Seven main feedback loops emerged from the constructed CLD representing involvement of the industry, government and academia in the innovation process within the construction innovation system (
Table 3). The relationships among the variables are dominated by reinforcing loops. The same reinforcing loop can have positive or negative impact on the system, depending on how the loop is triggered. In other words, reinforcing processes can be helpful for improving the innovation performance in the construction industry, or can serve to hinder the industry development. The positive description of the identified feedback loops is provided below.
R1 industry motivation: The innovation implementation leads to an increase in construction companies R&D activity as a result of quality improvement and client satisfaction as ones of the most essential industry motivation points. Subsequently, stronger involvement of construction organisations in research activities enhances further applied research.
R2 government’s role: By playing various roles in the construction innovation process a government has the power to make the industry and universities collaborate in order to provide the basis for absorbing and implementing R&D results. In other words, the strengthening of government intervention in the innovation process through various incentive mechanisms leads to greater partnerships and collaborations between the actors of the construction innovation system by supporting strategic innovative projects with greater industry participation. Subsequently, these connections build a robust foundation to the further development of public universities and research centres following by greater need for government involvement.
R3a, R3b practical application: Given the nature of construction innovation, it is necessary to transfer laboratory ideas and research results to the practical environment, which in turn leads to new opportunities for research centres to develop, test and evaluate new technological solutions.
R4 reduction of regulatory burden: The Russian government tends to take measures aimed at promoting the production of domestic innovative materials and technologies as a response to inflicted Western sanctions. The process affects changes in construction-related legislation, rules and building codes, that leads to simplifying administrative procedures. Hence, simplification of administrative procedures increases the chances of strengthening the contracts between universities, research centres and construction companies, which in turn enhances import substitution process.
R5 need for innovation: As a main client, the government is able to significantly influence and motivate construction companies to innovate through increasing the demand for cutting-edge products and processes in order to promote domestic science and further industry development.
There is a balancing loop B1 that represents side effects and consequences that can hinder the innovation process. Consequently, in order to limit the negative impacts of increasing costs and decreasing R&D activity, additional promotions (loops B2 and B3) are needed in order to boost the economic interest of firms without forcing them to wait for short-term economic benefits.
B1 expectation of short-term profit: Following a client’s requirements for implementing innovative solutions, a construction company, however, experiences significant business expenses. As a result of the increase in the quality of the final product, construction costs can be high and lead to low profits in the short-term prospective. Hence, construction companies prefer to stay conservative and do not invest in R&D. This leads to a drop in the industry’s interest in being innovative.
B2 support for innovation: High costs may be significant due to the implemented innovative solutions and make it hard for construction firms to compete. As a result, insufficient industry’s innovative activity takes place. This situation pushes the government to intervene by applying appropriate fiscal measures and incentive mechanisms in order to increase the attractiveness of research investments and stimulate not only those who implement innovations but also those who discover and develop them.
B3 overcoming isolation: In the majority of cases the insufficient technical and technological capabilities of construction companies hinder the industry’s ability, not only to implement innovative solutions but also to quickly adapt to new opportunities. In order to cope with the unwillingness of the industry to innovate, the government implements policies that promote science and invests in higher education and techno-parks to entice public R&D activity first. This in turn establishes integrated R&D collaborations required for effective implementation of technology-using strategies and research commercialisation and boost industry participation in the process [
6].
As the next step, stakeholder workshops were conducted to refine and extend the initial conceptual model created on the base of the structural analysis with MICMAC. As a result, a representation of the problematic situation of innovation diffusion within the construction innovation system was generated (
Figure 10). A comprehensive description of the step-by-step CLD extension and the feedback loops within the model is given in
SM-3.
4.3.3. System Archetypes
In order to more effectively understand the root causes of the challenges and complex management issues, visualise the high leverage interventions and predict the system behaviour, four system archetypes were identified in the CLD representing the construction innovation system in Russia. The archetypes include limits to growth, shifting the burden, tragedy of the commons and eroding goals. As an example, two archetypes are explained below: limits to growth and shifting the burden.
The
limits to growth archetype represents a process where accelerating growth is limited by a constraint that restricts growth or success. In other words, a period of growth initially starts due to the reinforcing loop, then is followed by a period of deceleration as the balancing loop inflicts limits and, eventually, pushes back on the reinforcing loop. As a result, a virtuous or a viscous cycle occurs due to the continuing efforts that lead to diminishing returns as limits are approached [
30]. A number of large and medium-sized construction firms in the Russian Federation were identified to have this archetype since the growth in the number of innovative companies becomes constrained as feedback dominance shifts from a reinforcing to a balancing loop (
Figure 11).
As detailed in
Supplementary Materials SM-3, imitative construction companies and firms involved in R&D accelerate the industry development. This reinforcing loop is the engine of the level of innovation in the construction sector growth. However, the process has a balancing loop which limits this growth, that is, approaching market saturation that gradually decreases the number of potential innovative companies (
Figure 11). Mainly large and medium-sized construction firms tend to have a potential capability to implement innovative solutions in the Russian context. Consequently, strategies to increase the market size become one alternative issue to improve the innovation performance. Moreover, the innovation rate is affected by companies’ readiness to adopt innovations from competitors or develop their own. Within the government program in which prospective companies have effectively received financial support, they have a tendency to invest in innovative solutions in order to become competitive and improve the business performance indicators.
The
shifting the burden archetype describes a situation where relatively simple quick fixes lead to apparent success and, as a result, become addictive in tackling the urgent problems before dealing with ambiguous and complicated situations. On the other hand, more fundamental solutions are not attractive as they take much longer to apply and require a large commitment of resources. However, these quick fixes have only temporary benefits and cause serious side effects that escalate the real problem [
30]. Consequently, the initial problem reoccurs over time with greater intensity that leads to significant delays in implementing long-term fundamental solutions to the problem. An ‘Innovators vs. imitators’ problematic situation was identified within the CLD to have this archetype as ‘firefighting’ behaviour which usually prevails among potential innovative construction organisations. There is a tendency to adopt innovative solutions from other successful companies (i.e., imitate others) rather than boosting innovation by collaborating with researchers and scientists (
Figure 12).
As discussed previously, when there is a need to implement innovations in the construction process, companies have two options to become innovative: to be involved of the R&D activity or adopt already known technological know-how. Both options have a significant impact on the industry development due to active innovation diffusion process. However, the imitative strategy is much less cost and labour intensive, and, therefore, tends to become a priority among companies willing to introduce novel solutions and technologies. Unfortunately, it is a quick fix gradually leading to diminution in the already relatively weak interest of firms in R&D progress (
Figure 12). In the long term, investing in local research and science would create know-how, new employment and industrial development. Therefore, the strategy to overcome low innovation performance problems should focus on strengthening the university-industry links and allocate resources to support domestic R&D.