Learning, knowledge, and the role of government: a qualitative system dynamics analysis of Andalusia’s circular bioeconomy

Learning,


Introduction
A common feature observed in government strategies for the development of a bio-based economy is the belief that a strong innovation system will play a key role in the realization of the sector's potential. In this context, the dynamics of learning and knowledge accumulation is critical and presents a key challenge as the bioeconomy is largely composed of companies with persistently low levels of digitalization (Bacco et al., 2019) and struggling to develop effective business models (Reim et al., 2019). For over 30 years, the innovation systems approach (Lundvall, 1992;Nelson, 1992) has provided an important theoretical framework to explain the complex interactions that take place between the different participants of the innovation process as well as the basis for policymaking in the fields of science, technology, innovation, and economic development. This concept has enjoyed vast popularity for the many advantages that it offers over the traditional linear models developed in the previous four decades. However, the inner dynamics of innovation systems remain somewhat unexplored, largely because innovation studies have often pursued a linear thinking approach while the innovation process is known to follow non-linear paths and involve feedback loops across all the stages.  Reim et al., 2019). In response to these discussions, the term "circular bioeconomy" was created, and some attempts have been made to define the concept but clear guidance for bioeconomy practitioners is still needed (Stegmann et al., 2020). A circular economy aims to maintain the value of products, materials, and resources as much as possible while minimising the generation of waste (European Commission, 2015), thus requiring interactions across several domains and the involvement of multiple players. However, despite this complexity, most of the analyses conducted until very recently have addressed the processes and components of innovation systems for the development of a circular bioeconomy in isolation. This is changing through the application of system dynamics modelling tools.
The system dynamics modelling approach was created during the late 1950s and early 1960s to understand the non-linear behaviour of complex systems and build models that capture their dynamic nature over time (Forrester, 1961;Meadows, 2009;Sterman, 2000). It provides powerful tools to examine cause-and-effect relationships, feedback mechanisms, non-linear effects, time delays and, accordingly, high complexity. Applications of system dynamics are increasingly found in a wide range of areas, including manufacturing, construction, infrastructure, software development, healthcare, population studies, waste management, water resources management, ecological and economic systems, and environmental management, among many others (Andersson  (Uriona and Grobbelaar, 2019) found that the contribution of the system dynamics approach to research on innovation systems has been limited and that, despite the high value offered by these tools, system dynamics modelling has not yet had the expected scientific impact in this domain.
Similarly, while there is a growing understanding that the application of linear approaches to analyze the complex mechanisms and interactions that occur in the development of the bioeconomy are often insufficient to get a good grasp of the dynamics governing this transition, systems thinking methods have only recently started to be applied in the study of these pathways (Bennich et al., 2018a(Bennich et al., , 2018bBlumberga et al., 2018;Stark et al., 2022). Work in this field is just beginning and, as a result, there is a myriad of areas where important research gaps exist. Thus, for example, despite the broad recognition that the bioeconomy is a knowledge-intensive sector that depends largely on public policies and programs, the systems thinking approach has found virtually no application to date in the literature about learning, knowledge, and the role of government in the transition to a circular bio-based economy (see Methodology, below).
Against this backdrop, our study seeks to increase the understanding of how the dynamics of innovation systems influence the development of the circular bioeconomy, exploring how knowledge and learning influence the performance of these processes, and identifying points where interventions could enhance the strengths and overcome the weaknesses to promote growth in this sector of the economy. We focus our analysis on the Andalusian bioeconomy because it is a key component of the region's economy, generating an annual turnover of about 29 billion euros and employing around 300.000 people (approximately 9% of the total) (Institute of Statistics and Cartography of Andalusia, 2022). The significance of this sector led the Regional Government of Andalusia to release a circular bioeconomy strategy in 2018 and become one of the first regions in Spain to acknowledge the opportunities that it offers for sustainable growth and competitiveness (Regional Government of Andalusia, 2018).
To achieve our objective, we address the following questions: 1) What are the underlying causal structures and feedback mechanisms that interact dynamically in Andalusia's bioeconomy system to shape the transition to a circular biobased economy in the region?
2) What potential learning and knowledge-related points exist in the system where targeted interventions could have significant impact?
3) What priority actions could be implemented at the identified intervention points that would have the highest probability of positive impact? 4) From a systems thinking perspective, what is the role of government in the transition to a circular bio-based economy?

Methodology
We apply qualitative systems modelling methods (Meadows, 2009;Sterman, 2000) to analyze Andalusia's circular bioeconomy and to conduct a qualitative assessment of key learning-and knowledge-related intervention points to develop this sector. System dynamics models are wellsuited for the representation of this type of system as they allow us to analyze complex situations, applying a comprehensive view of the whole and at the same time examining the causal relationships among each of its parts. Furthermore, they provide a valuable tool to build theory around behaviours observed within a system and assess the potential impact that management and policy actions could have on it.
In this study we use causal loop diagrams (CLDs) as they are an easy and powerful tool used in system dynamics modelling to provide a visual representation of the elements of a system, their interdependency relationships, and the feedback processes that exist between them. A CLD comprises a set of variables that are connected by arrows that are assigned either a positive (+) or negative (-) sign, according to how a dependent variable is affected when an independent variable changes. The connected variables, in turn, can form positive and negative feedback loops, which 6 are at the heart of system dynamics. These loops are positive or "reinforcing" (R) when a change in a variable circulates along the loop in a way that it reinforces the initial variation, generating growth or acceleration and having a destabilizing effect. And they are negative or "balancing" (B) when a change in a variable circulates along the loop in a way that counteracts the initial variation, acting as a stabilizing force. A feedback loop is deemed to have a reinforcing effect when all the relationships are positive or if it contains an even number of negative links, and it has a balancing effect if it contains an odd number of negative links. Lastly, the existence of lags in the cause-effect relationships between variables is another key aspect of system dynamics and implies that the effects of a change in a variable become evident not immediately, but after some time. A time delay is indicated in a CLD by a perpendicular double line marked in the arrow where it takes place.
Our methodology comprised four steps, as illustrated in Figure 1 and described below.

Literature Review
The first step consisted of a comprehensive review of the literature related to the application of systems thinking to the study of the development of bio-based sectors and the transition to the bioeconomy, with the objective of identifying the factors that influence performance in these processes and detecting research gaps. For this purpose, we applied an approach based on the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) method ( Figure   2), using different combinations of the keywords "system dynamics", "systems thinking", "bioeconomy", "bio-based economy", "transition", "innovation", and "innovation system" within article titles, keywords, and abstracts in the Scopus database. After several iterations, the searches that yielded a manageable number of relevant results (under 500) were "bio-based economy" AND "transition" (148 results), "system dynamics" AND "innovation system" (116 results), "bioeconomy" AND "transition" AND "factors" (65 results), "systems thinking" AND "innovation system" (60 results), "bioeconomy transition" (39 results), "system dynamics" AND "bioeconomy" (16 results), "systems thinking" AND "bioeconomy" (9 results), "bioeconomy" AND "transition" AND "variables" (8 results), "transition to the bioeconomy" (8 results), "system dynamics" AND "bioeconomy" AND "transition" (3 results), "transition to the bio-based economy" (2 results), "system dynamics" AND "bio-based economy" (2 results), and "systems thinking" AND "innovation system" AND "CLDs" (1 result). The duplicates were then discarded, a qualitative screening of the remaining articles was performed through a review of the abstracts followed by a full text review, and a total of 30 were retained due to their relevance for our study (Allas,  Grobbelaar, 2019). Furthermore, a search of the Scopus database with the keywords "system dynamics" AND "bioeconomy" AND "knowledge" was conducted and yielded merely two results, of which only one was relevant to our study but was focused on the health sector (Oriama and Pyka, 8 2021). This finding revealed an important research gap and led us to the decision of focusing the second part of our study on the identification of learning-and knowledge-related leverage points.
Concurrently, a review of the literature on the Andalusian bioeconomy and innovation system was conducted applying the same approach and the keywords "Andalusia", "bioeconomy", "bio-based economy" and "innovation system", to gain a perspective of the regional context. A total of five documents were retained after expanding the quest to Google Scholar and the Google search engine as the Scopus database did not yield any relevant result (Agency for Innovation and Development

Preparatory Analysis
Several definitions of circular bioeconomy were found in the literature due to the vast variety of sectors and activities that make up the bio-based economy sector (Bugge et al., 2016;Giampietro, 2019). Therefore, for the purpose of this study, we decided to focus on the Andalusian Circular Bioeconomy Strategy 2030 (ACBS) (Regional Government of Andalusia, 2018) and thus adopt the scope therein used to define the circular bioeconomy, i.e., the primary and agro-industrial production of food for human consumption are not included. Food products are considered a resource for the circular bioeconomy only if they are deemed unsuitable for human consumption due to non-compliance with regulations or loss of quality during their processing.
The ACBS document comprises four strategic lines and four cross-cutting lines of programmes  Once the scope was defined, the strategic lines of the ACBS were set as the boundaries of the system for this study and a first group of variables was selected from each one of them, using the 30 articles retained in our literature review as a guiding reference. These were subsequently submitted to further refining based on the Andalusia-related literature and the thematic connections among them was analyzed to further define the boundaries of the system.

Development of CLDs and Refinement of the Models
In the next step, the variables selected during the preparatory analysis were used to design a primary Subsequently, taking the primary CLDs as the starting point, the models were improved iteratively to portray the relationships among the key variables and factors formulated in the Andalusian strategy, providing a visual understanding of the causal relationships and the feedback loops that shape the reinforcing and balancing forces between the various components of the system. Along this process, several secondary or redundant elements were gradually eliminated to obtain a simple but comprehensive representation of the system with the lowest possible number of elements (Sterman, 2000). The resulting CLDs were subsequently merged into a series of integrated CLDs, to provide an overall view of the system.

Identification of Intervention Points and Targeted Actions
Lastly, the strategic lines and programmes of the ACBS were revisited for a full text analysis of its 17 prescribed measures to identify specific actions related to learning and knowledge that are known for their effectiveness in industrial development and that could be implemented to facilitate the transition to a circular bio-based economy. The actions thus identified were subsequently used to find the appropriate intervention points where their implementation could have meaningful impact (Meadows, 2009).
Throughout the entire process, the literature findings were complemented by the authors' combined experience of over three decades in the analysis, design, and implementation of public policies, strategies and programs for science, technology, innovation, and economic development in both government and the academic sector in North America and the European Union. In addition, preliminary versions of the CLDs and their intervention points were presented for feedback and comments at the ISPIM Innovation Conference 2021 (Berlin, Germany, June 20-23, 2021) and 30 experts at the XIII Agrifood Economics Congress of the Spanish Association of Agri-food Economy (Cartagena, Spain, September 1-3, 2021), which allowed the collection of further contributions that enriched the results.

Results
The data gathered from the literature review and analysis of the ACBS led to four conceptual CLDs portraying the causal relationships among 33 key variables identified from the four strategic lines of the ACBS: (1) sustainable generation and availability of biomass resources, (2) infrastructure and logistics management of biomass resources, (3) processing of biomass resources and capacity of industrial production of bioproducts and bioenergy, and (4) development of markets for bioproducts and bioenergy. Subsequently, after merging them into a series of integrated CLDs, a total of 20 key learning-and knowledge-related interventions points were identified, along with 52 targeted actions that could have meaningful impact on the system.

Sustainable Generation and Availability of Biomass Resources
The first strategic line formulated in the ACBS relates to increasing the availability of biomass resources produced sustainably for their subsequent conversion into bioproducts and bioenergy.
During the analysis of the document, several reinforcing feedback loops were identified that would lead to higher "Availability of sustainable biomass resources" (capitalized, Figure 4). As the main proponent and champion of the regional strategy, government investment in technology and training for sustainable biomass production ("Public investment in technology and training I") drives the development of "Skills in sustainable biomass production" as well as the "Deployment of sustainable technologies for biomass production" both directly (reinforcing feedback loops R1a, R1b, and R1c) and through the enhancement of private investment in these activities (reinforcing feedback loops R2a, R2b, and R2c). All these together trigger an increase in the "Share of land and water used for sustainable biomass production", which in turn leads to a higher "Volume of biomass produced sustainably". The higher "Availability of sustainable biomass resources" thus achieved consequently induces an increase in both the "Use of sustainable biomass in bioindustrial processes" and the "Use of biomass transformation and conversion technologies" (reinforcing feedback loops R1d and R2d), resulting in a higher "Volume of industrial production of bioproducts and bioenergy" and increased "Income from bioproducts and bioenergy" (as well as from related services the economic, environmental, and social benefits that accrue from these activities close the loop by prompting more public and private investment in technology and training for sustainable biomass production. The private sector also can promote these cycles (reinforcing feedback loops R3a, R3b, R3c, and R3d), but the impact on the system is lower if it does it alone.
Three balancing feedback loops were identified. Water and land of good quality are often limited resources, but this condition is exacerbated in Andalusia due to the region's geographic characteristics. Therefore, the more land is used to produce biomass for bioproducts and bioenergy, the lower the potential to further expand sustainable biomass production for these purposes (delayed balancing feedback B1). Likewise, the more biomass is used in bioindustrial processes, the lower the potential to further expand these activities (delayed balancing feedback B2). And as new entrants are attracted to the region's circular bioeconomy industry due to increasing income from bioproducts and bioenergy, higher competition for resources would eventually become a limiting factor (delayed balancing feedback B3).

Infrastructure and Logistic Management of Biomass Resources
The second strategic line described in the ACBS relates to optimizing the management and distribution of biomass resources from the points where they are generated to the bioindustries that use them as inputs. Several reinforcing feedback loops were identified during the analysis that would lead to higher "Economic viability of biomass collection and storage" (capitalized, Figure 5). R6a + Logistics and Transportation Infrastructure" and increasing the "Proximity to biomass resources", which in turn would contribute to boost the "Profitability of biomass utilization". Lastly, as the first results of this collaborative work become evident, the model would be replicated across the region through more public and private investment in logistics and transportation infrastructure.
And as in the previous model, the attractiveness of the growing market would attract new entrants which, over time, would have a negative impact on returns because of higher competition for the limited resources available (delayed balancing feedback B4).

Processing of Biomass Resources and Capacity of Industrial Production of Bioproducts and Bioenergy
The third strategic line defined in the ACBS comprises actions to support the development of a biobased industry that optimizes the use of biomass resources in Andalusia, especially through integrated biorefineries. As in the previous cases, several reinforcing feedback loops were identified ( Figure 6).  According to this model, investments in technology and training by the public and private sector would lead to the development of "New models of use of biomass resources and industrial CO2".
As synergies are achieved, new companies will be created, new biorefineries would be built, and existing facilities would be reconverted to increase the "Industrial use of transformation and conversion technologies" and expand the "Volume of industrial production of bioproducts and bioenergy". As in the previous models, private investment alone can generate positive results (reinforcing feedback loop R9), but public sector involvement can potentiate the system through direct investments, financial support, and positive market signals (reinforcing feedback loops R7 and R8). The income thus generated from bioproducts and bioenergy (as well from related services) would produce economic, environmental, and social benefits that would in turn encourage more public and private "Investment in technology and training" for biomass transformation and conversion. However, as in the previous models, once the biomass processing sector reaches a critical mass, its attractiveness would encourage the entry of new players up to a point where the competition for resources would become a limiting factor (delayed balancing feedback B3).

Development of Markets for Bioproducts and Bioenergy
The fourth and last strategic line formulated in the ACBS consists of actions aimed at consolidating the markets that already exist in Andalusia while promoting and supporting the development of national and international value chains for bioproducts and bioenergy. As described in Figure 7, several reinforcing feedback loops were identified that would lead to higher sales of bioproducts, bioenergy, and related bioeconomy services. Once again, synergistic "Investments in market preparedness" by the public and private sector would lead to enhancements in the collective "Market knowledge" about opportunities for bioproducts, bioenergy, and bioeconomy-related services. With this valuable information at hand, both government planning and corporate business plans would be upgraded to support "Market development activities" aimed at increasing "Market awareness about the benefits of using bioproducts and bioenergy" and triggering the "Cultural change in regional, national and international markets" that is needed to switch consumption towards products and energy obtained from sustainable biomass resources. Lastly, higher sales of bioproducts, bioenergy, and related bioeconomy services to local, regional, and national customers (reinforcing feedback loops R10a, R11a, and R12a), as well as in international markets (reinforcing feedback loops R10b, R11b, and R12b) would close the loop by increasing the "Profitability of circular bioeconomy companies", which would in turn stimulate more public and private investments in market development activities.
As in the previous strategic lines, the finite availability of biomass resources would eventually

Combined Causal Loop Diagrams and Key Intervention Points
While the individual CLDs in Figures 4 to 7    stock of knowledge about technologies for biomass transformation and conversion, the various types of bioproducts and bioenergy generated from these sources, and the rising income from these activities (IP7, IP8, and IP9). Above all of these, given that the circular bioeconomy is an emerging sector that will require ongoing government support for some time, political commitment to public investments in technology and training is key for its success. Table 1 contains a list of targeted learning-and knowledge-related actions identified from the ACBS and the authors' analysis that could be implemented with meaningful impact at these intervention points. • Implement mechanisms and tools to increase the interaction between all levels of the education system and the actors involved in the generation of biomass to promote technical advice, education and training in matters related to sustainable practices. Adjust the offer to the needs of the market.
( • Foster the incorporation of technical and business skills into the knowledge of people who work in bioindustries to improve the sustainability of companies and increase the addition of value to the region's biomass resources. • Develop guidelines and case studies based on regional, national, and international success stories to disseminate knowledge about new models of use of biomass resources and industrial CO2. (3.2.2) • Develop guidelines and case studies based on regional, national, and international success stories to disseminate knowledge about the planning and implementation of bioindustries and biorefineries.

IP8 Use of Biomass Transformation and Conversion Technologies
• Implement mechanisms and tools to increase the interaction between all levels of the education system and the bio-based industry to promote technical advice, education and training in matters related to the use of technologies for sustainable biomass transformation and conversion. Adjust the offer to the needs of the market. • Foster collaboration and the establishment of regional, national, and international alliances and multi-actor platforms to facilitate the transfer of knowledge and the adoption of sustainable technologies generated by the R&D and innovation system. As for the strategic lines 2 and 4 of the ACBS, negligible overlap was observed among them and 8 with the others, as shown in Figure 9.   • Improve knowledge about biomass resources and industrial sources of CO2 among public sector staff and policy makers in terms of those factors that determine their logistics, as well as about the infrastructure that is required to ensure the supply of biomass to operators, users, and bioindustries.
(1. • Promote knowledge among public sector staff and policy makers about potential users of biomass resources and infrastructure gaps in the region. (2.1.1) • Foster knowledge among public sector staff and policy makers about alternative ways of collaborative financing and publicprivate partnerships (PPP).
Private Investment in Infrastructure and Logistics • Upgrade technical and management skills in the private sector to increase the absorptive capacity of companies to receive public funding and improve the quality of their investment decisions in infrastructure and logistics for biomass collection, transportation, pretreatment, and storage.
(C.1.1) (C.1.2) • Promote knowledge within the private sector about alternative ways of collaborative financing and public-private partnerships (PPP). (C.1.2) • Foster initiatives to disseminate knowledge among local, national, and foreign investors about the region's competitive advantage in the bioeconomy and specific investment opportunities in infrastructure and logistics for biomass collection, transportation, pretreatment, and storage.

IP12
Public-Private Collaboration • Develop and regularly update an inventory of biomass resources in the region to improve public and private sector knowledge about the types and volumes available, their physical and chemical characteristics, their geographic location, and the distribution of their availability over time (seasonality). (1.1.1) • Improve public and private sector knowledge about both the infrastructure needed and available for the collection, transportation, pretreatment, and storage of the different types of biomass resources. (2.1.2) • Develop and regularly update a georeferenced inventory of potential users of biomass resources. (2.1.1) • Develop guidelines and case studies to improve public and private sector knowledge regarding best practices in public-private collaboration and about current and upcoming solutions for collection, storage, pretreatment, and transportation of biomass resources. (2.

IP14 Public Investment in Market Preparedness
• Improve knowledge among public sector staff and policy makers regarding the needs and gaps faced by the bio-based sector in matters related to market preparedness. improve the quality of their investment decisions in market preparedness.

IP16
Market Knowledge • Carry out and disseminate market studies and feasibility analyses to determine supply and demand, prices, and distribution channels for bioproducts, bioenergy and services linked to the circular bioeconomy. (4.1.1) • Prepare, regularly update, and disseminate prospective studies on consumption trends and new uses of bioproducts and bioenergy, as well as analyses of areas and/or sectors where there is potential for the introduction of bio-based products in their value chains. (4.1.2) • Foster interaction and knowledge exchange among bio-based companies and with actors of other industry sectors. Facilitate communication and cooperation among the different agents, especially with the nonrenewable sector, to promote the identification of potential synergies. • Monitor the evolution of the different links that make up the value chains of bioproducts and bioenergy (supply of raw materials, production, and commercialization) and promote knowledge exchange across the bio-based sector to improve efficiencies and ensure its ability to meet market demand in the medium and long term.    • Develop and disseminate a portfolio of guidelines and case studies about successful initiatives and business models for the introduction of bioproducts, bioenergy, and bioeconomy-related services in regional, national, and international markets.

39
The application of the system dynamics approach to the analysis of Andalusia's circular 40 bioeconomy provides important insights into the complexity of the system due to the existence of 41  R5b   IP3   IP4   IP5   IP6   IP8   IP7   IP9   IP12   IP13   IP11   IP16   IP17  IP15   +   IP1   IP2   IP10   IP14   IP18   IP19  IP20 non-linear processes, multiple feedback loops, and time delays. Likewise, the models generated in 42 this study provide tools for a better understanding of the potential impact that learning-and 43 knowledge-related interventions by governments and other actors could have on different parts of 44 the development of the circular bioeconomy. and social pull will play a leading role in the realization of the potential presented by the circular 48 bioeconomy and, in realizing this potential, it will be fundamental to keep in mind that an important 49 aspect of the innovation process is its heterogeneity across sectors, industries, and regions. In this 50 regard, assessing and measuring the underlying processes of learning and knowledge accumulation 51 for innovation has been an ongoing challenge for decades (Abramovitz, 1956;Dosi, 1982;Romer, 52 1990; Solow, 1957). This has had important repercussions for technology, innovation, and 53 economic development policymaking, which has strongly relied on a linear R&D-based innovation 54

69
Overall, the learning-and knowledge-related actions identified in this study to accelerate 70 Andalusia's transition from a linear bio-based economy to a sustainable circular bioeconomy 71 support the view that an innovation system that encourages a combination of STI and DUI activities 72 will have the greatest potential of success as each company will choose the learning mode that best 73 suits its scientific, technological, and geographic context (  regions that adopt technologies from leading regions need to incur expenses in the retraining of 90 human capital, organizational restructuring, and so on (Stoneman, 1983). Furthermore, as the 91 characteristics of new technologies are strongly influenced by the business environments in which 92 they are developed, large differences between regions can represent serious barriers for their 93 transfer. Among the factors that have been suggested to play a crucial role in a region's ability to 94 import technology are its political, commercial, industrial, and financial institutions, as well as 95 national characteristics such as market size and the relative supply of factors of production 96 (Abramovitz, 1986;Caselli and Coleman, 2001;Chen and Wang, 2021;Keller, 2004). 97 Technological change is therefore the combined result of innovation and learning activities within 98 domestic organizations, and of the interaction among them and with their environment. It thus 99 becomes obvious that firms, with their different combinations of intrinsic competencies and 100 business strategies, are key players in this process (Fagerberg, 1994).

101
As the costs of adoption of new technologies tend to exceed the benefits obtained, these are often 102 not adopted as soon as they become public (Stoneman, 1983). The process of technological 103 diffusion then occurs gradually, with a few firms adopting the technology first and the others 104 following in a process that can take more than a decade or even stop before its completion 105 (Detragiache, 1998). Remarkably, as the followers have the possibility of copying the adaptive 106 efforts of the pioneers and, at the same time, have access to more skilled labor trained by the early 107 adopters, the costs of adoption tend to decrease as more firms import the technology. Based on the 108 premise that early adopters create positive externalities for the followers, Detragiache (1998) 109 proposed that, once technology adoption starts, less developed regions tend to converge to the level 110 of developed regions as more firms adopt the new technology. Accordingly, different degrees of 111 economic convergence among regions are explained by differences in the rate of diffusion of the 112 imported technologies, whereas convergence failure occurs when the adoption costs are too high or 113 when technology diffusion stops before it is completed. Furthermore, this model can successfully 114 account for the fact that economies of transition regions are generally dualistic, i.e., small, traditional 115 firms normally operate along with modern enterprises. In this scenario, it has been argued that when 116 a technology is transferred from a developed region to less developed regions, the discrepancies 117 between the labor skills lead to marked differences in factor productivity and output per capita and, 118 for this reason, governments of less developed and transition regions need to increase their 119 investments in human capital formation (Acemoglu and Zilibotti, 2001). 120 In this regard, our findings show that the role of government in supporting learning and knowledge-121 related processes is key for the development of the circular bioeconomy, which can be explained by 122 the pervasiveness of information asymmetries in the sector, its intensity in knowledge and 123 innovation, and its position in the confluence of several technological areas. This is supported by a 124 recent causal mapping analysis of political structures in bioeconomic transitions based on the case 125 of renewable energy political lobbying in six countries (Palmer et al., 2022), and is in agreement 126 with the New Structural Economics framework postulated by Lin (2003Lin ( , 2010, whereby the 127 positive impact of government in this kind of scenario is higher when it seeks to help companies to 128 overcome information and coordination costs about new industries, markets, and technologies; 129 coordinate investment between companies and industries; and internalize the externalities linked to 130 information by compensating pioneering companies through tools such as guarantees and fiscal 131 incentives. Furthermore, from the innovation policy perspective, a combination of supply-driven 132 policies aimed at the commercialization of research results will be required to foster the STI mode 133 of learning, along with demand-driven policies aimed at supporting the DUI mode of learning for 134 the development of products or services to specific markets (Isaksen and Nilsson, 2013). Lastly, due 135 to the emerging nature of the concept, the circular bioeconomy sector is likely to require ongoing 136 government support for some time, and for this reason political commitment to public investments 137 in technology and training will be key to its success. In Andalusia, both the ACBS and the bill for 138 the Circular Economy Law of Andalusia (LECA) that has recently been sent by the Government 139 Council to the regional Parliament for deliberation are the two most important initiatives currently 140 underway in this direction.

142
The results outlined in this paper provide an initial understanding of the dynamics of Andalusia's 143 bioeconomy and the identification of intervention points where targeted actions could be undertaken 144 to accelerate the transition from a linear bio-based economy to a sustainable circular bioeconomy. 145 The models generated in this study provide tools for a better understanding of the potential impact 146 that interventions by governments and other actors could have on the development of the circular 147 bioeconomy. Overall, when confronted with the current scenario, the preliminary cause-effect 148 relationships and causal cycles herein described suggest that the Andalusian innovation system 149 needs greater collaboration and coordination along and across the triple helix to support the 150 development, commercialization, and diffusion of innovative solutions, all of which are necessary 151 for the development of critical mass in this emerging sector of the economy. While the CLDs herein 152 described provide structural insight into the system, an avenue for future research involves the 153 development of quantitative models using stock and flow diagrams to evaluate the sensitivity of the 154 intervention points, using historical data as a reference. 155 Ketzer, D., Schlyter, P., Weinberger, N., and Rösch, C. (2020). Driving and restraining forces for 307 the implementation of the agrophotovoltaics system technology: a system dynamics analysis.