The transition to bioeconomy and its implications for sustainable development: The case of Germany

Wen, Lanjiao Doctoral Thesis The transition to bioeconomy and its implications for sustainable development: The case of Germany Suggested Citation: Wen, Lanjiao (2024) : The transition to bioeconomy and its implications for sustainable development: The case of Germany, Universitätsund Landesbibliothek Sachsen-Anhalt, Halle (Saale), https://nbn-resolving.de/urn:nbn:de:gbv:3:4-1981185920-1203362 , https://hdl.handle.net/1981185920/120336 This Version is available at: https://hdl.handle.net/10419/312663 Standard-Nutzungsbedingungen: Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden. Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich machen, vertreiben oder anderweitig nutzen. Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, gelten abweichend von diesen Nutzungsbedingungen die in der dort genannten Lizenz gewährten Nutzungsrechte. Terms of use: Documents in EconStor may be saved and copied for your personal and scholarly purposes. You are not to copy documents for public or commercial purposes, to exhibit the documents publicly, to make them publicly available on the internet, or to distribute or otherwise use the documents in public. If the documents have been made available under an Open Content Licence (especially Creative Commons Licences), you may exercise further usage rights as specified in the indicated licence. https://creativecommons.org/licenses/by/4.0/ The transition to bioeconomy and its implications for sustainable development: The case of Germany Dissertation Zur Erlangung des Doktorgrades der Agrarwissenschaften (Dr.agr.) der Naturwissenschaftlichen Fakultät III Agrar‐ und Ernährungswissenschaften, Geowissenschaften und Informatik der Martin‐Luther‐Universität Halle‐Wittenberg vorgelegt von Frau Lanjiao Wen Gutachter: (1) Prof. Dr. Alfons Balmann (2) Prof. Dr. José Gil Roig Tag der Verteidigung: 9. Dezember 2024 i Acknowledgment Finally, after navigating the challenges of the COVID-19 pandemic and welcoming my child, I approach the final part of this dissertation with mixed emotions. This dissertation marks the culmination of my second PhD journey, a path filled with motivation, hardship, encouragement, frustration, and fulfilment. This journey has strengthened my resilience, helping me realize that dreams can be achieved through unremitting efforts. I have so much to be grateful for, particularly for those who have helped me reach this point. First and foremost, I would like to express my deepest gratitude to my supervisors. I am extremely thankful to my “Doktorvater”, Prof. Dr. Alfons Balmann, for allowing me to pursue this work with scientific and financial support. Prof. Dr. Alfons Balmann consistently identified areas for improvement in my work and offered constructive suggestions. I am also deeply grateful to my daily supervisor, Dr. Zhanli Jerry Sun, who offered me great freedom to explore my ideas and supported me with invaluable advice. Thank you for adopting me when I was helpless and holding me up when I was climbing the mountain of challenges. You are not only my supervisor but also my spiritual mentor. I cannot imagine finalizing this dissertation without your support. Second, I would like to sincerely thank my collaborators. I thank my collaborators, Dr. Lioudmila Chatalova, Prof. Dr. Anlu Zhang, Dr. Ir. Frans Hermans, Prof. Dr. Van Butsic, Prof. Dr.Zhiyong Fox Hu, Dr. Xin Gao, and Dr. Wenfang Pu for their supportive and constructive input. Special thanks to Dr. Lioudmila Chatalova who taught me a lot in science and writing skills, Prof. Dr. Anlu Zhang, one of my Thesis Advisory Committee (TAC) members, for your continued support and helpful advices regarding my research career and Dr. Ir. Frans Hermans, one of my TAC members, for your insightful comments regarding the bioeconomy and continued support and help. ii Third, I would also like to express my appreciation to my colleagues at IAMO. Thanks to the colleagues in the Bioeconomy group and China group where I found my research family. I would acknowledge the support from the BSE members, particularly Prof. Dr. Daniel Müller, Dr. Franziska Hauff, Dr. Stephan Brosig, Dr. Franziska Appel, Dr. Changxing Dong, Dr. Florian Schierhorn, Dr. Lijuan Miao, Dr. Brian Beadle, Mr. Florian Heinrich, and Ms. Claudia Grützmacher. I also appreciate the friendly help from other IAMO colleagues, especially Dr. Sarvarbek Eltazarov, Dr. Laura Moritz, Dr. Sören Prehn, Dr. Nodir Djanibekov and Prof. Dr. Thomas Herzfeld. Thanks to IAMO for supporting my stay and providing me a family-friendly working atmosphere. In addition, I am grateful for the support and assistance of Dr. Milad Abbasiharofteh, Prof. Dr. Sebastian Lakner, Prof. Dr. Justus Wesseler and Dr. Yan Jin. I extend thanks to Dr. Milad Abbasiharofteh for providing me the data and teaching me how to extract it; to Prof. Dr. Sebastian Lakner for supplying data on EFA; to Prof. Dr. Justus Wesseler and Dr. Yan Jin for their valuable feedback on my research idea about bioeconomy. There are countless colleagues who have helped me along the way, far too many to name individually. However, please know that each of you holds a special place in my heart. My last gratitude is to my family members and friends, who know I am not perfect yet love me unconditionally. Many thanks to my husband, Lvmeng, for your lasting love, encouragement and involvement in looking after our family members when I was absent; to my lovely daughter, Yogurt, for giving me huge courage to face life’s challenges without hesitation; to my parents, for setting me on the path to pursue my dreams, raising me, and helping to raise my daughter; to my parents-in-law for assisting us in setting up our home. Thanks to my relatives and friends in Europe and China for your companionship and support. I also thank myself for never giving up. Though my second PhD journey is nearing its end, my research career is far from over. I will continue moving forward, always striving to become a better version of myself. Halle (Saale), September 2024 Lanjiao Wen iii Summary Ensuring that climate neutrality, the bioeconomy, and economic competitiveness go hand in hand has become a key European goal for future sustainable development. Germany, as one of the leading countries in the modern bioeconomy, aims to boost its bioeconomy as a key strategy to achieve the Sustainable Development Goals (SDGs). The transition to bioeconomy depends not only on a sufficient biomass supply but also on supporting technological and institutional innovations. In Germany, the role that technological innovation can play in carbon emissions reduction in the agricultural system is far from clear. Currently, Germany has low R&D productivity due to a general production factor mismatch and low allocation efficiency. Additionally, little is currently known about the performance of institutional innovation in the bioeconomy. Therefore, a better understanding of the real impacts of technological and institutional innovation on sustainable development is critical for supporting policymaking and guiding the transition to bioeconomy. This dissertation focuses on the bioeconomy in Germany, examining the mechanisms by which technological and institutional innovations influence, promote, and support its development. Specifically, this dissertation i) evaluates the potential impact of R&D investments on carbon emissions through the dynamic interactions among agricultural carbon subsystems using a system dynamics modelling approach based on sectoral data (Chapter II), ii) estimates the potential mitigation effects of technological innovation on carbon emissions using an extended Spatial Durbin Model based on 401 NUTS-3 level panel data (Chapter Ⅲ), and ⅲ) examines the impacts of bioclusters, representing regional institutional innovation in Germany, on sustainable performance through the use of a super slacks-based measure (super-efficiency SBM), a series of quasi-natural experiments, and a mediating model based on 401 NUTS-3 level panel data (Chapter Ⅳ). iv Chapter II reports the modelling and analysis of various scenarios, where the simulations of the dynamic interactions in the agricultural carbon system from 2020 to 2050 suggest that R&D investments can have a mitigation effect on agricultural carbon emissions both directly and indirectly, with the direct effect being more significant. The result suggests that increasing the fallow land, improving the circular economy, and increasing R&D investment are effective strategies for reducing net carbon emissions. These strategies can provide an efficient and more sustainable pathway for the transition to bioeconomy in Germany. In Chapter III, the results of the study regarding the implication of a forest-based bioeconomy on carbon emissions are presented and suggest that technological innovations in a forest-based bioeconomy can reduce carbon emissions through promoting industrial upgrading and creating job opportunities related to the bioeconomy in local areas. Additionally, it can lower carbon emissions indirectly in neighbouring areas through the spillover effects of industrial upgrading and the size of the bioeconomy. These findings highlight the need for a coordinated approach to align technological innovation (as indicated by the number and application of patents), employment population, and industrial transition strategies. Chapter IV investigates the potential effects of bioclusters on green total factor productivity (GTFP), and the results indicate that developing bioclusters, including both Bioregions and green clusters, would have positive effects on GTFP, both directly and indirectly, essentially through technological innovation and market agglomeration. Furthermore, it reveals that different types of bioclusters have heterogeneous impacts on GTFP, with the greatest contribution arising from chemical green clusters. By analysing various aspects of the bioeconomy development and its implications, the findings in this dissertation contribute to the field and provide insights that can inform and support ongoing and future scientific and policy actions that guide the transition to bioeconomy in Germany. Keywords: Bioeconomy; Technological innovation; Institutional innovation; Bioclusters; Carbon emissions; Mitigation effects; Dynamic interactions; Spillover effects; GTFP; Germany v Zusammenfassung Die Sicherstellung, dass Klimaneutralität, Bioökonomie und wirtschaftliche Wettbewerbsfähigkeit Hand in Hand gehen, ist zu einem zentralen europäischen Ziel für eine nachhaltige Zukunftsentwicklung geworden. Deutschland, als eines der führenden Länder in der modernen Bioökonomie, zielt darauf ab, seine Bioökonomie als Schlüsselstrategie zur Erreichung der Ziele für nachhaltige Entwicklung zu fördern. Der Übergang zur Bioökonomie hängt nicht nur von einer ausreichenden Biomasseversorgung ab, sondern auch von der Unterstützung technologischer und institutioneller Innovationen. In Deutschland ist die Rolle, die technologische Innovation bei der Reduzierung von landwirtschaftlichen Kohlenstoffemissionen spielen kann, noch unklar. Derzeit weist Deutschland eine geringe F&E-Produktivität aufgrund eines allgemeinen Missverhältnisses der Produktionsfaktoren und einer niedrigen Allokationseffizienz auf. Zudem ist wenig über die Leistungsfähigkeit institutioneller Innovationen in der Bioökonomie bekannt. Daher ist ein besseres Verständnis der tatsächlichen Auswirkungen von technologischen und institutionellen Innovationen auf die nachhaltige Entwicklung entscheidend für die Unterstützung der politischen Entscheidungsfindung und die Steuerung des Übergangs zur Bioökonomie. Um die genannten Punkte zu adressieren, konzentriert sich diese Dissertation auf die Bioökonomie in Deutschland und untersucht die Einflussmechanismen technologischer und institutioneller Innovationen bei der Förderung und Unterstützung der Bioökonomie. Konkret bewertet diese Dissertation: (1) den potenziellen Einfluss von F&E-Investitionen auf die Kohlenstoffemissionen durch die dynamischen Interaktionen zwischen landwirtschaftlichen Kohlenstoffsubsystemen mithilfe eines SystemdynamikModellierungsansatzes auf Grundlage sektoraler Daten (Kapitel II), ii) die potenziellen Minderungseffekte technologischer Innovationen auf Kohlenstoffemissionen mithilfe eines erweiterten Spatial DurbinModells basierend auf Paneldaten auf NUTS-3-Ebene (Kapitel III); und iii) die Auswirkungen von Bioclustern, die regionale institutionelle Innovationen in Deutschland repräsentieren, auf die nachhaltige vi Leistung durch den Einsatz eines Super Slacks-basierten Maßes (super-effizientes SBM), eine Reihe von quasi-natürlichen Experimenten und ein Mediationsmodell basierend auf Paneldaten auf NUTS-3-Ebene (Kapitel IV). Kapitel II berichtet über die Modellierung und Analyse verschiedener Szenarien, bei denen die Simulationen der dynamischen Interaktionen im landwirtschaftlichen Kohlenstoffsystem von 2020 bis 2050 darauf hinweisen, dass F&E-Investitionen sowohl direkt als auch indirekt eine Minderung der landwirtschaftlichen Kohlenstoffemissionen bewirken können, wobei der direkte Effekt signifikanter ist. Das Ergebnis legt nahe, dass die Erhöhung der Brachfläche, die Verbesserung der Kreislaufwirtschaft und die Erhöhung der F&E-Investitionen wirksame Strategien zur Reduzierung der NettoKohlenstoffemissionen sind. Diese Strategien können einen effizienten und nachhaltigeren Weg für den Übergang zur Bioökonomie in Deutschland bieten. In Kapitel III werden die Ergebnisse der Studie über die Auswirkungen einer wald-basierten Bioökonomie auf die Kohlenstoffemissionen präsentiert und legen nahe, dass technologische Innovationen in einer waldbasierten Bioökonomie die Kohlenstoffemissionen durch die Förderung der industriellen Aufwertung und die Schaffung von Arbeitsplätzen im Zusammenhang mit der Bioökonomie in lokalen Gebieten reduzieren können. Zudem können sie indirekt die Kohlenstoffemissionen in benachbarten Gebieten durch die Spillover-Effekte der industriellen Aufwertung und die Größe der Bioökonomie senken. Diese Erkenntnisse unterstreichen die Notwendigkeit eines koordinierten Ansatzes zur Abstimmung von technologischer Innovation (angezeigt durch die Anzahl und Anwendung von Patenten), Beschäftigung, Bevölkerung und industriellen Übergangsstrategien. Kapitel IV untersucht die potenziellen Auswirkungen von Bioclustern auf die grüne totale Faktorproduktivität (GTFP) und die Ergebnisse zeigen, dass die Entwicklung von Bioclustern, einschließlich Bioregionen und grünen Clustern, positive Auswirkungen auf die GTFP sowohl direkt als auch indirekt haben würde, im Wesentlichen durch technologische Innovation und Marktagglomeration. xiii List of tables Table 2.1: Coefficients for carbon emissions calculation ........................................................................... 26 Table 2.2: Relative errors in the simulation (%) ......................................................................................... 28 Table 2.3: Scenario schemes for different scenarios ................................................................................... 30 Table 3.1: Descriptions of the variables ...................................................................................................... 52 Table 3.2: Moran’s I values of net carbon emissions from 2000 to 2021 ................................................... 60 Table 3.3: Estimation results from the Spatial Durbin Model and its extensions ....................................... 61 Table 3.4: Direct effect, indirect effect, and total effect of the model parameters...................................... 64 Table 4.1: Input-output indicators for GTFP .............................................................................................. 78 Table 4.2: Descriptive statistics of variables .............................................................................................. 83 Table 4.3: Estimation results of SDiD model for Bioregions ..................................................................... 88 Table 4.4: Estimation results of SDiD model for green clusters ................................................................. 89 Table 4.5: PSM-SDiD regression results .................................................................................................... 93 Table 4.6: Mediating regression results for Bioregions .............................................................................. 94 Table 4.7: Mediating regression results for green clusters ......................................................................... 96 Table 4.8: Regression results of DDD model for Bioregions ..................................................................... 97 Table 4.9: Regression results of DDD model for green clusters ................................................................. 99 xiv List of figures in the appendix Figure A. 1: Results for Placebo test ......................................................................................................... 150 xv List of tables in the appendix Table A. 1: Main variables and parameter indexes in the SD model during 2000-2050 .......................... 139 Table A. 2: Scenario design ...................................................................................................................... 146 Table A. 3: Biocluster classification at the 3 digit level ........................................................................... 146 xvi List of abbreviations BMBF German Federal Ministry of Education and Research BMEL German Federal Ministry of Food and Agriculture BMELV German Federal Ministry of Food, Agriculture and Consumer Protection BMU German Federal Ministry for the Environment, Nature Conservation and Nuclear Safety CAP Common Agricultural Policy EC European Commission EFA Ecological Focus Area Eurostat European Statistical Office EEA European Environment Agency DEA Data envelopment analysis DDD Difference in Difference in Differences SDiD Staggered Difference in Differences PSM-SDiD Propensity Score MatchingStaggered Difference in Differences FAO Food and Agriculture Organization GHG Greenhouse Gas GTFP Green total factor productivity IPCC Intergovernmental Panel on Climate Change NCE Net carbon emissions NUTS Nomenclature of territorial units for statistics R&D Research and Development SD System Dynamic Sd Standard deviation SDG Sustainable Development Goal SDM Spatial Durbin Model 1 1 Introduction Developing the bioeconomy has become a key strategy for facilitating the transition towards sustainability in the European Union (EU) and in many other regions worldwide. Many countries have launched bioeconomy strategies, ranging from dedicated bioeconomy strategies to be-related strategies and dedicated be-strategies (see Figure 1.1). Figure 1.1 displays an overview of the distribution of bioeconomy strategies around the world, indicating countries where strategies are already in place or are under development. Germany is included in the figure as one of the countries that has already launched a dedicated bioeconomy strategy. Figure 1.1: Bioeconomy strategies in place or under development around the world Source: bioökonomierat (2018). “International bioökonomiestrategien”. https://bioökonomierat.de/bioökonomie/international. 2 Although the term “bioeconomy” first started to become popular in the early 2000s, the concept of the bioeconomy is still relatively new and it is considered to still be in its early growth stage. The concept of the bioeconomy has a multifaceted breadth and depth of meaning, varying across paradigms, disciplines, and countries. While definitions of the bioeconomy may differ, they typically share many similarities (Wesseler and von Braun, 2017). According to the United Nations Food and Agriculture Organisation (FAO), the bioeconomy refers to “the production, use, and conservation of biological resources, including related knowledge, science, technology, and innovation to provide information, products, processes, and services to all economic sectors with the aim of moving towards a sustainable economy” (FAO, 2018). The bioeconomy was defined by McCormick and Kautto (2013) as an economy where the basic building blocks for materials, chemicals, and energy are derived from renewable biological resources. In the policy framework of the European Union, the bioeconomy is regarded as a key component for attaining smart and green growth (EC, 2012). According to the European Commission, the bioeconomy “encompasses the production of renewable biological resources and their conversion into food, feed, bio-based products, and bioenergy. This includes agriculture, forestry, fisheries, food, pulp, and paper production, as well as parts of the chemical, biotechnological, and energy industries” (EC, 2012). This definition is widely accepted by academic and political communities all around the world. With the goals of ensuring food security, managing depleting natural resources sustainably, reducing the dependence on non-renewable resources, adapting to climate change, and creating job opportunities, the bioeconomy aims to contribute to intelligent, sustainable, and inclusive growth that will allow the transition towards a green economy (OECD, 2011a, b; 2016). 1.1 Background: The transition towards bioeconomy in Germany 1.1.1 The bioeconomy in Germany The development of the bioeconomy in Germany is being driven by policy initiatives aimed at modernizing the economy in a sustainable, environmentally responsible and societally sensitive manner. With the aim to develop a cross-sectoral, knowledge and bio-based economy, the bioeconomy in Germany began with 3 the establishment of the Bioeconomy Council in January 2009. The council was established by the Federal Ministry of Education and Research (BMBF) and the Federal Ministry of Food, Agriculture, and Consumer Protection (BMELV) and was regarded as an independent advisory board for the German Federal Government. In 2010, Germany published the National Research Strategy Bioeconomy 2030, which was designed by the Bioeconomy Council, becoming one of the first countries to outline its national bioeconomy strategy. In 2013, Germany implemented the National Policy Strategy Bioeconomy, setting another important milestone for the bioeconomy. In early 2020, Germany launched the new National Bioeconomy Strategy, which laid down guidelines for policies on the bioeconomy as well as the measures for implementation. With these and later initiatives, Germany has set a pioneering pace as one of the first countries to formally set out and pursue a bioeconomy strategy in line with the EU Framework Programme for Research and Innovation and later in 2012, the EU Bioeconomy Strategy. Although revolutionary in their ambition to sustainably transform the entire society, these strategies started with a step backwards, namely by revisiting the potential of plant-based biomass. In April 2009, the National Biomass Action Plan was launched (BMELV/BMU, 2009; Goven and Pavone, 2015; Hagemann et al., 2016). Alongside 2009 amendments to the Renewable Energy Sources Act and a boom in renewable electricity uptake, this plan defined forest and agricultural biomass as one of the most promising domestic renewable energy sources that could significantly contribute to value creation, especially in rural areas (Troost et al., 2015). The plan envisaged a large-scale expansion of bio-based energy, including agricultural fuels. Recent European Union strategy papers on biodiversity (EC, 2020b) and food systems (EC, 2020c), along with recommendations by the National Academy of Sciences (Leopoldina, 2020), have further clarified the role of biomass in the bioeconomy. Alarmed by the increasingly pessimistic projections for climate development, soil, water, and air quality, as well as by the noticeable consequences of unsuitable development paths, recent debate has begun to shift the focus from alternative models of economic growth to prioritizing environmental protection. The Biodiversity and Food to Fork strategies (EC, 2020b; 2020c), the core components of the European Green 4 Deal (EC, 2019), set time-bound targets for the expansion of nature conservation areas, with the aim of achieving farmland biodiversity and land degradation neutrality within a decade. The Leopoldina stresses the major role of agriculture in reducing carbon emissions and biodiversity loss and recommends even more profound and immediate actions are needed (Leopoldina, 2020). These recommendations, including minimizing the use of fertilizers, pesticides, and herbicides, a large-scale shift to organic farming, and limiting farmland for biofuels and animal feed production, have direct implications for non-food biomass production. Despite the conflicting goals of sustainability and economic growth reflected in regional, national, and supranational agendas, the bioeconomy is gaining momentum along various dimensions (Bell et al., 2018). In 2005, the share of the bioeconomy in Germany accounted for 3.9% of gross value added and 5.2% of the labour force, while in 2019 these figures had risen to 19.9% and 13.5%, respectively (Bioeconomy Council, 2010; BMBF, 2020). This growth has been driven by significant research funding (EUR 20.3 billion in 2020, cf. BioStep (2016)), invested in “mapping and engineering the uncharted territories” of the technical and biotechnological knowledge and making them marketable (Aguilar et al., 2018). The transformation of the economy and especially of the chemical sector away from fossil-based resources (Schütte, 2018), along with the promotion of bioclusters and technology parks (Scarlat et al., 2015; BioSTEP, 2016), could accelerate this process by demanding more high-quality biomass (Budzinski et al., 2017; Efken et al., 2016). However, the envisaged production of high-value biomass-based goods and materials with economic and non-economic benefits may prove an unattainable vision (Brar et al., 2013), considering the already high imports of biomass for material and energy use (Leopoldina, 2012). 1.1.2 Technological innovation in the agricultural system in Germany In the context of Germany, the role of technological innovation in reducing agricultural carbon emissions is still unclear, as the agricultural system is complex and the productivity of R&D is difficult to estimate. By comparison, while Germany’s business R&D spending has increased by 3.3 % per year over the last decades, R&D productivity has fallen by an average of 5.2% per year (Boeing and Hünermund, 2020). This 5 aligns with the findings that R&D productivity is decreasing in Germany, particularly outside the bioeconomy context (Schäfer, 2014; Ugur et al., 2016). In 2016, the total public R&D funding for the bioeconomy in Germany amounted to around EUR 120 million (Imbert et al., 2017). Through R&D investments, a number of local biotech innovation networks and bioeconomy clusters in Germany have integrated biomass producers within a circular economy to support innovation and coupled subsystems (Kaiser and Prange, 2004; Mennicken et al., 2016; Wilde and Hermans, 2021). This highlights the interaction between innovation and coupled subsystems in the German plant-based bioeconomy. In addition, Germany aims to fulfil all its electricity needs from renewable sources by 2035, with two-thirds expected to come from bioenergy (Frondel et al., 2010; Mohmmed et al., 2019). The projected decrease in energybased emissions by 2030 with this energy transition due to the political incentives promoting renewable energy (754.883 Mt CO2, cf. Mohmmed et al., 2019) underscores the relevance of coupled subsystem for renewable energy and the agricultural production subsystem in the plant-based bioeconomy. 1.1.3 Forest-based bioeconomy in Germany The forest-based bioeconomy, encompassing the entire forest value chain, is considered to be a key player in the arena of promoting the bioeconomy for achieving decarbonization of the economy. In Germany, it has exhibited great potential for climate change mitigation by reducing carbon emissions (Hagemann et al., 2016; Purkus et al., 2018). Up to 2016, 570 policy documents linked to carbon-mitigation strategies, covering the whole value chain of the forest-based bioeconomy, have been launched at the EU level (Rivera León et al, 2016). It is argued that the forest-based bioeconomy can play both direct and indirect roles (e.g., carbon sequestration by the forest and soil, and substitution effects from bioenergy replacing fossil fuel, respectively) in carbon emission reduction (Seppälä et al., 2019; Jonsson et al., 2020; Kumeh et al., 2021). Forests, a key resource input and support system for the forest-based bioeconomy, are a main source of carbon sinks. It has been reported that forests and harvested wood products together sequester the equivalent of circa 10% of the EU’s greenhouse gas emissions (EU, 2022). Apart from forests and traditional wood products, the forest-based bioeconomy also covers efforts directed towards bioenergy, biochemicals, 6 textiles, cellulose and lignocellulosic bioplastics, packaging products, etc. (Wolfslehner et al., 2016). It can also contribute to climate change mitigation by promoting the use of wood to substitute fossil fuels and other materials (EU, 2022). However, this substitution process occurs at the sacrifice of more forest biomass. It is evident that without sufficient afforestation and forest resources, the mitigation strategies focusing on enhanced carbon storage in wood products and the substitution of fossil fuels and energy-intensive materials require biomass removals; thereby, in most cases, decreasing the carbon sequestration potential of the forests (Lindner et al., 2017). Thus, the real mitigation effects of the forest-based bioeconomy on carbon emissions require a better understanding. 1.1.4 Bioclusters in Germany Bioclusters are a special kind of clusters that operate with the explicit goal of promoting sustainable development by fostering the transition to a bioeconomy (Hermans, 2018). Bioclusters can play an important role in the sustainable transition to bioeconomy, especially in Germany. At the same time, bioclusters are characterized by coupled production systems, leading to stronger horizontal and vertical implications for industrial integration (Wesseler and von Braun, 2017). The circular production mode is prevalent in bioclusters carries a high expectation for linking innovation with climate neutrality (Biber‐ Freudenberger et al., 2020). This can help to tranform the prevalent linear production mode to a no-linear production mode for the whole of society as well. Furthermore, close cooperation among biotech companies, research institutes, technology parks, etc., can create sufficient scientific outputs and innovations to support the emerging bioeconomy. Especially for Germany, bioclusters, characterized by the heavy concentration of stakeholders and organizations, are proving to be important new technology impulse givers in this respect (Dorocki, 2014). The thriving bioclusters in Germany, with more than 770 biotechnology companies involved in 2021, have created substantial scientific outputs, offering opportunities to promote green efficiency and productivity at large (FMEACA, 2022). 13 As biomass sources mainly originate from agriculture and forest, the mitigation effects of forest and agriculture sectors on carbon emissions are distinguished. Research question: Will the transition of a forest-based bioeconomy reduce carbon emissions”? Given that there is currently limited understanding of the impact of the forest-based bioeconomy on carbon emissions in Germany, this dissertation aims to estimate the spatial impact of the forest-based bioeconomy, especially the role of technological innovation in the forest-based bioeconomy on carbon emissions. Using an extended Spatial Durbin Model and 401 NUTS-3 level panel data from 2000 to 2021, this dissertation measures the intra-regional and spillover effects of technological innovation, the size of the bioeconomy, industrial upgrading and their interactions on carbon emissions empirically. Research Objective 3 (Chapter IV): The third objective of this dissertation is to measure the impact of bioclusters on sustainable performance where carbon emissions are considered as an undesired output. Research question: Will the establishment of bioclusters promote green productivity”? Given the research gap about the mechanisms of institutional innovation’s effects on green productivity, this dissertation, focusing on Germany at the NUTS-3 level, aims to estimate the causal effects of bioclusters on green productivity mediated by technological innovation. The dissertation uses a quasinatural experiment, including a series of methods, like difference in differences (DiD), Staggered DiD(SDiD), PSM-SDiD, and difference in difference in differences (DDD), and a mediating model to estimate the impact of establishing bioclusters on green total factor productivity (GTFP) in Germany. 1.4 Structure of the dissertation This dissertation examines the implications of developing the bioeconomy for promoting sustainable development in Germany, analysing it from two perspectives and three influencing mechanisms. One perspective is obtained from an impact evaluation. Using NUTS-3 panel data from 2000 to 2021, this dissertation examines the impacts a bioeconomy in Germany would have on carbon emission reduction and 14 green productivity empirically, providing empirical evidence regarding the efficiency of the bioeconomy strategy for policymakers and contributing to the relevant body of literature. The other perspective is obtained from a scenario simulation process. This dissertation anticipates the potential mitigation effect of technological innovation on agricultural carbon emissions as well as the dynamic interactions among subsystems from 2020 to 2050, with the simulation period chosen to fit in Germany’s goal of attaining carbon neutrality by 2050. Three influencing mechanisms are proposed covering structural changes in the production factors, namely changes in land use and labour, and industrial structure; technological innovation and its spatial diffusion; and institutional incentives in the bioeconomy. The remainder of the dissertation is organized as follows. Chapter Ⅱ and Ⅲ discuss the potential mitigation effects of agriculture and forestry under the transition to bioeconomy on carbon emissions, respectively. Chapter Ⅳ analyses the influences of institutional innovations in the bioeconomy on the green total factor productivity, while the final chapter presents the conclusion part (Chapter Ⅴ) with some policy suggestions and methodological implications. The structure of the dissertation is illustrated in Figure 1.2. Biomass supply Technological innovation Institutional innovation Bioeconomy Case of Germany Chapter Ⅰ Introduction Policy implications Methodological implications Implications Germany and other countries Chapter Ⅴ Synthesis Chapter Ⅱ Mitigation effect of agriculture on carbon emissions System dynamic modeling; Effects of R&D investments and dynamic interactions Chapter Ⅲ Mitigation effect of forest-based bioeconomy Intraregional and spillover effects of technological innovation; Spatial Durbin Model Chapter Ⅳ Implication of bioclusters on green productivity Causal effects of bioclusters on green productivity; Quasi-natural experiment; Mediating model Sector NUTS-3 Figure 1.2: Structure of the dissertation Source: Own operations. 15 2 Potential mitigation effects of technological innovation in the plant-based bioeconomy on carbon emissions 1 2.1 Objectives and theoretical framework 2.1.1 Background and organization Agriculture is the cornerstone of the European Union’s bioeconomy due to its role as a primary biomass supplier and as an important sector with great potential for carbon emission reduction. Agriculture can both generate carbon emissions through farming activities and industrial processes and at the same time act as a carbon sink via plant photosynthesis and land use changes (Pataki et al., 2006; Wang et al., 2012; Gutzler et al., 2015). This dual role complicates the measurement of agriculture’s actual impact on carbon emissions. The introduction of technological innovation, which is achieved through R&D investment and is an important pillar in the bioeconomy (Schütte, 2018), further adds to the difficulty. This is because such innovation can reduce carbon emissions by improving resource efficiency and enhancing carbon sequestration, but it can also increase emissions due to the rebound effect. So far, there is little knowledge about whether and how R&D investments can mitigate agricultural carbon emissions with the transition to the bioeconomy. Many studies suggest that technological innovation has the potential to reduce carbon emissions by improving agricultural productivity, reducing the need for cultivated land, enhancing innovation efficiency, and promoting a circular economy (Xiong et al., 2016; Frank et al., 2019; Balsalobre-Lorente et al., 2019; Nwakae et al., 2020). R&D investment, often used as an indicator of technological innovation, has the potential to reduce carbon emissions by advancing biotechnological innovations, improving the production efficiency of biorefineries, and promoting 1 Author statement: Lanjiao Wen (conceptualization, methodology, software, writing-original draft, and revision); Dr. Zhanli Sun (conceptualization, revision and supervision); Dr. Lioudmila Chatalova (conceptualization and revision); Prof. Dr. Anlu Zhang (conceptualization); Prof. Dr. Alfons Balmann (revision and supervision). 16 bioenergy (e.g., biogas). However, some agricultural literature argues that the increase in carbon emissions caused by technological innovation may exceed the reductions they offer. This is because the application of certain technologies, such as biotechnology and breeding, requires more biomass and energy to support the transition to a bioeconomy (Henle et al., 2008; Fleiter et al., 2012; Iris and Lam, 2019). The growing demand for biomass can also lead to land use conflicts, biodiversity loss, and an overuse of chemicals and energy (Deininger, 2013; Liobikiene et al., 2020). With the transition to the bioeconomy, the agricultural carbon emission system becomes more complex, comprising the land use subsystem, agricultural production subsystem, innovation subsystem, coupled subsystem (including resource recycling and upgrading use), and the socioeconomic subsystem. Specifically, in the plant-based bioeconomy, the productivity of technological investments can be affected by many factors, like the input-output relationship, industrial integration, biomass recycling/upcycling use, and innovation efficiency. This not only makes the agricultural carbon system more complex and challenging to quantify, but also highlights the roles of the innovation subsystem and coupled production subsystem, as well as the dynamic interactions among subsystems in carbon emission reduction. Furthermore, the decreasing R&D productivity has been reported in Germany (Schäfer, 2014; Ugur et al., 2016), which makes the role of technological innovation in reducing carbon emissions within the agricultural system even less clear. Therefore, understanding how R&D investments impact carbon emissions through the dynamic interactions among subsystems in the agricultural system is a major question that needs to be answered. The present study aims to contribute to projecting agricultural emissions by detailing the impact of R&D investments on carbon emissions. Focusing on the agricultural sector in Germany, this study applies a system dynamics modelling approach to simulate the potential impact of R&D investments on carbon emissions through considering the dynamic interactions among the agricultural carbon subsystems. To present the net effect of R&D investments on carbon emissions, this study takes carbon sinks and carbon 17 emission reduction into account, including, e.g., carbon sequestration from land use and plant production, and carbon reduction from re-/upcycling of agricultural residues and by-products. The rest of this chapter is organized as follows. Section 2.1.2 outlines the relationship among carbon emissions, carbon sinks, and carbon emission reduction in the agricultural system. Section 2.2 introduces the data and methodology used for simulating net carbon emissions and the dynamic interactions in the agricultural system. Section 2.3 describes the design of scenarios for the simulation. The results and policy implications are summarized and discussed in section 2.4, while the last section 2.5 concludes the analysis. 2.1.2 Theoretical framework Figure 2.1 illustrates an overview of the subsystems considered under the nexus between agricultural sustainability and carbon emissions. Crop production and animal husbandry are the main agricultural activities and carbon sources as well. Their production involves input and output flows and is associated with carbon emissions from both the farm processes and livestock, while carbon sequestration saved by the green landscape and the recycling and reuse of biomass can reduce the emissions to some degree, forming an agricultural carbon cycle. Since the agricultural carbon cycle is associated with the production process, economic environment, land use cover change, technological level, and producing structure (Lu and Guldmann, 2012; Gu et al., 2019), five subsystems in the plant-based bioeconomy are defined and modelled in the present study. The five subsystems, namely the land use subsystem, agricultural production subsystem, innovation subsystem, coupled production subsystem (including resource recycling and upgrading use), and the socioeconomic subsystem (see Figure 2.1), closely interact with each other. 18 Agricultural carbon emission system Carbon sink Green landscape (photosynthesis, carbon sequestration) Crop production Animal husbandry Carbon emissions Plant farming (irrigation, fertilizer, pesticides, ploughing…… ) Carbon cycle Input factors Livestock (animal manure…… ) Recycling use (agricultural residues, biogas production) Outputs Subsystems in the plant-based bioeconomy Socioeconomic subsystem Agricultural production subsystem Land use subsystem Coupled production subsystem Innovation subsystem GDP; social investment; R&D investment…… Farmland; grassland; Fallow land; EFA…… R&D investment; biotechnology; bioclusters…… Biorefinery; agricultural residues; animal manure…… Farmland & grassland; fertilizer; biotechnology …… Figure 2.1: Overview of the subsystems Source: Own operations. R&D investment can significantly improve technological innovation and industrial integration (e.g., vertical integration) (Wesseler et al., 2015; Wesseler and von Braun, 2017). Thus, the innovation subsystem is crucial in the plant-based bioeconomy. Along with industrial integration, value chain integration through cascading or circular resource utilization can facilitate agricultural production by switching from a traditional linear mode to a non-linear mode, making the interactions between the agricultural production subsystem and coupled production subsystem more relevant to each other. As R&D investment can promote a cascading use efficiency for biomass, e.g. by promoting its use for biogas production, while new patents and biorefineries can contribute to secondary GDP (GDP-2) (Sorda et al., 2013; Grando et al., 2017), they 19 closely link the circular economy with R&D investment in the plant-based bioeconomy (Theuerl et al., 2019; Kardung et al., 2021). This implies that the innovation subsystem closely interacts with the coupled production subsystem and socioeconomic subsystem. In addition to the innovation subsystem and coupled production subsystem, the agricultural production subsystem, socioeconomic subsystem and land use subsystem are the basic components in the agricultural carbon system. Reverting agricultural land use with natural/perennial vegetation (for instance, fallow land) is regarded as one of the most efficient ways to accumulate organic carbon in soil (Post and Kwon, 2000; Schulp et al., 2008). Thus, sustainable farmland policies, such as the Common Agricultural Policy (CAP) and greening reform of the CAP, have been adopted in Germany to improve the biodiversity of farmland. Especially, the greening reform of the CAP introduced ecological focus areas (EFAs). Even though EFAs have not yet been evidenced to have a positive effect on improving biodiversity as researchers expectated (Pe'Er et al., 2017), they have proven to be effective for carbon sequestration (Ottoy et al., 2018). Therefore, the EFA is considered in the agricultural carbon emission system. In this study, the land use subsystem mainly includes farmland (cropland and grassland) and green land. Green land is the sum of the fallow land, ecological focus area (EFA), and grassland, which is associated with a net carbon sink. As arable land is associated with the socioeconomic and agricultural production subsystems, the land use subsystem directly interacts with the socioeconomic subsystem and agricultural production subsystem. The agricultural production subsystem covers the planting and livestock, and also their associated carbon emissions. Ploughing and irrigation, which are positively related with the arable land area, together with the capital inputs involved in planting (fertilizer, pesticides, and diesel) are positively related to carbon emissions. Agricultural production subsystem interacts with all the other subsystems directly. This is because agricultural production provides biomass and agricultural residuals for the coupled production subsystem, and R&D investment overall (R&Dvest) and R&D investment specifically in agriculture (AR&D) from the innovation subsystem are beneficial for improving agricultural productivity. Additionally, the basic input factors, such as capital (agricultural investment) and labour, are 20 associated with the socioeconomic subsystem, while the agricultural land is part of the land use subsystem. The innovation subsystem is mainly represented by R&D investment overall (R&Dvest), R&D investment specifically in agriculture (AR&D), R&D staff, and patents. The coupled production subsystem includes bioenergy production and other secondary industries that use biorefineries for production. As it depends on the biomass supply and technological investments, this subsystem directly interacts with the innovation subsystem and agricultural production subsystem. The socioeconomic subsystem is denoted by GDP, secondary GDP (GDP-2), tertiary GDP (GDP-3), labour, population, social investment (Invest) and agricultural investment (Ainvest). It directly interacts with the agricultural production subsystem, land use subsystem and innovation subsystem. This subsystem may not produce agricultural carbon emissions directly, but the activities associated with other subsystems can generate carbon emissions and sequester carbon at the same time. The effects of the dynamic interactions among R&D investments, land use change and innovation efficiency on agricultural carbon emissions are studied in four scenarios in addition to the base scenario, namely (1) land effect, (2) structure effect, (3) technological effect and (4) their combined effect. All of them are simulated for the period from 2020 to 2050 (see Figure 2.2). 21 Sustainability-guided bioeconomic strategies Net carbon emissions simulation SD scenario schemes S1-Land effect: sustainable land use management S2-Structural effect: circular economy S3-Technological effect: technological investments Historical trajectory of carbon emissions Model validation Subsystems Modify Agricultural production subsystem Socioeconomic subsystem Coupled production subsystem Innovation subsystem land use subsystem Carbon emissions projection from 2020 to 2050 S4-Combined effect Figure 2.2: Simulation progress for net carbon emissions Source: Own operations. 2.2 Methodology With a system dynamics (SD)approach, the net carbon emissions from 2020 to 2050 and the dynamic interactions in the system are simulated. The net carbon emissions over these three decades are projected as Germany aims to attain carbon neutrality by 2045. 22 2.2.1 Data source and assumptions The carbon emissions data used in this study were calculated on the basis of land use data and emission parameters for different land use types and different agricultural activities at the federal level in Germany. The historical period was from 2000 to 2019 since the concept of knowledge-based bioeconomy originated in the 2000s (Patermann and Aguilar, 2018). The emission parameters were derived from the Intergovernmental Panel on Climate Change (IPCC) (2019) and related studies. The land use data from 2000 to 2019 were collected from the statistics of the German Federal Ministry of Food and Agriculture (BMEL), Federal Office of Statistics and Thünen-Institut für Ländliche Räume. The other secondary data, such as socioeconomic data, for 2000–2019 were collected from the European Statistical Office (Eurostat), Federal Office of Statistics, and BMEL. This study assumes that R&D investments have a mitigation effect on agricultural carbon emissions mainly through internal interactions with regard to the land use subsystem and innovation subsystem. In addition, R&D investments are assumed to be positive with social investment and GDP. Besides, agricultural R&D investments are assumed to be positive with R&D investments. 2.2.2 Structure of the SD model The SD model developed by Jay W. Forrester is a decision-making tool that has been widely used to simulate the complicated behaviour and feedback of real systems (Forrester, 1970). It involves the use of stocks, flows and feedback loops to represent the interdependencies within a system. As an advanced simulation tool, it provides enhanced capabilities for visualization, scenario analysis, and user interactivity, supporting multi-method modelling and combining system dynamics with agent-based modelling. Owing to the complexity discussed previously, the SD model is employed in this study for simulating carbon emissions over the period from 2020 to 2050. The reason for choosing this model is that it offers advantages for integrated and quantitative simulation in the short and medium term (Fong et al., 2009; Fu et al., 2015; Gu et al., 2019). 29 In scenario 2, the structural effect is reflected by the development of circular economy. Due to the significant role of biomass in the circular economy and the target to minimize agricultural residues in agriculture (Sherwood, 2020; Sharma et al., 2021), the development of the circular economy is represented by the increasing supply of biomass and increasing share of agricultural residuals sent to the biorefinery. In the first subcase (Structure 2-1), all the plants are assumed to have faster increase rates than that in the base scenario. Alao maize, which usually generates a great number of agricultural residues, is set to have the highest growth rate (3%) from 2020 to 2050 among the plants for biogas production. In the second subcase (Structure 2-2), to highlight the role of biorefineries as raw materials for the industry sector, it is assumed that the share of agricultural residuals for biogas would reduce from 0.4 in 2020 to 0.2 in 2050. Furthermore, in Structure 2-3, both cases are considered. Direct R&D investment can not only improve production efficiency in the agricultural production subsystem, but also promote the scale of the circular economy by elevating the level of residues pretreatment and the resource use intensity for biorefineries (Amidon et al., 2011; Tayeh et al., 2020). To display the role of R&D investment in the agricultural production subsystem and coupled production subsystem, R&D investment in agriculture (AR&D) and the general R&D investment (R&D) are taken into account in scenario 3. In Tech 3-1, the share of agricultural R&D (AR&DR) is set to increase to 6% in 2020. At the same time, in Tech 3-2, the share of R&D investment in GDP (R&DR) is assumed would increase to 0.06 in 2020. Combining Tech 3-1 and Tech 3-2, AR&DR and R&DR are set to increase in Tech 3-3. The scheme in the Combine 4 scenario includes the Land scenario, Structure 2-3, and Tech 3-3. 30 Table 2.3: Scenario schemes for different scenarios Scenario Schemes Base scenario Base: Ratios are set as same as that in 2019 Scenario 1-Land effect Land: Increasing the ratio of fallow land to 0.05 in 2020 Scenario 2-Structural effect Structure 2-1: Increasing the supply of agricultural biomass; Structure 2-2: Increasing the share of agricultural wastes for biorefineries (decreasing the share of waste for biogas from 0.4 in 2020 to 0.2 in 2050); Structure 2-3: Increasing both agricultural biomass and the share of agricultural wastes for biorefineries Scenario 3Technological effect Tech 3-1: Increasing the share of agricultural R&D in R&D investment (AR&DR) to 0.05 in 2020; Tech 3-2: Increasing the share of R&D investment in GDP (R&DR) to 0.06 in 2020; Tech 3-3: Increasing both AR&DR and R&DR Scenario 4Combined effect Combine: Increasing the increment ratio of fallow land to 0.05 in 2050, increasing both agricultural biomass and the share of biorefinery, and increasing both AR&DR and R&DR 2.4 Results and analysis 2.4.1 Historical tendency of net carbon emissions Figure 2.5 shows the historical tendencies of the Net carbon emissions (NCE), carbon emissions (CE), carbon sinks (CS), and carbon emissions reduction (CR) in German agriculture during the period 2000 to 2019 varying by years. Despite the agricultural carbon emissions fluctuating from 10.5 million tons to 9 million tons from 2000 to 2019, an overall downward trend could be observed. This is likely because of the reducing use of fertilizer and ploughing and the decreasing number of cattle and sheep due to technological 31 improvements in the production for farming and husbandry (Jantke et al., 2020). Furthermore, the agricultural carbon sink amount decreased slightly from 1.25 million tons in 2000 to 1.07 million tons in 2013. After 2014, this figure increased gradually; especially after 2015, whereby it increased rapidly to 1.36 tons in 2019. Biogas had a minor replacement effect on carbon emission reduction at the beginning of the period (2000-2005) when the biogas electricity production was relatively low (445 Mio. kWh in 2000 and 1696 Mio. kWh in 2005), but such production grew robustly after 2006, reaching 29,245 Mio. kWh in 2017. This contributed to a rapid growth in carbon emissions reduction, increasing from 0.28 million tons in 2006 to 2.41 million tons in 2019 (see Figure 2.5). Net carbon emissions (NCE) in German agriculture kept decreasing in the period 2000 to 2019. This tendency was similar to the tendency for carbon emissions from 2000 to 2006, as the carbon sink and carbon emissions reduction remained nearly constant during this period. Since 2006, NCE has declined gradually but in the opposite direction with that of carbon emissions due to the rapid increase in carbon emission reduction. Due to the policy reforms promoting biogas, such as guaranteed feed-in tariffs, biorefineries are encouraged to produce biogas. During the period from 2006 to 2019, the net carbon emissions decreased by almost a third, from 7.79 million tons in 2006 to 5.34 million tons in 2019. 32 Figure 2.5: Net agricultural carbon emissions, carbon emissions, carbon sink and carbon emissions reduction from 2000 to 2019 2.4.2 Scenario analysis Using the SD model, the net carbon emissions in the agricultural system are simulated under different scenarios for the period 2020 to 2050 (see Figure 2.6). According to Figure 2.6, net carbon emissions are projected to decrease rapidly during the period from 2020 to 2050. Net carbon emissions under the Structure 2-1 (increase agricultural biomass), Structure 2-3 (increase both agricultural biomass and biorefinery), and Combine (combined effect) scenarios will remain positive during 2020 to 2050. The results from Structure 2-1, Structure 2-3, and Combine imply that increasing biomass production may support the circular economy, but the increased carbon emissions during the production process cannot be offset by the reduced carbon emissions through R&D investment alone. Compared with the Base scenario, the net carbon 33 emissions under the Land (increase the ratio of fallow land), Structure 2-2 (increase the share of biorefineries), Tech 3-2 (increase the share of R&D investment), and Tech 3-3 (increase both agricultural R&D investment and R&D investment) scenarios are smaller. The result of net carbon emissions under Tech 3-2 is the lowest (-2.82 Mt in 2050). This highlights the role of R&D investment in mitigating carbon emissions directly. While the results for Land and Structure 2-2 not only indicate that increasing the amount of fallow land and developing the circular economy can reduce carbon emissions, but also imply that R&D investment can indirectly mitigate carbon emissions through improving the production efficiency of biorefineries and by increasing the amount of green land. The agricultural carbon emissions calculated in this study correspond to nearly one sixth of carbon dioxide equivalents reported by the Thünen institute (61.8 Mt in 2019, cf. Rösemann et al., 2021). This big difference may result from two aspects: Once, as mentioned previously, the calculation by Rösemann et al. (2021) includes all the GHG emissions from German agriculture, with all kinds of agricultural activities and animals of livestock taken into consideration, whereas fewer production management activities (e.g., fertilizer, ploughing and irrigation) and only three kinds of animals (caw, sheep, and pig) are considered in this study; Second, the carbon emissions calculated by Rösemann et al. is based on carbon dioxide equivalents, while our results are based on carbon equivalents. 34 Figure 2.6: Net agricultural carbon emissions under different scenarios from 2000 to 2050 The agricultural carbon emissions and their causal tree during 2000 to 2050 are shown in Figure 2.7. The obtained results show that agricultural carbon emissions will gradually decrease from 2020 to 2050. The similar tendencies for CE-1 (carbon emissions from plant farming) and carbon emissions in the causal tree (in the right of Figure 2.7) indicate that agricultural carbon emissions are mainly caused by farming activities. This adds to the body of the relevant literature by uncovering the mechanism of the technological effects on emissions reduction. Specifically, it shows that R&D and AR&D investments can significantly lower agricultural carbon emissions. The amount of carbon emissions under Structure 2-1 is the lowest (decreasing to 4.33 Mt in 2050), and the figures under Combine (4.44 Mt in 2050) and Structure 2-3 (4.78 Mt in 2050) are less than that under the Base case (5.12 Mt in 2050), illustrating their greater impacts on carbon emissions from the improved circular economy with the increasing biomass supply. 35 Figure 2.7: Carbon emissions and their causal trees under different simulation scenarios from 2000 to 2050 Carbon sinks, as stimulated in Structure 2-2, scenario 3 (Tech 3-1, 3-2, and 3-3), and in the Base case, are predicted to continuously increase from 2020 to 2050 (see Figure 2.8). While carbon sinks under scenario 1 (Land) and 4 (Combine) first grow rapidly after 2020, they are then forecast to show a sharp decline in 2033, before gradually increasing again from 2034 to 2050. This sharp change is mainly related to the carbon sinks by green land (CS-1), according to the causal tree. Due to the maximum restriction of the EFA, the simulation equation for the EFA changes after 2033 (see the equations in Table A.1). The higher values in the Land scenario (6.83 Mt in 2050) suggest that sustainable land use management and increasing biomass supply will aid the amount of carbon sinks available. 36 Figure 2.8: Carbon sinks and their causal trees under different simulation scenarios from 2000 to 2050 The observations from the projected carbon emissions reduction from biogas support the findings that biogas production has the potential to improve carbon emission reduction if R&D investment leads to cleaner production (Meyer et al., 2012; Ersoy and Ugurlu, 2020). The projected carbon reduction decreases gradually after 2020. This arises from the replacement effect of bioenergy on the carbon emission reduction. Comparison among the different scenarios indicates that the combined effect has the greatest impact on carbon emissions reduction (1.87 Mt in 2050). The values under Tech 3-2 and Tech 3-3 show stable decrease (1.3 Mt in 2050). This implies that R&D investment can promote the carbon emission reduction at large (see Figure 2.9). 37 Figure 2.9: Carbon emissions reduction under different simulation scenarios from 2000 to 2050 2.4.3 Sensitivity analysis The sensitivity of R&D investment’s impact on net carbon emissions is shown in Figure 2.10. In order to test the sensitivity, the change rate for the share of R&D in GDP (R&D-to-GDP ratio) is set to increase and decrease by 25%, 15%, 10%, and 5%. The change rates of net carbon emissions represent the percentage difference between net carbon emissions at varying R&D-to-GDP ratios and net carbon emissions under the base scenario. Figure 2.10 shows that most of the net carbon emissions are similar when the change rate of R&D-to-GDP ratio is changed from -25% to 25%. This implies our results are relatively reliable. The negative relationship between the change rates of net carbon emissions and the change rates of the R&Dto-GDP ratio illustrates that net carbon emissions are sensitive to R&D investment. Regarding the tendency of the change rates of net carbon emissions, the tendency showed a decrease from 2000 to 2011 and then an increase afterwards. Among simulated rates, the change rate is the largest when the change rate of R&Dto-GDP ratio increased to 25%, reaching 290% in 2011 and returning to 8.7% in 2019. This may be because of the model restriction to carbon emissions reduction. 38 Figure 2.10: Result of the sensitivity analysis 2.5 Discussion and conclusions 2.5.1 Discussion The above analysis shows that R&D investment can contribute to a reduction in agricultural carbon emissions during 2020 to 2050, helping realize carbon neutrality at the sector level ahead of 2045. The findings regarding the decline in carbon emissions simulated by the SD model under different scenarios add to the body of relevant literature by projecting the impact of R&D investments on carbon emissions and uncovering the dynamic interactions among the various subsystems in the plant-based bioeconomy. Methodologically, the SD model is a cutting-edge approach to understanding and managing the complexities of agricultural practices and their impact on carbon emissions. Unlike general projection models, such as time series regression (e.g. ARIMA) and back-propagation networks, which require strict 45 emissions in the regional eco-economic system from 2000 to 2021 in Germany? 2) How does the diffusion of technological innovation in the forest-based bioeconomy affect carbon emission at the county level? 3) How does the spatial spillover effect of technological innovation in the forest-based bioeconomy determine the emissions reduction potential and direction? By addressing these points, the present study aims to contribute to the currently limited knowledge about the impact of the forest-based bioeconomy on carbon emissions in Germany. The remainder of this chapter is organized as follows. Section 3.1.2 outlines the relationship between carbon emissions and the forest-based bioeconomy in Germany. Section 3.2 introduces the methods and data sources. The results are summarized in Section 3.3. The last section (3.4) concludes this part of the work. 3.1.2 Conceptual framework: Carbon emissions and the forest-based bioeconomy in Germany Human activities rely on land-related ecosystem services, necessarily affecting the ecosystem’s carrying capacity. Anthropogenic impacts–from unavoidable changes in land cover for creating living and production space to avoidable environmental harms–are associated with carbon emissions (Pataki et al., 2006). Land use change driven by socioeconomic dynamics such as urbanization and industrialization is one of the largest contributors to carbon emissions today. At the same time, it directly affects the ecosystem’s capacity to sequester carbon in soils, the forests, and geological formations (Bockstael et al., 1995; Pataki et al., 2006). The carbon cycle is further affected by physical processes in the lithosphere, which are, however, largely outside of human control. Figure 3.1 illustrates carbon emissions production and regulation in a stylized eco-economic system, representing the economic activities that may occur in a regional eco-economic system. In line with the European Commission, Germany has an ambition to develop a bioeconomy that depends largely on the forest-based sector (Giurca and Späth, 2017). The forest-based bioeconomy in Germany not only produces traditional wood products, such as woodwork, pulp and paper, and wood for bioenergy (Jochem et al., 2015), but also aims to maximize value increment in the whole value chain, which should result in high-value products and offering more job opportunities. Technological innovation is regarded as 46 a key pillar for the forest-based bioeconomy. Both policy makers and scholars acknowledge that slow technological development can hinder the development of the forest-based bioeconomy (BMBF, 2011) and thus confine many relevant technological developments to the laboratory and pilot scale (Hagemann et al., 2016). According to the theory of endogenous growth, knowledge spillover through learning by doing has a significant positive effect on economic growth (Arrow, 1962). As an endogenous input factor, technological innovation in the forest-based bioeconomy together with other production factors, such as labour, land, and capital, can promote new resource allocation efficiency and economic growth. This will increase the productivity and competitiveness of local industries in the value chain of the bioeconomy. The knowledge spillover effect driven by technological innovation among the value chain will also stimulate industrial integration and labour division through optimizing the factor substitution efficiency, facilitating bioclusters and spatial industrial patterns to develop. Due to the spatial diffusion of innovations, a higher degree of technological innovation can also improve the competitiveness of adjacent areas (Vaitsos, 1978). Industrial clusters can accelerate industrial upgrading by increasing the competitiveness of the involved industries and their capacities for value-added generation, economic diversification, and employment creation. From this perspective, intra-cluster competitiveness not only contributes to economic growth, but also leads to the reduction of carbon emissions (Gautam, 2014; Cui et al., 2021). While adjacent counties may provide abundant urban land for industrial growth (Guastella et al., 2017; Gao et al., 2020), they have the potential to offer extra job opportunities too, thus attracting labour to move to the areas. 47 Ecological economic system Biological processes and human activities Physical processes Ecological subsystem (photosynthesis of forests, carbon sequestration, etc.) Economic subsystem (production &consumption) Lithosphere (volcanic eruption, carbon precipitation, climate change, etc.) Forest-based bioeocnomy Carbon cycle Carbon emissions Carbon sink Labor mobility Value added Effects of spatial agglomeration on carbon emissions Industries spatial partterns Labor division Industrial restructuring Technological innovation Industrial competition Technology diffusion Industrial restructuring Labor mobility regions adjacent regionslocal Technological innovation Figure 3.1: Carbon cycle in a regionally integrated land use system Source: Own representation. 3.2 Methodology In this research, the analysis entails three steps: First, the net carbon emissions are calculated for the period from 2000 to 2021. Next, the size of the bioeconomy at each county/city is quantified to assess the regional development of bioeconomy. Then, a Spatial Durbin Model is employed to estimate the impacts of the forest-based bioeconomy on net carbon emissions. 48 3.2.1 Estimations of the net carbon emissions at the county-level Net carbon emissions are calculated as the sum of carbon emissions and carbon sinks associated with the use of arable land and construction land, which have the highest energy consumption (Zhao et al., 2015; Zhang et al., 2015). In the present study, carbon emissions for construction land are calculated indirectly as the product of the energy consumption per unit of GDP (𝑇𝑖) and GDP of the secondary and tertiary industries (𝑀𝑖) in county i (Wen et al., 2021). Although crops produced on arable land can, to some extent, absorb carbon emissions, the use of fertilizers, agricultural machinery, and irrigation systems generate high net emissions (Yang et al., 2016). Therefore, carbon emissions (ECi) for each county/city are defined as: 𝐸𝐶𝑖=𝜂𝑎∙𝐶𝑖+𝑇𝑖∙𝑀𝑖 (3-1) where 𝜂𝑎 is the carbon emissions parameter for arable land and Cii is the arable land size. The carbon sink (ESi) of county i comprises the carbon sequestration in forests, grassland soils, and water areas, calculated as the product of the carbon emissions parameter 𝛿𝑗 and land size Sij for each land use type j: 𝐸𝑆𝑖=∑𝛿𝑗∙𝑆𝑖𝑗 (3-2) where j=1,2,3 and represent forest, grassland and water, respectively. The net carbon emission (NECi) for county i is then: 𝑁𝐸𝐶𝑖=𝐸𝐶𝑖−𝐸𝑆𝑖 (3-3) 3.2.2 Measuring the size of the bioeconomy The size of the bioeconomy varies depending on the definitions and approaches taken. As the majority of studies regarding bioeconomy are qualitative conceptual papers with different definitions, the measurements of the size of the bioeconomy differ accordingly. The diverse definitions and lack of harmonized approaches for comparison are major challenges for quantitative analysis of the contribution of the bioeconomy towards sustainability. Kuosmanen et al. (2020) concluded that there are basically four 49 types of approaches to measure the size of the bioeconomy, namely the output-based approach by novaJRC, Finnish bioeconomy statistics, the physical supply and use approaches developed by JRC, Statistics Netherlands–CBS, and the Thünen Institute’s methodology. Among these, Thünen’s approach not only offers the advantage in highlighting the role of resource-based materials flows in the process of production, which is consistent with the definition provided by the German bioeconomy strategy (BMEL, 2014), but also has the advantage of reflecting the direct socioeconomic contribution of the bioeconomy as it focuses on the sectoral level. For this, this study adopts the Thünen Institute’s approach to measure the size of the bioeconomy in Germany. Since “value added” has been proven to be preferable to that of gross output to avoid repeated calculations (Kuosmanen et al., 2020), this study measures the size of the bioeconomy by employing the Thünen Institute’s approach as reported by Iost et al. (2019) and considering the indicators gross value added and employment. According to Iost et al. (2019), the agricultural sector, including agriculture, forestry, and fishing, is considered to be 100% bio-based. For the manufacturing sector, the bio-based share used in this study is the average (𝑏𝑚     ) of the different bio-based shares of the sub-manufacturing sectors (at the 4-digit level), including food and feed, textile, leather, wood and wood products, paper and paperboard, printing, chemicals, pharmacy, plastics, furniture, and others. As it is difficult to bring the service sector data in line with the NACE sectors, the average bio-based share (𝑏𝑜    ) of other experimental developments based on natural science and engineering is used as the bio-based share for the service sector (the bio-based shares of relevant NACE sectors at the 4-digit level are shown in Table A.3). Two dimensionalities for the size of bioeconomy in Germany are calculated as below: 𝐵𝑉𝑖=𝑉𝐴𝑖+𝑉𝑀𝑖∗𝑏𝑚     +𝑉𝑂𝑖∗𝑏𝑜    (3-4) 𝐵𝐸𝑖=𝐸𝐴𝑖+𝐸𝑀𝑖∗𝑏𝑚     +𝐸𝑂𝑖∗𝑏𝑜    (3-5) where 𝐵𝑉𝑖 and 𝐵𝐸𝑖 are the value added of bioeconomy and the number of employees in the bioeconomy for county i respectively; VA, VM and VO denote the value added for the agriculture sector, manufacturing 50 sector, and service sector respectively; EA, EM and EO presents the number of employees in the agriculture sector, manufacturing sector, and service sector respectively. 3.2.3 The Spatial Durbin Model A Spatial Durbin Model (SDM) is developed to estimate the impact of a forest-based bioeconomy on carbon emissions. Prior to modelling, the global spatial autocorrelation index (Moran’s I) is calculated to test for spatial autocorrelation and spatial heterogeneity (Odland,1988; Geniaux and Martinetti, 2018; Feng and Chen, 2018): 𝑀𝑜𝑟𝑎𝑛′𝑠 𝐼 =∑ ∑ 𝑊𝑖𝑘 𝑛 𝑘=1 (𝑁𝐶𝐸𝑖 𝑛 𝑖=1 −𝑁𝐶𝐸       )(𝑁𝐶𝐸𝑘−𝑁𝐶𝐸       )/𝑉2∑ ∑ 𝑊𝑖𝑘 𝑛 𝑘 𝑛 𝑖=1 (3-6) with the mean (𝑁𝐶𝐸 −𝑁𝐶𝐸       ), variance of net carbon emissions (𝑉) and spatial weight matrix (Wik). The values of Moran’s I index are within the range of [-1, 1], indicating either positive or negative spatial correlation among counties (Bai et al., 2012; Anselin, 2013; Gao et al., 2020). If the value is zero, then the counties are not spatially correlated. Numerous studies suggest that carbon emissions, being affected by the natural environment and human activities, have regional spillover effects (Jun et al., 2017; Wang et al., 2018; Wang et al., 2019). The range of this effect, however, varies depending on the model in use. The advantage of the SDM is that–other than the spatial lag model (SLM) and the spatial error model (SEM)–it can capture the spatial correlation of dependent variables and the spatial spillover effects of independent variables (LeSage and Pace, 2010). Furthermore, the SDM usually has a higher level of goodness-of-fit compared with other spatial panel models (Wen and Liao, 2019). Since the null hypothesis of random effects is rejected (Prob>chi2=0.000 according to Hausman’s test), a SDM with fixed effects is applied. The impact of the forest-based bioeconomy on carbon emissions is twofold. Apart from the size of the bioeconomy, the number of patents in the forest-based bioeconomy and its rate of application (as proxy for technological innovation), as well as their interactions with the size of bioeconomy are selected as the core variables. In accordance with Grossman and Krueger (1995), two dimensionalities of the size of the 51 bioeconomy in Germany, namely the value added of bioeconomy (BV) and number of employees (BE) in the bioeconomy, are used to estimate the impact of the bioeconomy scale on carbon emissions. The number of patents (Number) in the forest-based bioeconomy and their application rate in the current year (Rate) are used to present two aspects of technological innovation: the former denotes the intensity of technological innovation and the latter denotes the transformation efficiency of scientific achievements, respectively (Popp et al., 2003; Harrahill et al., 2023). By accounting for the socioeconomic control variables, namely industrial upgrading (Structure), which is the ratio of GDP of the tertiary sectors to GDP of the industrial sectors, the labour density (Labour), which is the amount of labour per ha, the size of urban construction area (Urban) and per capita GDP (PerGDP), the basic SDM in the present analysis can be written as SDM 1: 𝑙𝑛𝑁𝐶𝐸 =𝜌𝑊𝑙𝑛𝑁𝐶𝐸+𝜕1𝑙𝑛𝐵𝑉+𝜕2𝑙𝑛𝐵𝐸+𝜕3𝑙𝑛𝑁𝑢𝑚𝑏𝑒𝑟+𝜕4𝑙𝑛𝑅𝑎𝑡𝑒+𝜕5𝑙𝑛𝐿𝑎𝑏𝑜𝑢𝑟+ 𝜕6𝑙𝑛𝑃𝑒𝑟𝐺𝐷𝑃+𝜕7𝑙𝑛𝑆𝑡𝑟𝑢𝑐𝑡𝑢𝑟𝑒+𝜕7𝑙𝑛𝑈𝑟𝑏𝑎𝑛+𝜑1𝑊𝑙𝑛𝐵𝑉+𝜑2𝑊𝑙𝑛𝐵𝐸+𝜑3𝑊𝑙𝑛𝑁𝑢𝑚𝑏𝑒𝑟+ 𝜑4𝑊𝑙𝑛𝑅𝑎𝑡𝑒+𝜑5𝑊𝑙𝑛𝐿𝑎𝑏𝑜𝑢𝑟+𝜑6𝑊𝑙𝑛𝑃𝑒𝑟𝐺𝐷𝑃+𝜑6𝑊𝑙𝑛𝑆𝑡𝑟𝑢𝑐𝑡𝑢𝑟𝑒+𝜑6𝑊𝑙𝑛𝑈𝑟𝑏𝑎𝑛+𝛾𝑙𝑛+𝜀 (3-7) With the spatial autocorrelation coefficient given by 𝜌, spatial weight matrix 𝑊, spatial lag of the dependent variable 𝑊𝑙𝑛𝑁𝐶𝐸, spatial lag of the explanatory variables 𝑊𝑙𝑛𝑋, matrix of the explanatory variables 𝑋, an n×1 vector of ones 𝑙𝑛, vectors of respective regression coefficients 𝜕, 𝜑, 𝛾 for X, 𝑊𝑙𝑛𝑋 and 𝑙𝑛 and the error term 𝜀. Growth-pole theory and empirical observations suggest that the more developed an area is, the stronger its spatial agglomeration effect on neighbouring regions (integration effect), because a higher development level creates centripetal forces on capital, technology and labour (Wen et al., 2016). The effect of technological innovation on net carbon emissions can, therefore, be mediated through the interacting input and output factors (resource allocation). In the extended model (SDM 2), this can be captured by accounting for interactions between the Number and mediating variables BV, BE, Structure, and PGDP: 52 𝑙𝑛𝑁𝐶𝐸 =𝜌𝑊𝑙𝑛𝑁𝐶𝐸+𝜕1𝑙𝑛𝐵𝑉+𝜕2𝑙𝑛𝐵𝐸+𝜕3𝑙𝑛𝑁𝑢𝑚𝑏𝑒𝑟+𝜕4𝑙𝑛𝑅𝑎𝑡𝑒+𝜕5𝑙𝑛𝐿𝑎𝑏𝑜𝑢𝑟+ 𝜕6𝑙𝑛𝑃𝑒𝑟𝐺𝐷𝑃+𝜕7𝑙𝑛𝑆𝑡𝑟𝑢𝑐𝑡𝑢𝑟𝑒+𝜕7𝑙𝑛𝑈𝑟𝑏𝑎𝑛+𝜷𝟏𝒍𝒏𝑵𝒖𝒎𝒃𝒆𝒓∗𝒍𝒏𝑩𝑽+𝜷𝟐𝒍𝒏𝑵𝒖𝒎𝒃𝒆𝒓∗𝒍𝒏𝑩𝑬+ 𝜷𝟑𝒍𝒏𝑵𝒖𝒎𝒃𝒆𝒓∗𝒍𝒏𝑺𝒕𝒓𝒖𝒄𝒕𝒖𝒓𝒆+𝜷𝟒𝑳𝒏𝑵𝒖𝒎𝒃𝒆𝒓∗𝒍𝒏𝑷𝒆𝒓𝑮𝑫𝑷+𝜑1𝑊𝑙𝑛𝐵𝑉+𝜑2𝑊𝑙𝑛𝐵𝐸+ 𝜑3𝑊𝑙𝑛𝑁𝑢𝑚𝑏𝑒𝑟+𝜑4𝑊𝑙𝑛𝑅𝑎𝑡𝑒+𝜑5𝑊𝑙𝑛𝐿𝑎𝑏𝑜𝑢𝑟+𝜑6𝑊𝑙𝑛𝑃𝑒𝑟𝐺𝐷𝑃+𝜑6𝑊𝑙𝑛𝑆𝑡𝑟𝑢𝑐𝑡𝑢𝑟𝑒+ 𝜑6𝑊𝑙𝑛𝑈𝑟𝑏𝑎𝑛+𝛾𝑙𝑛+𝜀 (3-8) where 𝛽 is a regression coefficient vector of the interactions. Considering the scale effects of GDP and labour on carbon emissions, the expected signs for BV and BE are positive (+). Technological innovation, denoted as the number of patents (Number) in the forest-based bioeconomy and the transformation rate (Rate) are assumed to have a negative impact on carbon emissions. Similarly, urban agglomeration (Urban and Labour) and economic growth (PGDP) will rather increase carbon emissions (Nakicenovic, 2000), suggesting the signs for Urban, Labour and PGDP are expected to be positive (+). Upgrading the industrial structure (Structure), by contrast, may lead to lower carbon emissions (-). Table 3.1 gives an overview of all the model variables. Table 3.1: Descriptions of the variables Name Units Obs Mean Std.Dev Min Max Sign NCE 106 tons 8822 610291.2 951385.7 62372.67 10600000 BV million Euro 8822 1371.585 2056.334 133.294 27605.67 + BE 103 persons 8822 24.739 27.5 3.831 357.031 + Number - 8822 17.64 25.68 0 271 - Rate % 8822 0.18 0.218 0 1 - Structure % 8822 0.520 0.327 0.037 3.994 - 53 Labour 103 persons 8822 24.903 14.777 2.421 73.560 + PGDP 103 Euro 8822 31.280 14.646 11.209 195.809 + Urban ha 8822 11978.92 7479.399 1212 62906 + 3.2.4 Data source The present study combined information from multiple data sources. The net carbon emissions, including direct carbon emissions and indirect carbon emissions as well as carbon sinks, were calculated on the basis of the land use data and emission parameters for different land use types at the county level. The emission parameters were derived from the Intergovernmental Panel on Climate Change (IPCC) (2021) data and related studies. The land use data from 2000 to 2021 were collected from the Thünen Land Atlas and Regional Database Germany. Due to the administrative division adjustment, some counties have been deleted and adjusted according to the counties/districts in 2021. For instance, Osterode am Harz has been adjusted as a municipality in the county of Göttingen since 2016, The annual socioeconomic data for 401 counties from 2000 to 2021 were gathered mainly from the Federal Office of Statistics of Germany, Regional Database Germany, Eurostat Database, and Federal Agency for Agriculture and Food (BMEL). The patent data in the forest-based bioeconomy from 2000 to 2021 were collected from the Organization for Economic Co-operation and Development (OECD) statistics. 3.3 Results and analysis 3.3.1 Spatiotemporal distribution of net carbon emissions Figure 3.2 summarizes the results for net carbon emissions (NCE) in Germany and for its division in Eastern and Western Germany in the period from 2000 to 2021. The results show a gentle downward trend in net emissions in Germany. A rapid drop in carbon emissions in 2009 followed by a rise again in 2010 may be because of the impact of economic depression caused by the global financial crisis at the time. Later, due to the economic decline resulting from the Covid-19 pandemic, carbon emissions dropped during the pandemic but climbed again significantly in 2021 to erase the earlier drop. Further, the much higher carbon 54 emissions in Western Germany than in Eastern Germany reflects the higher economic growth and industrial development in the former, and hence its higher associated NCE. Western Germany shared the same tendency of carbon emissions with Germany while Eastern Germany had a relatively small and stable fraction of net carbon emissions. Figure 3.2: Net carbon emissions in Germany during the period 2000-2021 Source: Own representation. The spatial distribution of net carbon emissions at the NUTS-3 level, as shown in Figure 3.3, changed in intensity over time. In addition to the city-states, like Berlin and Hamburg, the highest NCE (> 2000×103 tons) was produced in the western counties, indicating clustering patterns in Western Germany, while the lowest NCE (< 500×103 tons) was produced in the eastern counties. Noteworthy, the rapid decline of NCE 61 2020 0.071 -0.003 0.029 2.569 0.005 2021 0.072 -0.003 0.029 2.614 0.004 Source: Own calculation. Table 3.3 summarizes the results of the parameter estimations by using two versions of the spatial Durbin model, as described in section 3.2.3. The first model (SDM 1) includes the size of the bioeconomy, technological innovation, and socioeconomic parameters (cf. Equation 3-7), while its extensions control correspondingly for the effects of the interaction terms (SDM 2, cf. Equations 3-8). Varying the variables in two models reveal only insignificant effects, indicating the models’ stability and robustness. Table 3.3: Estimation results from the Spatial Durbin Model and its extensions Main effect on NCE (ρWlnNCE+∂lnX) Spillover effect of X on NCE (φWlnX) SDM 1 SDM 2 Interaction effect SDM 1 SDM 2 Interaction effect lnBV 0.198*** (0.006) 0.131*** (0.007) 0.055*** (0.013) 0.004 (0.013) lnBE 0.055*** (0.046) 0.116*** (0.007) 0.084*** (0.014) 0.100*** (0.015) lnNumber 0.002*** (0.005) -0.053*** (0.005) -0.002* (0.001) -0.021** (0.010) lnRate 0.001*** (0.000) 0.001* (0.000) 0.001 (0.001) 0.001 (0.001) lnPGDP 0.680*** (0.008) 0.024*** (0.002) -0.518*** (0.142) 0.007 (0.004) lnStructure -0.025*** (-3.24) -0.026*** (0.002) -0.013* (0.007) -0.0113** (0.004) lnLabour 0.007*** (0.000) -0.009*** (0.002) 0.001 (0.002) 0.004 (0.003) lnUrban -0.102*** -0.001 0.115*** 0.003** 62 (0.007) (0.001) (0.012) (0.001) LnNumber* lnBV 0.716*** (0.008) -0.465*** (0.016) LnNumber* lnBE -0.0157*** (0.003) -0.032*** (0.006) LnNumber* lnPGDP 0.00591*** (0.001) 0.001 (0.002) LnNumber* lnStructure -0.076*** (0.007) 0.063*** (0.011) ρ 0.420*** (0.016) 0.429*** (0.0135) R2 0.622 0.609 lgt_theta -4.304*** (0.392) -4.302*** (0.392) sigma2_e 0.001*** (1.07e-5) 0.001*** (9.86e-6) Log-Likelihood 19610.264 19719.822 Source: Own calculation. Note: t-statistics in parentheses; *statistical significance on p<0.10 level, ** p<0.05 level, *** p<0.01 level; 8822 observations. The results show that within SDM 1 the estimated carbon emissions (NCE) are positively correlated with the value added of the bioeconomy (BV), employees in the bioeconomy (BE), per capita GDP (PGDP), and labour density (Labour). This finding is in line with existing literature, arguing that economic factors are the main drivers of higher emissions (Wang et al., 2018; Zhang et al., 2020). An increase in urban construction land (Urban), in contrast, reduces NCE (-0.102***), because Germany is highly developed and has an advanced industrial division. German factories with high carbon emissions tend to be the less labour-intensive industries, while urban areas in Germany already have a high level of land development (Li et al., 2020). Industrial upgrading (Structure), as expected, drives down emissions significantly (- 63 0.025***), because it stimulates industrial transition towards greater sustainability (Bai et al., 2023; Mehmood et al., 2024). Contrary to expectations, an increase in the number of patents in the forest-based bioeconomy (Number) and the ratio of patents applied for in the current year (Rate), seems to give rise to the higher carbon emissions, respectively (0.002*** and 0.001***, respectively). This can be explained by the fact that technological innovation of forest-based bioeconomy in Germany contributes to economic growth, which ultimately lead to higher NCE levels (Khan et al., 2023). Further, the coefficients of the spatially lagged independent variables suggest that the value added of bioeconomy (BV), employees in bioeconomy (BE), labour density (Labour) and urban construction land (Urban) have significantly positive spillover effects in terms of higher emissions on neighbouring counties. The spillover effects of Number, Structure, and PGDP are significantly negative, indicating that counties with high levels of GDP and technological innovation, and thus more developed industries attract more technological investment and natural resources from neighbouring counties (Gao et al., 2020). The results for SDM 2 suggest that technological innovation in the forest-based bioeconomy can reduce net carbon emissions, given a stronger value added for the bioeconomy, and more jobs for employees in the bioeconomy, a higher per capita GDP, and industrial upgrading. The significantly negative spillover effect of Number highlights the role of spatial diffusion of technological innovation in the forest-based bioeconomy in reducing carbon emissions. As shown in Table 3.3, there is a strong interaction between Number and BV, BE, PGDP, and Structure compared to SDM 1. The significantly negative interactions between Number and BE and Structure reveal that, considering the impacts of BE and Structure on carbon emissions, an increase in technological innovation can start to reduce net carbon emissions. The significant positive interactions between Number and BV and PGDP indicate an opposite effect on emissions. Promoting technological innovation in the forest-based bioeconomy can consequently mitigate carbon emissions when combined with a greater number of employees in the bioeconomy and industrial 64 upgrading. The negative spillover effects of the interactions between Number and BV and BE on carbon emissions imply that an increase of technological innovation in the forest-based bioeconomy given the increase of BV and BE can reduce the carbon emissions in neighbouring counties, while the positive spillover effect of the interaction between Number and Structure reflects the opposite case. Table 3.4 displays the direct effect, indirect effect and the total effect of the parameters in SDM 1 and SDM 2. The total effect of Number on carbon emissions is negative whether in SDM 1 or SDM 2. This confirms the mitigation effect of technological innovation in the forest-based bioeconomy on carbon emissions. Further, the consistent direction of the direct effects of interaction between Number and BV, BE, PGDP, and Structure with their total effects not only implies their higher direct effects than direct effects but also stresses the combined action of technological innovation in the forest-based bioeconomy, labour structural change and industrial upgrading on carbon emissions. Table 3.4: Direct effect, indirect effect, and total effect of the model parameters SDM 1 SDM 2 Direct Indirect Total Direct Indirect Total lnBV 0.207*** (0.006) 0.230*** (0.019) 0.436*** (0.019) 0.137*** (0.007) 0.098*** (0.019) 0.235*** (0.022) lnBE 0.061*** (0.004) 0.177*** (0.022) 0.238*** (0.023) 0.130*** (0.007) 0.253*** (0.022) 0.382*** (0.024) lnNumber 0.002*** (0.001) -0.002 (0.002) -0.000 (0.003) -0.056*** (0.006) -0.074*** (0.016) -0.130*** (0.018) lnRate 0.001*** (0.000) 0.002 (0.002) 0.003 (0.002) 0.001* (0.000) 0.002 (0.001) 0.003 (0.002) lnPGDP 0.664*** (0.009) -0.387*** (0.024) 0.278*** (0.025) 0.025*** (0.002) 0.029*** (0.007) 0.054*** (0.007) lnStructure -0.027*** -0.037*** -0.064*** -0.028*** -0.038*** -0.066*** 65 (0.003) (0.010) (0.011) (0.002) (0.007) (0.008) lnLabour 0.007*** (0.001) 0.006 (0.004) 0.013*** (0.004) -0.009*** (0.002) -0.001 (0.005) -0.010 (0.006) lnUrban -0.098*** (0.007) 0.121*** (0.017) 0.022 (0.017) -0.001 (0.001) 0.004* (0.002) 0.003 (0.003) LnNumber* lnBV 0.701*** (0.009) -0.262*** (0.022) 0.439*** (0.024) LnNumber* lnBE -0.019*** (0.003) -0.064*** (0.008) -0.083*** (0.009) LnNumber* lnPGDP 0.006*** (0.001) 0.005** (0.003) 0.012*** (0.003) LnNumber* lnStructure -0.074*** (0.008) 0.050*** (0.017) -0.023 (0.017) Note: t-statistics in parentheses; *statistical significance at p<0.10 level, ** p<0.05 level, *** p<0.01 level; 8822 observations. 3.4 Discussion and conclusions 3.4.1 Discussion The analysis shows that technological innovation in the forest-based bioeconomy can contribute to decoupling economic development from emissions production. The obtained results are in line with recent studies, which found that technological innovation negatively affects carbon emissions (Erdoğan et al., 2020; Zhao et al., 2021). The results contribute to the body of the relevant literature by showing that the forest-based bioeconomy, combined with technological innovation in the forest-based bioeconomy and the number of employees in the bioeconomy, can improve the net carbon emissions performance within a region and through spatial spillover effects empirically. As shown by the values of coefficients, the number of patents in the forest-based bioeconomy (-0.053***), its interactions (LnNumber*lnBE and LnNumber*lnStructure), industrial upgrading (-0.108***), and labour intensity (-0.009***) can significantly 66 lower net carbon emissions. Likewise, the value added of bioeconomy (0.131***), number of employees in the bioeconomy (0.116***), application rate of patents in the forest-based bioeconomy (0.001*), and PGDP (0.024***) can all aid in reducing emissions. This is consistent with the latest findings suggesting that the growth of the bioeconomy results in a higher demand for biomass, while carbon tax will accelerate market opportunities for bio-based alternatives (Philippidis et al., 2024). Technological innovation in the forestbased bioeconomy also strengthens industrial upgrading (-0.026***), and enhances industry competition and labour division, contributing in this way to a more sustainable transition of industry and society (Bai et al., 2023; Mehmood et al., 2024). The key effect of the forest-based bioeconomy on carbon emissions is shown to be twofold, determined by the substitution/complementarity of the resources exchanged and technological diffusion among the counties. Resource substitution can drive industrial upgrading and the optimization of resource allocation in counties with high levels of technological innovation and large numbers of employees in the bioeconomy. The resulting negative effect on carbon emissions spills over to their neighbouring counties. Resource complementarity, for its part, weakens administrative barriers and strengthens regional cooperation, which explains the inconsistent spatial diffusion of carbon emissions (Figure 3.4) and value added of bioeconomy (Figure 3.6). This requires an efficient regulation of technological innovation and resource allocation in the bioeconomy to prevent their negative externalities (Zilberman et al., 2013). Other than the studies by Jonssen et al. (2021), in which the climate-change mitigation effect of the forestbased bioeconomy is investigated by considering the increased carbon storage in harvested wood products (HWP) at the EU level, the present analysis suggests that the contribution of the forest-based bioeconomy to carbon mitigation can not only be reflected by the carbon sinks in HWP but also through technological innovation as well as its spillover effects. The combinations of technological innovation and the number of employees in the bioeconomy and industrial upgrading further highlight their overlapping effects on carbon emissions. These observations allow concluding that an alignment of technological innovation with 67 industrial upgrading and a structural change in employment is needed to boost the forest-based bioeconomy to reduce carbon emissions (Halonen et al., 2022; Hetemäki et al., 2022). 3.4.2 Conclusions The forest-based bioeconomy, accompanied by the demanding forest biomass, technological innovation and value-added production, affects both the carbon footprint of economic activities and the carbon sink capacity of the ecological environment. Boosting the forest-based bioeconomy to benefit from its potential to promote carbon emissions reduction is a priority in the series of bioeconomy strategies in Germany. The study estimated the spatial impact of a forest-based bioeconomy, especially technological innovation in the forest-based bioeconomy, on carbon emissions. The analysis used the Spatial Durbin Model and countylevel panel data for 401 counties/cities and arrived at four main conclusions, as described below. First, for the observed period 2000 to 2021, the carbon emissions of 401 counties/cities in Germany have been found to be spatially autocorrelated and exhibit clustering patterns in Western Germany, largely reflecting the regional economic development. Second, technological innovation in the forest-based bioeconomy reveals a significant negative spillover effect on carbon emissions, indicating a role of technological diffusion in reducing carbon emissions from the local county/city to the periphery. Yet, the inconsistent diffusion trajectory of carbon emissions and the number of patents in the forest-based bioeconomy implies a high emissions reduction potential of a forest-based bioeconomy. Third, technological innovation in the forest-based bioeconomy can reduce carbon emissions through industrial upgrading and increasing job opportunities in the bioeconomy. Fourth, it can also lower carbon emissions through the negative spillover effect of industrial upgrading and increasing the size of the bioeconomy. 68 4 Impacts of institutional innovation in the bioeconomy on green productivity 3 4.1 Background and objectives 4.1.1 Institutional background of bioclusters in Germany German biocluster is an innovative strategy, serving as an institutional incentive for stimulating biotech industries and economic transformation. Bioclusters in Germany can date back to the 1970s when policymakers worldwide started to focus on biotechnology as a key innovation strategy (Fornahl et al., 2011; Dorocki, 2014). Despite Germany creating a national law on genetic modifications in 1978 to support biotechnology, it eventually fell well behind the global leaders in the following decades. It has been argued that Germany was the least biotechnology development-friendly country in the Western world in the early 1990s (Dohse and Staehler, 2008). However, in 1995, the German Federal Ministry of Education and Research (BMBF) announced the BioRegio competition to speed up the late-starting biotech industry. At that moment, there were only 70 biotech companies in Germany (BMBF, 2004). In this programme, winning regions could get preferential access to federal funding to realize their biotech investment plans (Dohse, 2000). After this initial programme, several others followed, like BioFuture, BioProfile and BioChance (Fornahl et al., 2011). The launch of a series of strategies regarding developing bioclusters has helped Germany reclaim its leading role in the bioeconomy. The German Biotechnology Report 2011 pointed out that the German biotech industry was back on a growth path in 2010, with 400 biotech companies and 809 million euros R&D expenditure (Ernst & Young, 2011). To better organize the Bioregions, the Council of BioRegions in Germany (AK-BioRegio) (also known as the alliance of the German Biotechclusters) was officially founded at the beginning of 2004 in Leipzig. The 3 Author statement: Lanjiao Wen (conceptualization, methodology, software, writing-original draft, and revision); Dr. Zhanli Sun (conceptualization, revision and supervision); Dr. Ir. Frans Hermans (conceptualization, data curation and revision); Prof. Dr. Alfons Balmann (revision and supervision). 69 tasks of AK-BioRegio includes five parts, namely meta-networking, biotech partnering, developing a trend radar (analysis of trends in biotechnology), innovation promotion, the provision of know-how for political decision-makers, and best practice exchange. Before 2004, biocluster managers in the BioRegio competition were competitors for public funding. But after 2004, with the establishment of AK-BioRegio, they established a network for sharing experiences and mutual learning. This further strengthened innovation diffusion and cooperation among companies, institutes, universities, and other stakeholders in the value chain of the bioeconomy. At the same time, the funding source was broadened to include private R&D investment. Now 24 members from Bioregions have come together to optimize and coordinate their regional activities in the interests of German biotechnology. 4.1.2 Objectives and organization The bioclusters, as an institutional innovation, create suitable ecosystems for the growth of the bioeconomy by linking biotech companies, research institutes and universities, technology parks, and relevant stakeholders in a geographic region together. As such, bioclusters can foster collaboration, innovation, knowledge exchange, and supply chain integration. Thus, bioclusters can greatly contribute to the development of the bioeconomy and facilitate the sustainable transition to the bioeconomy. To date, both national and supernational strategies have tended to focus on the sustainability of the bioeconomy, such as the EU Green Deal (“From Farm to Fork”, “Circular Economy Action Plan”, etc.). However, there is still little empirical evidence on whether and how the establishment of bioclusters affects green productivity. To fill in this research gap, this chapter, focusing on Germany at the NUTS-3 level, aims to estimate the causal effects of bioclusters on green productivity mediated by technological innovation. We attempt to answer the following research questions: 1) Does the establishment of bioclusters increase green productivity? If so, by how much? 2) How does the presence of bioclusters affect green productivity increases through technological innovation in the bioeconomy, and how can this effect be assessed through patent data? 70 The remainder of the paper is organized as follows. Section 4.2 outlines the background of bioclusters in Germany and provides the theoretical analysis and hypotheses. Section 4.3 introduces the study area, data, and methodology. The results are summarized and discussed in section 4.4. Finally, the discussion and policy implications are provided in section 4.5. 4.2 Theoretical analysis and hypotheses This section discusses the direct effects and indirect effects of bioclusters on green total factor productivity, where the indirect effects include technological, agglomeration, and structural effects. As the purpose of implementing bioclusters is to cultivate new dynamics for economic growth and to promote green development through innovation, bioclusters may have a direct impact on green productivity. Simultaneously, the establishment of bioclusters can contribute to promoting technological innovation, clustering innovation factors, and the transformation of industry structures. Therefore, technological, agglomeration, and structural effects arise that can indirectly influence green productivity (see Figure 4.1). Biocluster Technological innovation diffusion MRT ROT Industrial upgrading & restructuring Talent aggregation Capital aggregation Factors flow Energy consumption Production cost R&D investment Carbon emissions Economic growth Industry structure Labor structure Factors allocation G T F P Agglomeration effect Structural effect Cost of inputs Desirable output Undesirable output Note: GTFP=Green total factor productivity; MRT=Marginal rate of transformation; ROT=Return on investment + + Direct effect Direct effect Scientific & technological support Functions of bioclusters Resource integration Agents cooperation Input-output system Figure 4.1: Direct and indirect effects of bioclusters on GTFP Source: Own representation. 77 and Food (BMEL). The patent data in the forest-based bioeconomy from 2000 to 2021 were collected from Organization for Economic Co-operation and Development (OECD) Statistics. The records of bioclusters were from the German Trade and Invest survey (2022), which is supported by the Ministry of Education and Research and Federal Ministry for Economic Affairs and Climate Action, and the European Cluster Collaboration Platform. 4.3.2 Measuring regional green productivity with Super-efficiency SBM A super-efficiency Slacks-based measure (super-efficiency SBM) model with undesirable outcomes was employed to estimate the GTFP. Based on Tone(2002), the model is specified as below. 𝐺𝑇𝐹𝑃 =𝑚𝑖𝑛 1 𝑛∑𝑥𝑖 𝑥𝑖𝑜 ⁄ 𝑛 𝑖=1 1 𝑐1+𝑐2(∑𝑦𝑟 𝑑 𝑐1 𝑟=1 𝑦𝑟𝑜 𝑑 ⁄+∑𝑦𝑙𝑛𝑑 𝑐2 𝑙=1 𝑦𝑙𝑜 𝑛𝑑 ⁄ ) 𝑠.𝑡. ∑ 𝛾𝑥𝑖 𝑛 𝑖=1,≠0 ≤𝑥; ∑ 𝛾𝑦𝑖𝑑 𝑛 𝑖=1,≠0 ≥𝑦𝑟 𝑑; ∑ 𝛾𝑦𝑖𝑑 𝑛 𝑖=1,≠0 ≤𝑦𝑙𝑛𝑑; 𝑥 ≥𝑥𝑜; 𝑦𝑑≤𝑦𝑜 𝑑; 𝑦𝑛𝑑 ≤𝑦𝑜 𝑛𝑑; 𝑦𝑑≥0,𝛾 ≥0 (4-1) where 𝐺𝑇𝐹𝑃 is the urban land green use efficiency, and o is the production decision unit (401 decision units in total). Each decision unit has n inputs, c1 desired outputs, and c2 non-desired outputs. xio presents the input i of decision unit o; 𝑥 denotes the redundancy of input; 𝑦𝑟𝑜 𝑑 and 𝑦𝑙𝑜 𝑛𝑑 are the desired and undesired outputs of the decision unit o, respectively. 𝑦𝑑and 𝑦𝑛𝑑are the redundancy of the desired and non-desired outputs, respectively, and 𝛾 is the weight vector. The inputs and outputs are shown in Table 4.1. The sizes of the urban areas, employment population and energy consumption are used to present the fixed capital input, labour input and energy input, respectively. Gross domestic product (GDP) is the desired output and carbon emissions are the undesired output. In this study, the carbon emissions are net carbon emissions and are calculated as the sum of carbon emissions and carbon sink associated with land use at the county level (Wen et al., 2021). 78 Table 4.1: Input-output indicators for GTFP Inputs Descriptions Capital Size of the urban areas, as the sum of the settlement area and transport areas, in each county. Labour Total employment population in each county. Energy consumption Energy use of companies in the manufacturing sector at the NUTS-3 level. Desired outputs GDP Gross domestic product at the NUTS-3 level. Undesired output Carbon emissions The net carbon emission (NECi) for county i is shown as below: 𝑁𝐸𝐶𝑖=𝜂𝑎∙𝐴𝑖+𝑇𝑖∙𝑀𝑖−∑𝜂𝑗∙𝑆𝑖𝑗 , where 𝜂𝑎is the carbon emissions parameter for arable land; Aii is the arable land size; j=1,2,3 and represent respectively forest, grassland and water; Sij is the land size for each land use type j; 𝜂𝑗 is the product of carbon emissions parameter collected from the IPCC (2021); 𝑇𝑖 is the product of the energy consumption per unit of GDP; and 𝑀𝑖 is the GDP of the secondary and tertiary industries in county i. 4.3.3 Measuring the impact of Bioregions on green productivity with a staggered DiD The difference in differences (DiD) method is a widely-used quasi-experimental technique for estimating the effect of a specific intervention or treatment by comparing the changes in outcomes before and after the intervention (Goodman-Bacon, 2021). Considering the bioclusters established in different years, a staggered difference in differences (SDiD) is employed to compare the net effect on GTFP before and after the development of bioclusters. This approach addresses the limitation of traditional DiD that requires it to satisfy a stable unit treatment value assumption (SUTVA) while ignoring spillovers. To measure the treatment effect, counties/cites with bioclusters (Bioregions and green clusters in this study) are considered 79 as the treatment group and counties/cites without bioclusters, excluding those surrounding the counties/cities with bioclusters, are regarded as the control group with consideration of the spatial spillover effect. It assumes that the treatment and control groups display a parallel trend—the two groups would have followed similar trends over time (Slaughter, 2001). According to Hermans (2018), bioclusters are clusters that specialize in various fields of the bioeconomy with the explicit goal of promoting sustainable development. To better promote close cooperation among biotech companies, research institutes, and technology parks, Germany, one of the best environments for biotechnology R&D worldwide, has established the Council of BioRegions (AK-BioRegio) since 2004. The Bioregions of Germany are regional initiatives set up for the advancement of modern biotechnology in Germany. Up to 2022, there were 24 active members of Bioregions, ranging from those at the local level to state level. In the basic model shown in the following regression, counties/cites with bioclusters from Bioregions are the treatment group: 𝐺𝑇𝐹𝑃 = 𝜕0+𝜕1𝐵𝑖𝑜𝑟𝑒𝑔𝑖𝑜𝑛𝑖𝑡 +𝜕𝑐𝑐𝑜𝑛𝑡𝑟𝑜𝑙𝑖𝑡 +𝜇𝑖+𝜏𝑖+𝛿𝑖𝑡 (Model 4-1) with dummy variable Bioregion to show whether county/city i has a biocluster from Bioregions or not; control variable control; constant term 𝜕0; estimated parameters (𝜕1 and 𝜕𝑐); individual effect 𝜇𝑖, time effect 𝜏𝑖, and random perturbation term 𝛿𝑖𝑡. Bioregion is denoted as Bioregion=dt×du, where du determines whether they are qualified as innovative cities (yes=1, no=0), and dt determines whether they have already been designated as innovative cities (yes=1, no=0). Specifically, policy assigns a value of 1 to a city in year t and onward if it attains recognition as an innovative pilot city; otherwise, it receives a value of 0. In model 4-2, the value added of bioeconomy (BV), and employees in bioeconomy (BE) are introduced to model 4-1 to further examine the heterogeneous effects of Bioregions on GTFP, as shown below with the estimated parameters (𝜕2 and 𝜕3 ). 𝐺𝑇𝐹𝑃 = 𝜕0+𝜕1𝐵𝑖𝑜𝑟𝑒𝑔𝑖𝑜𝑛𝑖𝑡 +𝜕2𝐵𝑉𝑖𝑡 +𝜕3𝐵𝐸𝑖𝑡 +𝜕𝑐𝑐𝑜𝑛𝑡𝑟𝑜𝑙𝑖𝑡 +𝜇𝑖+𝜏𝑖+𝛿𝑖𝑡 (Model 4-2) 80 The patent application rate in the current year (Ratio) and regional share (reg_share) are used to present the characteristics of technological innovation (see Model 4-3). The patent application rate in the current year (Ratio) is the proportion of the number of patents applied for in the current year, which is relevant for assessing the transformation efficiency of scientific research and achievements (Harrahill et al., 2023). The regional share (reg_share) means the share of addresses of inventors in cases where an address is allocated to more than one region, therefore indicating regional cooperation (Maraut et al., 2008; Maraut and Martínez, 2014). 𝐺𝑇𝐹𝑃 =𝜕0+𝜕1𝐵𝑖𝑜𝑟𝑒𝑔𝑖𝑜𝑛𝑖𝑡 +𝜕2𝐵𝑉𝑖𝑡 +𝜕3𝐵𝐸𝑖𝑡 +𝜕4𝑅𝑒𝑔_𝑠ℎ𝑎𝑟𝑒𝑖𝑡 +𝜕5𝑅𝑎𝑡𝑖𝑜𝑖𝑡 +𝜕𝑐𝑐𝑜𝑛𝑡𝑟𝑜𝑙𝑖𝑡 +𝜇𝑖+ 𝜏𝑖+𝛿𝑖𝑡 (Model 4-3) 4.3.4 Measuring the impact of green clusters on green productivity with a staggered DiD As bioclusters are heterogeneous entities, varying widely in structure, evolution, and goals (Zechendorf, 2011), the impact of differential types of bioclusters on regional green productivity is considered in this study. To be specific, the green clusters that work in green sectors and/or technologies are selected from the European Cluster Collaboration Platform. According to Hermans (2018; 2021), green clusters operate with the goal of sustainable development, so they can be included as bioclusters and classified into four types, namely agricultural agglomeration, green chemistry clusters, bioeconomy districts and life science clusters. There are 29 green clusters chosen from the data European Cluster Collaboration Platform. The basic model of the SDiD is shown as below, where counties/cites with green clusters are the treatment group. 𝐺𝑇𝐹𝑃 = 𝛽0+𝛽1𝐺𝑐𝑙𝑢𝑠𝑡𝑒𝑟+𝛽𝑐𝑐𝑜𝑛𝑡𝑟𝑜𝑙𝑖𝑡 +𝜇𝑖+𝜏𝑖+𝛿𝑖𝑡 (Model 4-4) with dummy variable Gcluster to show whether county/city i has a green cluster or not. Gcluster is denoted as Gcluster =dt×du, where du determines whether they are qualified as innovative cities (yes=1, no=0), and dt determines whether they have already been designated as innovative cities (yes=1, no=0). Specifically, 81 policy assigns a value of 1 to a city in year t and onward if it attains recognition as an innovative pilot city; otherwise, it receives a value of 0. In the extended model 4-5, the value added of bioeconomy (BV) and the number of employees in the bioeconomy (BE) are introduced. 𝐺𝑇𝐹𝑃 = 𝛽0+𝛽1𝐺𝑐𝑙𝑢𝑠𝑡𝑒𝑟+𝛽2𝐵𝑉𝑖𝑡 +𝛽3𝐵𝐸𝑖𝑡 +𝛽𝑐𝑐𝑜𝑛𝑡𝑟𝑜𝑙𝑖𝑡 +𝜇𝑖+𝜏𝑖+𝛿𝑖𝑡 (Model 4-5) When considering the technological innovation by adding the patent application rate in the current year (Ratio) and regional share (reg_share) in model 4-6, get the following form. 𝐺𝑇𝐹𝑃 = 𝛽0+𝛽1𝐺𝑐𝑙𝑢𝑠𝑡𝑒𝑟+𝛽2𝐵𝑉𝑖𝑡 +𝛽3𝐵𝐸𝑖𝑡 +𝛽4𝑅𝑎𝑡𝑖𝑜𝑖𝑡 +𝛽5𝑅𝑒𝑔_𝑠ℎ𝑎𝑟𝑒𝑖𝑡 +𝛽𝑐𝑐𝑜𝑛𝑡𝑟𝑜𝑙𝑖𝑡 +𝜇𝑖+ 𝜏𝑖+𝛿𝑖𝑡 (Model 4-6) 4.3.5 Meditating model Based on Böckerman and Ilmakunnas (2009), a mediating model is developed to estimate the indirect effect of bioclusters on GTFP. With the introduction of the mediating variable 𝑀𝑖𝑡 (Equation 4-4) into model 43, the extended SDiD model 4-7 is shown as below. 𝑀𝑖𝑡 =𝛼0+𝛼1𝐵𝑖𝑜𝑟𝑒𝑔𝑖𝑜𝑛+𝛼𝑐𝑐𝑜𝑛𝑡𝑟𝑜𝑙𝑖𝑡 +𝜇𝑖+𝜏𝑖+𝛿𝑖𝑡 (4-4) 𝐺𝑇𝐹𝑃 =κ0+𝜅1𝐵𝑖𝑜𝑟𝑒𝑔𝑖𝑜𝑛𝑖𝑡 +𝜅2𝐵𝑉𝑖𝑡 +𝜅3𝐵𝐸𝑖𝑡 +𝜅4𝑅𝑒𝑔_𝑠ℎ𝑎𝑟𝑒𝑖𝑡 +𝜅5𝑅𝑎𝑡𝑖𝑜𝑖𝑡 +𝜅𝑚𝜕𝑐𝑀𝑖𝑡 + 𝜅𝑐𝑐𝑜𝑛𝑡𝑟𝑜𝑙𝑖𝑡 +𝜇𝑖+𝜏𝑖+𝛿𝑖𝑡 (Model 4-7) Where 𝛼 and 𝜅 are the respective regression coefficients. Similarly, adding 𝑀𝑖𝑡 ′ (Equation 4-5) into model 4-6 gives the extended SDiD, as shown in model 5-8, with vectors for the respective regression coefficients 𝜉. 𝑀𝑖𝑡 ′ =𝜁0+𝜁1𝐺𝑐𝑙𝑢𝑠𝑡𝑒𝑟+𝜁𝑐𝑐𝑜𝑛𝑡𝑟𝑜𝑙𝑖𝑡 +𝜇𝑖+𝜏𝑖+𝛿𝑖𝑡 (4-5) 𝐺𝑇𝐹𝑃 =𝜉0+𝜉1𝐺𝑐𝑙𝑢𝑠𝑡𝑒𝑟+𝜉2𝐵𝑉𝑖𝑡 +𝜉3𝐵𝐸𝑖𝑡 +𝜉4𝑅𝑎𝑡𝑖𝑜𝑖𝑡 +𝜉5𝑅𝑒𝑔_𝑠ℎ𝑎𝑟𝑒𝑖𝑡 +𝜉𝑚𝑀𝑖𝑡 ′ +𝜉𝑐𝑐𝑜𝑛𝑡𝑟𝑜𝑙𝑖𝑡 + 𝜇𝑖+𝜏𝑖+𝛿𝑖𝑡 (Model 4-8) 82 The indirect effect of bioclusters on GTFP encompasses three distinct effects: technological, agglomeration, and structural, denoted respectively by technological innovation (Patent), market capacity (MC), and industrial structure (ST). Usually, the number of patents (Patent) is denoted as the intensity of technological innovation, while the application rate of each patent can denote the transformation efficiency of scientific research and achievement (Popp et al., 2003; Harrahill et al., 2023). The number of patents (Patent) is used to signify the technological effect, as patent applications can serve as indicators of both the quantity and quality of innovations within a county/city during a specific timeframe (Liu et al., 2023). The ratio of the tertiary industry to the secondary industry (ST) serves as an indicator of industrial upgrading, whereby a higher ratio signifies a more advanced industrial structure within the county/city. Market capacity (MC) is utilized to represent the agglomeration effect, given that enterprises often gravitate towards regions with robust market potential, fostering the clustering of resources within such counties/cities (Wu and Shao, 2016). MC is calculated using Equation 4-6. 𝑀𝐶𝑖𝑡 =𝑆𝑇𝐺𝐷𝑃𝑖𝑡 𝑑𝑖𝑡 +∑𝑆𝑇𝐺𝐷𝑃𝑘𝑡 𝑑𝑖𝑘 𝑖≠𝑘 ;𝑑𝑖𝑡 =2 3(𝑎𝑟𝑒𝑎𝑖 𝜋)1 2 (4-6) where 𝑀𝐶𝑖𝑡 denotes the market potential of county/city i. 𝑆𝑇𝐺𝐷𝑃𝑖𝑡 and 𝑆𝑇𝐺𝐷𝑃𝑘𝑡 denote the output value of secondary and tertiary industries in county/city i and k respectively in year t. 𝑑𝑖𝑡 denotes the internal distance of county/city i in year t, and areai denotes the urban area of city county/city i. 𝑑𝑖𝑘 is the distance between county/city i and k, calculated using latitude and longitude data. 4.3.6 Variables Considering the positive contribution to economic output, the expected signs for BV and BE are positive (+). As the higher level of bioclusters, the more competitive the clusters are, the sign for the local level of biolcusters (Level-1) is expected to be negative, while for the regional and state level the signs are positive. Technological innovation in the bioeconomy is assumed to have a positive impact on regional green productivity, where the signs for the number of patents (Patent) and patents’ transformation rate (Ratio) are positive, while that for regional share (Reg_share) is negative. In terms of the type of green clusters, all 83 types are assumed to have a positive impact on green productivity. Among them, green chemical clusters (Type-2) may have the highest contribution. Similarly, upgrading the industrial structure (Structure), and number of De-domain (Domain) rather increase the regional productivity (Dierckx and Stroeken, 1999), suggesting the signs for Structure and Domain are expected to be positive (+). The development of the Internet has a significant effect on promoting improvements to GTFP in this region and the surrounding areas, but also suggests that the long-term effect is greater than the short-term effect (Yu, 2022). Table 4.2 gives an overview of all the model variables. Table 4.2: Descriptive statistics of variables Name Units Mean Std.Dev Min Max Sign GTFP - 0.205 0.125 0.062 1.354 BV 10 million Euro 1371.585 2056.334 133.294 27605.67 + BE 103 persons 24.739 27.5 3.831 357.031 + Bioregion 0.041 0.198 0 1 + Level-1 0.004 0.065 0 1 + Level-2 0.019 0.138 0 1 + Level-3 0.031 0.174 0 1 + Gcluster 0.037 0.190 0 1 + Type-1 0.002 0.040 0 1 + Type-2 0.007 0.084 0 1 + Type-3 0.033 0.177 0 1 + Patent 19.659 29.506 0 389 + 84 Ratio % 0.181 0.217 0 1 + Reg_share % 0.664 0.310 0 1 - Structure % 0.520 0.327 0.037 3.994 + Domain 27566.59 45225.02 0 630403 + 4.4 Results and analysis 4.4.1 Spatiotemporal distribution of net carbon emissions Figure 4.3 summarizes the results for the average annual GTFP in Germany in the period 2000 to 2021. The results show an upward trend in GTFP over the study period, indicating an increasing growth in regional green productivity. However, the average value for the GTFP stays below 0.5 during the study period, showing a relatively low green productivity in Germany. Specifically, the average GTFP from 2000 to 2015 grew slowly, albeit with a slight drop in 2009 due to the delayed impact of the global financial crisis. After 2015, the average GTFP increased more rapidly, increasing from 0.21 in 2015 to 0.441 in 2021. In particular, during the period from 2018 to 2021, the GTFP grew at an even higher rate, despite the economic decline after the Covid-19 pandemic. The obtained results are not surprising, given the series of strategies from the government targeted primarily at boosting the bioeconomy to deal with the increasing climate crisis (BMBF, 2020). In addition, the COVID-19 pandemic also contributed to the reduction of carbon emissions, leading to the sharp increase of GTFP. 85 Figure 4.3: Average annual GTFP in Germany during the period 2000–2021 The spatial distribution of GTFP at the county level, as shown in Figure 4, changed in intensity over time. Western and southern counties tend to have higher GTFPs. Counties in eastern Germany (e.g. Wittenberg and Salzlandkreis) have lower but increasing GTFPs observed. Noteworthy, here, most counties surrounding developed municipalities, like Frankfurt and Munich, have large development potential and a relatively high absorbing capacity for investments from more developed economic centres). As a result, the number of counties with the lowest GTFP (<0.2) declined steadily over the observed time period. In all other categories (> 0.2), there was a clear increase trend over time. By overlapping the spatial distribution of GTFP with bioclusters, we found that most counties with bioclusters (e.g., Potsdam) tend to have higher GTFPs. And they showed spillover effects on surrounding counties from 2000 to 2021. For instance, the 86 number of cities/counties around Düsseldorf with a GTFP greater than 0.2 gradually increased from 2000 to 2021. From 2015 to 2021, both the GTFP and its spatial spillover effects continued to grow sharply. Figure 4.4: Spatial county-level distribution of GTFP in 2000, 2005, 2010, 2015, and 2021 4.4.2 SDiD regression results Table 4.2 summarizes the results of the parameter estimations using three versions of the SDiD model, developed and described in section 4.3.2. The first model (Model 4-1) only includes the treatment group (Bioregion) and the control variable (Domain), while its extensions control correspondingly for the effects of the size of the bioeconomy (Model 4-2) and of technological innovation (Model 4-3). Varying the variables, reveals relatively insignificant effects, indicating the models’ stability and robustness. 93 that establishing bioclusters can improve the GTFP. Other coefficients show similar values, indicating a robust result. Table 4.5: PSM-SDiD regression results Variables GTFP Bioregion 0.020*** (3.090) Gcluster 0.025*** (3.240) BV 0.0001*** (34.630) 0.0002*** (45.390) BE -0.005*** (-12.170) -0.009*** (-23.970) Rate 0.075*** (17.930) 0.071*** (18.150) Reg_share -0.035*** (-8.940) -0.032*** (-9.18) Domain 3.09e-7*** (6.590) 6.70e-7*** (12.190) Fe Yes Yes 94 N 6226 6494 R2 0.146 0.278 Note: ***p<0.01, **p<0.05, *p<0.1. Bioregion denotes the key explanatory variable. Fe indicates time fixed and individual fixed. N indicates the total sample size. R2 denotes the coefficient of determination. 4.4.5 Mediating effects of bioclusters on GTFP The three mediators for technological, agglomeration, and structural effects in Table 4.6 (0.008, 0.121, and -0.004, respectively) show that the agglomeration effect is much greater than the other two. It means bioclusters primarily promote GTFP by enhancing technological innovation and market capacity. Column (1) in Table 4.5 displays a positive and significant impact of bioclusters on local GTFP in the results for the SDiD base regression, providing a basis for the tests of the mediating effects (from column 2 to column 7). All the mediating effects passed the bootstrap tests. In column (3), the significant coefficients of Bioregion and Patent indicate there is a strong mediation effect (0.01), suggesting that bioclusters can promote technological innovation in the local city. Despite the coefficient of Patent being negative, the results from the bootstrap test confirm its validity ([0.201,0.275]). This contributes to the promotion of green production technology and pollution control technology, minimizing emissions, and consequently improving the GTFP. Thus, this finding supports H2. The mediating effect of industrial upgrading is negative because of the negative coefficient of ST in column (7). Table 4.6: Mediating regression results for Bioregions Mediating model for Bioregions Variables GTFP (1) Patent (2) GTFP (3) MC (4) GTFP (5) ST (6) GTFP (7) Bioregion 0.020*** -0.236 0.022*** 175.498*** -0.009* -0.053*** 0.014** 95 (3.090) (-0.220) (3.090) (8.07) (-1.650) (-5.680) (2.170) Patent -0.001*** (-11.73) MC 0.0002*** (51.230) ST -0.117*** (-12.910) BV BE Mediating effect 0.008 0.121 -0.004 Control Yes Yes Yes Yes Yes Yes Yes Fe Yes Yes Yes Yes Yes Yes Yes N 6226 6226 6226 6226 6226 6226 6226 R2 0.146 0.090 0.159 0.084 0.175 0.099 0.156 Note: ***p<0.01, **p<0.05, *p<0.1. Bioregion denotes the key explanatory variable. Fe indicates time fixed and individual fixed. N indicates the total sample size. R2 denotes the coefficient of determination. The three mediators for technological, agglomeration, and structural effects in Table 4.7 (-0.002, 0.03, and 0.005, respectively) show a similar result, whereby the agglomeration effect is much greater than the other two. It means bioclusters primarily promote GTFP by enhancing market capacity. Unlike the results in 96 Table 4.6, the coefficient of Gcluster in column (6) is negative. This implies that the establishment of green clusters can restrict industrial upgrading. This may be due to the external economies generated by green clusters, which boost the output value of secondary industries and increase their share in the overall economy. The coefficient of Gcluster in column (2) is significantly positive, indicating that green bioclusters can promote the intensity of technological innovation. However, the mediating effect of technological innovation on GTFP is negative, meaning that the force of green clusters that promote technological innovation is greater than its mediating effect through technological innovation on GTFP. These results align with the findings of Graf and Broekel (2020), who highlight the impact of Bioregion initiatives in promoting technological innovation only during the funding period. This may be because a simple financial injection into projects to support technological innovation is not able to cause the emergence of clusters for decoupling economic growth from pollution (Kamath et al., 2022). Table 4.7: Mediating regression results for green clusters Mediating models for green clusters Variables GTFP (1) Patent (2) GTFP (3) MC (4) GTFP (5) ST (6) GTFP (7) Gcluster 0.018*** (2.320) 3.685*** (2.630) 0.021*** (2.780) 133.318** (4.98) -0.003 (-0.45) -0.044*** (-3.87) 0.013* (1.70) Patent -0.001*** (-13.07) MC 0.0002*** (50.62) ST -0.111*** (-13.03) Mediating effect -0.002 0.030 0.005 97 Control Yes Yes Yes Yes Yes Yes Yes Fe Yes Yes Yes Yes Yes Yes Yes N 6512 6512 6512 6512 6512 6512 6512 R2 0.260 0.069 0.203 0.085 0.137 0.011 0.219 Note: ***p<0.01, **p<0.05, *p<0.1. Bioregion denotes the key explanatory variable. Fe indicates time fixed and individual fixed. N indicates the total sample size. R2 denotes the coefficient of determination. 4.4.6 Heterogeneity analysis with difference-in-difference-in-differences (DDD) In this study, the level of Bioregions and types of green clusters, are introduced to further explore the heterogeneous implications of bioclusters on regional green productivity. The DDD model is employed in the present study to identify causal effects by comparing differences in changes in outcome variables before and after the intervention between treatment and control groups. Through structuring a third dimension of the treatment group (Bioregion*Level and Gcluster*Type, respectively in this study), DDD is used to identify the heterogeneous treatment effects of intervention policies across groups. For Bioregions, three levels of Bioregions, namely local level (Level-1), regional level (Level-2), and state level (Level-3) are introduced. The results shown in Table 4.8 illustrate that even though all levels of Bioregions can contribute to the increase of GTFP, the regional Bioregions have a significantly positive impact on GTFP. The higher the level of the bioclusters, the lower the parameter is. This implies a lower level of bioclusters may involve a closer cooperation between firms and research institutions as bioclusters highly depend on industry– university–research integration (Youtie and Shapira, 2008; Jeong et al., 2023). Table 4.8: Regression results of DDD model for Bioregions Variables GTFP Bioregion*Level-1 0.046 (1.54) 98 Bioregion*Level-2 0.032** (2.91) Bioregion*Level-3 0.010 (1.25) BV 0.0001*** (35.63) 0.0001*** (35.38) 0.0001*** (35.03) BE -0.004*** (-12.09) -0.004*** (-12.06) -0.005*** (-12.12) Ratio 0.076*** (18.13) 0.076*** (18.04) 0.076*** (18.06) Reg_share -0.0348*** (-8.99) -0.035*** (-8.95) -0.035*** (-8.99) Domain 3.23e-7*** (6.91) 3.13e-7*** (6.69) 3.20e-7*** (6.83) Fe Yes Yes Yes N 6226 6226 6226 R2 0.142 0.144 0.141 Note: ***p<0.01, **p<0.05, *p<0.1. Bioregion denotes the key explanatory variable. Fe indicates time fixed and individual fixed. N indicates the total sample size. R2 denotes the coefficient of determination. 99 For green clusters, three types of green clusters, namely agricultural agglomeration (Type-1), green chemistry clusters (Type-2), and bioeconomy districts (Type-3) are introduced to estimate the heterogeneity. The results shown in Table 4.9 illustrate that all kinds of bioclusters can promote GTFP. Among them, the higher coefficient of green chemistry clusters (Type-2) indicates a greater contribution. The coefficients of the other variables are quite similar, showing the results are relatively robust. Table 4.9: Regression results of DDD model for green clusters Variables GTFP Gcluster*Type-1 0.019*** (1.31) Gcluster*Type-2 0.077*** (3.78) Ccluster*Type-3 0.018*** (2.66) BV 0.0001*** (41.91) 0.0001*** (42.11) 0.0001*** (41.95) BE -0.009*** (-24.93) -0.009*** (-25.19) -0.009*** (-25.08) Rate 0.078*** (19.65) 0.078*** (19.60) 0.078*** (19.69) Reg_share -0.032*** (-8.94) -0.032*** (-8.88) -0.032*** (-8.95) Domain 3.48e-7 (6.83) 3.45e-7 (6.78) 3.27e-7 (6.36) Fe Yes Yes Yes 100 N 6512 6512 6512 R2 0.226 0.263 0.262 Note: ***p<0.01, **p<0.05, *p<0.1. Bioregion denotes the key explanatory variable. Fe indicates time fixed and individual fixed. N indicates the total sample size. R2 denotes the coefficient of determination. 4.5 Discussion and conclusions 4.5.1 Discussion The analysis shows that the development of bioclusters, as an institutional innovation, can contribute to improving regional green productivity. The obtained results are in line with recent studies, which found that technological innovation caused by regulation policies can be a new engine to stimulate economic growth and protect the environment at the same time (Alvarez-Herranz et al., 2017; Balsalobre-Lorente et al., 2020). The results contribute to the body of the relevant literature by showing that the policy of developing bioclusters may improve the green total factor productivity directly and indirectly through mediating effects from technological innovation, factor agglomeration, and industrial upgrading. As shown by the values of coefficients in Table 4-5, both Bioregion (0.02***) and Gcluster (0.025***) as well as the value added of bioeconomy (0.0001*** and 0.0002***, respectively) and patent application rate in the bioeconomy (0.075*** and 0.071***, respectively) can significantly promote the increase of GTFP. The mediating effect of factor agglomeration is found to be the biggest, as shown in Tables 4-6 (0.121) and 4-7 (0.03). Differing from Du and Li (2022), where the mediating effect includes government strategic leadership in addition to technological innovation and industry upgrading, this study instead considers the agglomeration effects. Specifically, the agglomeration effect is found to be the strongest mediating force when bioclusters affect GTFP. This highlights the roles of production factor flows and resource allocation in green urban efficiency, contributing new insights to the literature in this field (Rusiawan et al., 2015; Atesagaoglu et al., 2017; Liu and Xin, 2019). Compared with previous studies where technological innovation is denoted by the number of patents or R&D investment (Du and Li, 2019; Luo et al., 2022), the present analysis suggests 101 that the regional share of patents and the ratio of patent applications that are highly linked with bioclusters can be extended to indicate the level of technological innovation. These observations allow the conclusion that an alignment of bioclusters with technological innovation, regional market capacity, and industrial structure is needed to promote green economic growth and sustainable transition to bioeconomy (Chen et al., 2021; Van Lancker et al., 2016; Wilde and Hermans, 2024). The heterogeneity analysis shows that the impact of bioclusters on GTFP varies with the types of bioclusters. The obtained results show that all kinds of bioclusters can promote GTFP, but chemical green clusters make the greatest contribution, which is in line with recent studies, which found that green technological innovation can promote GTFP (Liu et al., 2024; Zhang and Wang, 2022). Methodologically, this study addresses the variation in treatment timing when employing DiD (Callaway and Sant’Anna, 2021) by combining SDiD and PSM in the exogeneity test and DDD in the heterogeneous analysis (Du and Li, 2022; Guo and Zhong, 2022). In this way, the mechanism and synergies behind regional green productivity, not detectable from federalor state-level data (Gurney et al., 2019; Li et al., 2020; Wang and Jiang, 2020), can be illuminated. In addition, the observed finding that a lower level of bioclusters has a larger impact on GTFP also implies that lower-level bioclusters are more efficient for improving regional GTFP. Using data at the NUTS-3 level can inform the development of tailored, regionally specific policy instruments for GTFP improvement. 4.5.2 Conclusions In the context of the increasing demands for environmental and economic alignment in various countries worldwide, boosting the bioeconomy to improve regional productivity through technological innovation and institutional innovation is expected to be an efficient pathway towards future sustainable development. Focusing on the 401 NUTS-3 level (counties/cities) in Germany from 2000 to 2021, the study estimates the impact of bioclusters on green productivity mediated by technological innovation, industrial upgrading, and market capacity. We used a series of DiD models (SDiD, PSM-SDiD, and DDD) and mediating models, and arrived at four main conclusions. 102 First, for the observed period 2000 to 2021, the regional productivity of the 401 counties/cities were found to be spatially autocorrelated and exhibited clustering patterns, largely reflecting the overlap between bioclusters and regional productivity. Second, developing bioclusters, no matter as Bioregions or green clusters, has a positive effect on GTFP, both in a direct manner and indirect manner through technological innovation and market agglomeration. Third, different types of bioclusters have heterogeneous impacts on GTFP, with chemical green clusters making the greatest contribution. Furthermore, the larger the level of the bioclusters, the less contribution the bioclusters makes to GTFP. Fourth, regional GTFP can be improved through the positive effect of the value added of bioeconomy, while there are negative effects of the number of employees in the bioeconomy, regional share of patents in the bioeconomy, and the level of bioclusters. The findings point to a high potential of establishing bioclusters to improve the regional green total factor productivity performance. This potential can be realized if policy measures account for the regional heterogeneity in economic strength and resource endowment at a possibly reasonable spatial scale. The study thus advocates for the transition to a bioeconomy, while highlighting the role of different regional drivers of GTFP. Still, due to the limited availability of some county-level data, especially for R&D investment, and sectoral diversity, the findings need to be verified by further research, which should also account for the impact of the more recent programmes for bioclusters. 109 compensation and non-food biomass supply in the decision making of farmers and their impacts on the structural change of agriculture should be measured in the future. Adding farmers’ behaviours under the transition to bioeconomy into the green reform of the CAP will facilitate a more cohesive and effective bioeconomic transition, but also promote a unified approach to sustainable development. Third, comparative studies between Germany and other countries with different bioeconomic strategies will be useful and yield valuable insights into best practices and potential pitfalls. This dissertation mainly focuses on the case of Germany, the leading country in promoting bioeconomy. However, many transition countries, like China and India, have launched bioeconomy strategies. Specifically, the bioeconomy in China got off to a late start and had a relatively small scale, but developed rapidly (the average annual growth rate has exceeded 20%). Up to 2018, China is the third largest bioethanol producer (2050 thousand tons in 2018) in the world after the United State (44100 thousand tons in 2018) and Brazil (21280 thousand ton in 2018) (FAO, 2018). In particular, there are 11 provinces that have been selected as pilot areas to promote the production and application of bioethanol. Therefore, a comparison between Germany and China may provide heterogeneous experiences for other countries around the world. 110 References Aguilar, A., Twardowski, T., & Wohlgemuth, R. (2019). Bioeconomy for sustainable development. Biotechnology Journal, 14(8), 1800638. Aguilar, A., Wohlgemuth, R., & Twardowski, T. (2018). Perspectives on bioeconomy. New Biotechnology, 40: 181-184. Alemu, M. (2020). Trend of Biotechnology Applications in Pest Management: A Review. International Journal of Applied Sciences and Biotechnology, 8(2), 108-131. Altman, I., Klein, P.G., & Johnson, T.G. (2007). Scale and Transaction Costs in the U.S. Biopower Industry. Journal of Agricultural & Food Industrial Organization, 5(1), 19. Alvarez-Herranz, A., Balsalobre-Lorente, D., Shahbaz, M., & Cantos, J. M. (2017). Energy innovation and renewable energy consumption in the correction of air pollution levels. Energy Policy, 105, 386397. Amidon, T. E., Bujanovic, B., Liu, S., & Howard, J. R. (2011). Commercializing biorefinery technology: A case for the multi-product pathway to a viable biorefinery. Forests, 2(4), 929-947. Anselin, L., 2013. Spatial econometrics: methods and models. Springer Science & Business Media. Appel, F., Ostermeyer-Wiethaup, A., & Balmann, A. (2016). Effects of the German Renewable Energy Act on structural change in agriculture–The case of biogas. Utilities Policy, 41, 172-182. Arrow, K. J. (1969). The organization of economic activity: issues pertinent to the choice of market versus nonmarket allocation. The analysis and evaluation of public expenditure: the PPB system, 1, 5973. Auty, M. R. (2007). Natural resources, capital accumulation and the resource curse. Ecological Economics, 61, 627-634. 111 Ayrapetyan, D., Befort, N., & Hermans, F. (2022). The role of sustainability in the emergence and evolution of bioeconomy clusters: An application of a multiscalar framework. Journal of Cleaner Production, 376, 134306. Bachmaier, J., Effenberger, M., & Gronauer, A. (2010). Greenhouse gas balance and resource demand of biogas plants in agriculture. Engineering in Life Sciences, 10 (6), 560-569. Backhouse, M., Lorenzen, K., Lühmann, M., Puder, J., Rodríguez, F., & Tittor, A. (2017). BioökonomieStrategien im Vergleich. Gemeinsamkeiten, Widersprüche und Leerstellen. Bioeconomy & Inequalities, Working Paper Nr. 1, Jena. Bai, C. E., Ma, H., & Pan, W. (2012). Spatial spillover and regional economic growth in China. China Economic Review, 23(4), 982-990. Bai, T., Qi, Y., Li, Z., & Xu, D. (2023). Digital economy, industrial transformation and upgrading, and spatial transfer of carbon emissions: The paths for low-carbon transformation of Chinese cities. Journal of Environmental Management, 344, 118528. Baležentis, T., Streimikiene, D., Zhang, T., & Liobikiene, G. (2019). The role of bioenergy in greenhouse gas emission reduction in EU countries: An Environmental Kuznets Curve modelling. Resources, Conservation and Recycling, 142, 225-231. Bala, B. K., Arshad, F. M., & Noh, K. M. (2017). System dynamics. Modelling and Simulation, 274. Balmann, A. (1997). Farm-based modelling of regional structural change: a cellular automata approach. European Review of Agricultural Economics, 24(1), 85-108. Balsalobre-Lorente, D., Driha, O. M., Bekun, F. V., & Osundina, O. A. (2019). Do agricultural activities induce carbon emissions? The BRICS experience. Environmental Science and Pollution Research, 26(24), 25218-25234. Batáry, P., Gallé, R., Riesch, F., Fischer, C., Dormann, C. F., Mußhoff, O., ... & Tscharntke, T. (2017). The former Iron Curtain still drives biodiversity–profit trade-offs in German agriculture. Nature Ecology & Evolution, 1(9), 1279-1284. 112 Barlas, Y. (1996). Formal aspects of model validity and validation in system dynamics. System Dynamics Review: The Journal of the System Dynamics Society, 12(3), 183-210. Baumol, W.J., 1996. Entrepreneurship: productive, unproductive, and destructive. J. Bus. Ventur. 11 (1), 3-22. Beer, L., & Heise, H. (2020). Citizens' Attitudes towards Greening as Part of Common Agricultural Policy: Results of a Panel Survey. German Journal of Agricultural Economics, 69(3), 173-182. Bell, J., Paula, L., Dodd, T., Németh, S., Nanou, C., Mega, V., & Campos, P. (2018). EU ambition to build the world’s leading bioeconomy—Uncertain times demand innovative and sustainable solutions. New Biotechnology, 40, 25-30. Benbi, D. K. (2018). Carbon footprint and agricultural sustainability nexus in an intensively cultivated region of Indo-Gangetic Plains. Science of The Total Environment, 644, 611-623. Bianco, V., Cascetta, F., & Nardini, S. (2024). Analysis of the carbon emissions trend in European Union. A decomposition and decoupling approach. Science of The Total Environment, 909, 168528. Biber‐Freudenberger, L., Ergeneman, C., Förster, J. J., Dietz, T., & Börner, J. (2020). Bioeconomy futures: Expectation patterns of scientists and practitioners on the sustainability of bio ‐based transformation. Sustainable Development, 28(5), 1220-1235. Bioökonomierat. (2010). Bio-economy innovation—research and technological development to ensure food security, the sustainable use of resources and competitiveness. BMBF & BMEL. 2023. Bioeconomy in Germany: Opportunities for a bio-based and sustainable future. Berlin, Germany, URL: https://www.bmbf.de/SharedDocs/Publikationen/de/bmbf/FS/31106_Biooekonomie_in_Deutschl and_en.pdf?__blob=publicationFile&v=5. BMBF (2020). National Bioeconomy Strategy. Federal Ministry of Education and Research, Division “Bioeconomy, Material Biomass Use”. URL: https://www.bmbf.de/bioeconomy. 113 BMBF. (2004). BioRegions in Germany: Strong impulses for national technological development. Berlin. https://d-nb.info/971374856/34 BMBF. (2011). National Research Strategy BioEconomy 2030 Our Route towards a Biobased Economy. Berlin. BMBF. (2020). National Bioeconomy Strategy. Federal Ministry of Education and Research, Division “Bioeconomy, Material Biomass Use”, URL: www.bmbf.de/bioeconomy. BMELV (2009): Aktionsplan der Bundesregierung zur stofflichen Nutzung nachwachsender Roh-stoffe. Bundesministerium für Ernährung, Landwirtschaft und Verbraucherschutz. BMELV/BMU. (2009). Nationaler Biomasseaktionsplan für Deutschland. Beitrag der Biomasse für eine nachhaltige Energieversorgung. Bundesministerium für Ernährung, Landwirtschaft und Verbraucherschutz/Bundesministerium für Umwelt, Naturschutz und Reaktorsicherheit. BMU. (2007). Nationale Strategie zur biologischen Vielfalt. Bundesministerium für Umwelt, Naturschutz und Reaktorsicherheit. Böckerman, P., & Ilmakunnas, P. (2009). Unemployment and self‐assessed health: evidence from panel data. Health economics, 18(2), 161-179. Bockstael, N., Costanza, R., Strand, I., Boynton, W., Bell, K., & Wainger, L. (1995). Ecological economic modeling and valuation of ecosystems. Ecological Economics, 14(2), 143-159. Boeing, P., & Hünermund, P. (2020). A global decline in research productivity? Evidence from China and Germany. Economics Letters, 197, 109646. Buckley, P. J., Chapman, M. (1998). The perception and measurement of transaction costs, International Business. Palgrave Macmillan, London, 57-86. Budzinski, M., Bezama, A., & Thrän, D. (2017). Monitoring the progress towards bioeconomy using multiregional input-output analysis: The example of wood use in Germany. Journal of Cleaner Production, 161, 1-11. 114 Carter, M. R. (1984). Resource allocation and use under collective rights and labour management in peruvian coastal agriculture. The Economic Journal, 94(376), 826-846. Charnes, A., Cooper, W. W., & Rhodes, E. (1978). Measuring the efficiency of decision making units. European Journal of Operational Research, 2(6), 429-444. Chatalova L., & Balmann A. (2017). The hidden costs of renewables promotion: The case of crop-based biogas, Journal of Cleaner Production, 168 (12), 893-903. Chatalova, L., Müller, D., Valentinov, V., & Balmann, A. (2016). The Rise of the Food Risk Society and the Changing Nature of the Technological Treadmill. Sustainability, 8(6), 584. Chen, H., Yi, J., Chen, A., Peng, D., & Yang, J. (2023). Green technology innovation and CO2 emission in China: Evidence from a spatial-temporal analysis and a nonlinear spatial durbin model. Energy Policy, 172, 113338. Chhetri, N., Chaudhary, P., Tiwari, P. R., & Yadaw, R. B. (2012). Institutional and technological innovation: Understanding agricultural adaptation to climate change in Nepal. Applied Geography, 33, 142150. Coase R. H. (1937). The Nature of the Firm. Economica, 4(16), 386-405. Coase R. H. (1960). The Problem of Social Cost. Journal of Law and Economics, 25(3), 1-44. Cooper, W.W., Seiford, L.M., Tone, K. (2006). Introduction to Data Envelopment Analysis and Its Uses: With DEA-Solver Software and References. New York, Springer. Costa, J. C., Sousa, D. Z., Pereira, M. A., Stams, A. J. M., & Alves, M. M. (2013). Biomethanation potential of biological and other wastes. In Biofuel Technologies (pp. 369-396). Springer, Berlin, Heidelberg. Dahlman, C. J. (1979). Problem of externality. Journal of Law and Economics, 22(1), 141–162. D'Amato, D., Veijonaho, S., & Toppinen, A. (2020). Towards sustainability? Forest-based circular bioeconomy business models in Finnish SMEs. Forest policy and economics, 110, 101848. De Clercq, M., Vats, A., & Biel, A. (2018). Agriculture 4.0: The future of farming technology. Proceedings of the World Government Summit, Dubai, UAE, 11-13. 115 De Roest, K., Ferrari, P., & Knickel, K. (2018). Specialisation and economies of scale or diversification and economies of scope? Assessing different agricultural development pathways. Journal of Rural Studies, 59, 222-231. Deininger K. (1995). Collective agricultural production: A solution for transition economies? World Development, 23(8):, 1317-1334. Deininger K. (2013). Global land investments in the bio-economy: evidence and policy implications, Agricultural Economics, 44(s1), 115-127. Dierckx, M. A., & Stroeken, J. H. (1999). Information technology and innovation in small and mediumsized enterprises. Technological Forecasting and Social Change, 60(2), 149-166. Dietz, T., Börner, J., Förster, J. J., & Von Braun, J. (2018). Governance of the bioeconomy: A global comparative study of national bioeconomy strategies. Sustainability, 10(9), 3190. Dohse, D. (2000). Technology policy and the regions—the case of the BioRegio contest. Research Policy, 29(9), 1111-1133. Donbesuur, F., Ampong, G. O. A., Owusu-Yirenkyi, D., & Chu, I. (2020). Technological innovation, organizational innovation and international performance of SMEs: The moderating role of domestic institutional environment. Technological Forecasting and Social Change, 161, 120252. Dorocki, S. (2014). Spatial diversity of biotechnology centres in Germany. Quaestiones Geographicae, 33(2), 151-169. Du, K., & Li, J. (2019). Towards a green world: How do green technology innovations affect total-factor carbon productivity. Energy Policy, 131, 240-250. Dulá, J.H. (2007). Mining Nonparametric Frontiers. In: Joe Zhu, Wade D. Cook (eds.), Modeling Data Irregularities and Structural Complexities in Data Envelopment Analysis. Springer Science & Business Media; pp: 155-170. Duranton, G., & Puga, D. (2004). Micro-foundations of urban agglomeration economies. In Handbook of regional and urban economics (Vol. 4, pp. 2063-2117). Elsevier. 116 EC. (2012). Innovating for Sustainable Growth: A Bioeconomy for Europe; COM (2012) final; European Commission: Brussels, Belgium. EC. (2018a). A sustainable Bioeconomy for Europe: Strengthening the connection between economy, society and the environment: Updated Bioeconomy Strategy. European Commission (EC), Brussels. EC. (2018b). Outcome Report on the 2017 Bioeconomy Policy Day; European Commission: Seville, Spain, 2018. EC. (2019). Green Deal of the European Union, 2019, URL: https://ec.europa.eu/info/strategy/priorities2019-2024/european-green-deal_en. EC. (2019). The European Green Deal. Communication from the Commission to the European Parliament, the Council, the European Economic and Social Committee and the Committee of the Regions. Brussels, 11.12.2019, COM (2019) 640 final. EC. (2020a). Making Europe's businesses future-ready: A new Industrial Strategy for a globally competitive, green and digital Europe. European Commission. Press release IP/20/416. EC. (2020b). EU Biodiversity Strategy for 2030: Bringing nature back into our lives. Communication from the Commission to the European Parliament, the Council, the European Economic and Social Committee and the Committee of the Regions. Brussels, 20.5.2020, COM (2020) 380 final. EC. (2020c). Farm to Fork Strategy: For a fair, healthy and environmentally-friendly food system. European Commission, URL: https://ec.europa.eu/food/farm2fork_en. EC. (2019). Communication from the Commission to the European Parliament, the European Council, the Council, the European Economic and Social Committee and the Committee of the RegionsThe European Green Deal. European Commission. URL: https://eurlex.europa.eu/resource.html?uri=cellar:b828d165-1c22-11ea8c1f01aa75ed71a1.0002.02/DOC_1&format=PDF. EC. (2020d). Farm to Fork Strategy Action Plan. URL: https://ec.europa.eu/food/sites/food/files/safety/docs/f2f_action-plan_2020_strategyinfo_en.pdf. 117 Efken, J., Dirksmeyer, W., Kreins, P., & Knecht, M. (2016). Measuring the importance of the bioeconomy in germany: Concept and illustration. NJAS - Wageningen Journal of Life Sciences, 77, 9-17. Eggertsson, T. (1990). Economic Behavior and Institutions. Cambridge, Cambridge University Press. Ehrenfeld, W., & Kropfhäußer, F. (2017). Plant-based bioeconomy in Central Germany–a mapping of actors, industries and places. Technology Analysis & Strategic Management, 29(5), 514-527. El-Chichakli, B., von Braun, J., Lang, C., Barben, D., & Philp, J. (2016). Policy: Five cornerstones of a global bioeconomy. Nature News, 535(7611), 221. Emerick, K., de Janvry, A., Sadoulet, E., & Dar, M. H. (2016). Technological innovations, downside risk, and the modernization of agriculture. American Economic Review, 106(6), 1537-61. Erdoğan, S., Yıldırım, S., Yıldırım, D. Ç., & Gedikli, A. (2020). The effects of innovation on sectoral carbon emissions: Evidence from G20 countries. Journal of Environmental Management, 267, 110637. Erjavec, K., & Erjavec, E. (2015). ‘Greening the CAP’–Just a fashionable justification? A discourse analysis of the 2014–2020 CAP reform documents. Food Policy, 51, 53-62. Ernst & Young. (2011). German Biotechnology Report 2011 from Ernst & Young. URL: https://analyticalscience.wiley.com/content/news-do/deutscher-biotechnologie-report-2011-vonernst-amp-young. Ersoy, E., & Ugurlu, A. (2020). The potential of Turkey's province-based livestock sector to mitigate GHG emissions through biogas production. Journal of Environmental Management, 255, 109858. EU. 2022. Forest-based bioeconomy for climate change mitigation. European Commission's Knowledge Centre for Bioeconomy. URL: https://knowledge4policy.ec.europa.eu/bioeconomy/topic/forestbioeconomy-cc-mitigation_en. Eurostat. 2020, URL: https://ec.europa.eu/eurostat/databrowser/ view/ef_lus_main/default /table?lang=en. FAO. (2020). Emissions due to agriculture. Global, regional and country trends 2000–2018. FAOSTAT Analytical Brief Series No 18. Rome. FAO. (2021). URL: https://unstats.un.org/sdgs/metadata/files/Metadata-15-03-01.pdf. 118 Feiz, R., Johansson, M., Lindkvist, E., Moestedt, J., Påledal, S. N., & Svensson, N. (2020). Key performance indicators for biogas production—methodological insights on the life-cycle analysis of biogas production from source-separated food waste. Energy, 200, 117462. Fernandez-Cornejo, J., Gempesaw, C. M., Elterich, J. G., & Stefanou, S. E. (1992). Dynamic measures of scope and scale economies: an application to German agriculture. American Journal of Agricultural Economics, 74(2), 329-342. Fleiter, T., Fehrenbach, D., Worrell, E., & Eichhammer, W. (2012). Energy efficiency in the German pulp and paper industry–A model-based assessment of saving potentials. Energy, 40(1), 84-99. FMEACA (Federal Ministry for Economic Affairs and Climate Action). (2022). Biotechnology Clusters in Germany, German Trade and Invest, Berlin, Germany, URL: https://www.gtai.de/resource/blob/64096/67cce1d0ca00742fd6a8d8c41e285f64/20221013_FS_Bi otechnology_WEB.pdf. FNR (Fachagentur Nachwachsende Rohstoffe e.V), Bioenergy in Germany facts and figures (2020). URL: http: https://www.fnr.de/fileadmin/allgemein/pdf/broschueren/broschuere_basisdaten_bioenergie_2020 _engl_web.pdf Fong, W. K., Matsumoto, H., & Lun, Y. F. (2009). Application of System Dynamics model as decision making tool in urban planning process toward stabilizing carbon dioxide emissions from cities. Building and Environment, 44(7), 1528-1537. Fornahl, D., Broekel, T., & Boschma, R. (2011). What drives patent performance of German biotech firms? The impact of R&D subsidies, knowledge networks and their location. Papers in Regional Science, 90(2), 395-419. Forrester, J. W. (1987). Lessons from system dynamics modeling. System Dynamics Review, 3(2), 136149. Forrester, J. W. (1970). Urban dynamics. IMR; Industrial Management Review (pre-1986), 11(3), 67. 125 Khan, I., Zhong, R., Khan, H., Dong, Y., & Nuţă, F. M. (2023). Examining the relationship between technological innovation, economic growth and carbon dioxide emission: Dynamic panel data evidence. Environment, Development and Sustainability, 1-20. Klärle, M. (2018). Landmanagement 4.0 – Mehrwert durch künstliche Intelligenz. Landtechnik, 73(2): 3738. Kumar, D., & Singh, V. (2020). Bioconversion of Processing Waste from Agro‐Food Industries to Bioethanol: Creating a Sustainable and Circular Economy. Waste Valorisation: Waste Streams in a Circular Economy, 161-181. Kumeh, E. M., Kyereh, B., Birkenberg, A., & Birner, R. (2021). Customary power, farmer strategies and the dynamics of access to protected forestlands for farming: Implications for Ghana's forest bioeconomy. Forest Policy and Economics, 133, 102597. Kuosmanen, T., Kuosmanen, N., El-Meligi, A., Ronzon, T., Gurria, P., Iost, S., & M’Barek, R. (2020). How big is the bioeconomy. Reflections from an economic perspective, 49. Ladu, L., Imbert, E., Quitzow, R., & Morone, P. (2020). The role of the policy mix in the transition toward a circular forest bioeconomy. Forest Policy and Economics, 110, 101937. Lakes, T., Garcia-Marquez, J., Müller, D., Lakner, S., & Pe’er, G. (2020). How green is greening? A finescale analysis of spatio-temporal dynamics in Germany (No. 17 (2020)). FORLand-Working Paper. Lebuhn, M., Munk, B., & Effenberger, M. (2014). Agricultural biogas production in Germany-from practice to microbiology basics. Energy, Sustainability and Society, 4(1), 1-21. Lee, C. C., & Lee, C. C. (2022). How does green finance affect green total factor productivity? Evidence from China. Energy Economics, 107, 105863. Lee, K., & Malerba, F. (2017). Catch-up cycles and changes in industrial leadership: Windows of opportunity and responses of firms and countries in the evolution of sectoral systems. Research Policy, 46(2), 338-351. 126 Leopoldina (2012). Bioenergy – Chances and Limits. Statement of the National Academy of Sciences Leopoldina, Halle (Saale). Leopoldina (2020). Global Biodiversity in Crisis – What can Germany and the EU do about this? Discussion paper 24. The National Academy of Sciences Leopoldina, Halle (Saale). LeSage, J. P., & Pace, R. K. (2009). Spatial econometric models. In Handbook of applied spatial analysis: Software tools, methods and applications (pp. 355-376). Berlin, Heidelberg: Springer Berlin Heidelberg. Lewandowski, I. (2015). Securing a sustainable biomass supply in a growing bioeconomy. Global Food Security, 6, 34-42. Li, H., Zhang, M., Li, C., & Li, M. (2019). Study on the spatial correlation structure and synergistic governance development of the haze emission in China. Environmental Science and Pollution Research, 26, 12136-12149. Li, Z., Wang, F., Kang, T., Wang, C., Chen, X., Miao, Z., ... & Zhang, H. (2022). Exploring differentiated impacts of socioeconomic factors and urban forms on city-level CO2 emissions in China: Spatial heterogeneity and varying importance levels. Sustainable Cities and Society, 84, 104028. Lin, J. Y., & Monga, C. (2011). Growth identification and facilitation: the role of the state in the dynamics of structural change. World Bank policy research working paper, (5313). Lindner, M., Hanewinkel, M., & Nabuurs, G. J. (2017). How can a forest-based bioeconomy contribute to climate change adaptation and mitigation?. In Towards a sustainable European forest-based bioeconomy (No. 8, pp. 77-85). European Forest Institute. Liobikiene, G., Chen, X., Streimikiene, D., & Balezentis, T. (2020). The trends in bioeconomy development in the European Union: Exploiting capacity and productivity measures based on the land footprint approach. Land Use Policy, 91, 104375. Liu, Z., Deng, Z., Davis, S. J., & Ciais, P. (2024). Global carbon emissions in 2023. Nature Reviews Earth & Environment, 5(4), 253-254. 127 Lovrak, A., Pukšec, T., & Duić, N. (2020). A Geographical Information System (GIS) based approach for assessing the spatial distribution and seasonal variation of biogas production potential from agricultural residues and municipal biowaste. Applied Energy, 267, 115010. Lovrić, N., Lovrić, M., & Mavsar, R. (2020). Factors behind development of innovations in European forest-based bioeconomy. Forest Policy and Economics, 111, 102079. Luhas, J., Mikkilä, M., Kylkilahti, E., Miettinen, J., Malkamäki, A., Pätäri, S., ... & Toppinen, A. (2021). Pathways to a forest-based bioeconomy in 2060 within policy targets on climate change mitigation and biodiversity protection. Forest Policy and Economics, 131, 102551. Luo, Y., Lu, Z., Salman, M., & Song, S. (2022). Impacts of heterogenous technological innovations on green productivity: An empirical study from 261 cities in China. Journal of Cleaner Production, 334, 130241. Maraut, S., & Martínez, C. (2014). Identifying author–inventors from Spain: methods and a first insight into results. Scientometrics, 101(1), 445-476. Maraut, S., Dernis, H., Webb, C., Spiezia, V., & Guellec, D. (2008). The OECD REGPAT database: a presentation. Masten, S.E. (2000). Transaction-cost economics and the organization of agricultural transactions." Industrial organization. Emerald Group Publishing Limited, 173-195. McCann, L., Colby, B., Easter, K.W., Kasterine, A., Kuperan, K.V. (2005). Transaction cost measurement for evaluating environmental policies. Ecological Economics, 52, 527-542. McCann, L. M. (2009). Transaction costs of environmental policies and returns to scale: The case of comprehensive nutrient management plans. Applied Economic Perspectives and Policy, 31(3), 561-573. McCormick, K., & Kautto, N. (2013). The bioeconomy in Europe: An overview. Sustainability, 5(6), 25892608. McSherry, M. E., & Ritchie, M. E. (2013). Effects of grazing on grassland soil carbon: A global review. Global Change Biology, 19(5), 1347-1357. 128 Mehmood, S., Zaman, K., Khan, S., & Ali, Z. (2024). The role of green industrial transformation in mitigating carbon emissions: Exploring the channels of technological innovation and environmental regulation. Energy and Built Environment, 5(3), 464-479. Mennicken, L., Janz, A., & Roth, S. (2016). The German R&D program for CO2 utilization—innovations for a green economy. Environmental Science and Pollution Research, 23(11), 11386-11392. Meyer-Aurich, A., Schattauer, A., Hellebrand, H. J., Klauss, H., Plöchl, M., & Berg, W. (2012). Impact of uncertainties on greenhouse gas mitigation potential of biogas production from agricultural resources. Renewable Energy, 37(1), 277-284. Mohmmed, A., Li, Z., Arowolo, A. O., Su, H., Deng, X., Najmuddin, O., & Zhang, Y. (2019). Driving factors of CO2 emissions and nexus with economic growth, development and human health in the Top Ten emitting countries. Resources, Conservation and Recycling, 148, 157-169. Nakićenović, N. (2000). Greenhouse gas emissions scenarios. Technological Forecasting and Social Change, 65(2), 149-166. North, D. C., & Wallis, J. J. (1994). Integrating institutional change and technical change in economic history a transaction cost approach. Journal of Institutional and Theoretical Economics. 150(4), 609-624. Nwaka, I. D., Nwogu, M. U., Uma, K. E., & Ike, G. N. (2020). Agricultural production and CO2 emissions from two sources in the ECOWAS region: New insights from quantile regression and decomposition analysis. Science of The Total Environment, 748, 141329. OECD (Organization for Economic Co-operation and Development). (2016). Key biotechnology indicators Updated O. 2016. http://www.oecd.org/sti/biotech/keybiotechnologyindicators.htm. OECD (Organization for Economic Cooperation and Development). (2011a). Farm Management Practices to Foster Green Growth. Paris: OECD. OECD (Organization for Economic Cooperation and Development). (2011b). Towards Green Growth. Paris: OECD. 129 Ottoy, S., Angileri, V., Gibert, C., Paracchini, M. L., Pointereau, P., Terres, J. M., ... & Dicks, L. V. (2018). Impacts of selected Ecological Focus Area options in European farmed landscapes on climate regulation and pollination services: A systematic map protocol. Environmental Evidence, 7(1), 110. Pahle, M., Tietjen, O., Osorio, S., Egli, F., Steffen, B., Schmidt, T. S., & Edenhofer, O. (2022). Safeguarding the energy transition against political backlash to carbon markets. Nature Energy, 7(3), 290-296. Pan, X., Li, M., Wang, M., Zong, T., & Song, M. (2020). The effects of a Smart Logistics policy on carbon emissions in China: A difference-in-differences analysis. Transportation Research Part E: Logistics and Transportation Review, 137, 101939. Pant, D., Misra, S., Nizami, A. S., Rehan, M., van Leeuwen, R., Tabacchioni, S., ... & Elst, K. (2019). Towards the development of a biobased economy in Europe and India. Critical Reviews in Biotechnology, 39(6), 779-799. Pataki, D. E., Alig, R. J., Fung, A. S., Golubiewski, N. E., Kennedy, C. A., McPherson, E. G., ... & Romero Lankao, P. (2006). Urban ecosystems and the North American carbon cycle. Global Change Biology, 12(11), 2092-2102. Patermann, C., & Aguilar, A. (2018). The origins of the bioeconomy in the European Union. New Biotechnology, 40, 20-24. Pe'Er, G., Dicks, L. V., Visconti, P., Arlettaz, R., Báldi, A., Benton, T. G., ... & Scott, A. V. (2014). EU agricultural reform fails on biodiversity. Science, 344(6188), 1090-1092. Pe'Er, G., Zinngrebe, Y., Hauck, J., Schindler, S., Dittrich, A., Zingg, S., ... & Lakner, S. (2017). Adding some green to the greening: improving the EU's Ecological Focus Areas for biodiversity and farmers. Conservation letters, 10(5), 517-530. Perrin, R., & Winkelmann, D. (1976). Impediments to technical progress on small versus large farms. American Journal of Agricultural Economics, 58(5), 888-894. 130 Petrick, M., & Kloss, M. (2018). Identifying agricultural factor productivity from micro-data: a review of approaches with an application to EU countries. German Journal of Agricultural Economics, 67(2), 67-69. Phan-huy, C., Göswein, V., & Habert, G. (2023). Climate-effective use of straw in the EU bioeconomy— comparing avoided and delayed emissions in the agricultural, energy and construction sectors. Environmental Research Letters, 18(12), 124004. Philippidis, G., Álvarez, R. X., Di Lucia, L., Hermoso, H. G., Martinez, A. G., M'barek, R., ... & Verkerk, P. J. (2024). The development of bio-based industry in the European Union: A prospective integrated modelling assessment. Ecological Economics, 219, 108156. Piotrowski, S., Carus, M., & Carrez, D. (2016). European bioeconomy in figures. Industrial Biotechnology, 12(2), 78-82. Popp, D., Juhl, T. P., & Johnson, D. K. (2003). Time in purgatory: Determinants of the grant lag for US patent applications. NBER Working Paper No. 9518. Popp, J., Kovács, S., Oláh, J., Divéki, Z., & Balázs, E. (2021). Bioeconomy: Biomass and biomass-based energy supply and demand. New Biotechnology, 60, 76-84. Post, W. M., & Kwon, K. C. (2000). Soil carbon sequestration and land‐use change: processes and potential. Global Change Biology, 6(3), 317-327. Pradhan, R. P., Arvin, M. B., & Bahmani, S. (2018). Are innovation and financial development causative factors in economic growth? Evidence from a panel granger causality test. Technological Forecasting and Social Change, 132, 130-142. Purkus, A., Hagemann, N., Bedtke, N., & Gawel, E. (2018). Towards a sustainable innovation system for the German wood-based bioeconomy: Implications for policy design. Journal of Cleaner Production, 172, 3955-3968. 131 Rebolledo-Leiva, R., Moreira, M. T., & González-García, S. (2023). Progress of social assessment in the framework of bioeconomy under a life cycle perspective. Renewable and Sustainable Energy Reviews, 175, 113162. Riordan, M. H., & Williamson, O. E. (1985). Asset specificity and economic organization. International Journal of Industrial Organization, 3(4), 365-378. Rivera León, L., Bougas, K., Aggestam, F., Pülzl, H., Zoboli, E., Ravet, J., Griniece, E., Vermeer, J., Maroulis, N., Ettwein, F., Van Brusselenm J. & Green, T. 2016. An assessment of the cumulative cost impact of specified EU legislation and policies on the EU forest-based industries. Brussels: DG GROW. Rørstad, P. K., Vatn, A., & Kvakkestad, V. (2007). Why do transaction costs of agricultural policies vary?. Agricultural Economics, 36(1), 1-11. Scarlat, N., Dallemand, J. F., & Fahl, F. (2018). Biogas: Developments and perspectives in Europe. Renewable Energy, 129, 457-472. Scarlat, N., Dallemand, J. F., Monforti-Ferrario, F., & Nita, V. (2015). The role of biomass and bioenergy in a future bioeconomy: Policies and facts. Environmental Development, 15, 3-34. Schäfer, A. (2014). Technological change, population dynamics, and natural resource depletion. Mathematical Social Sciences, 71, 122-136. Scheftelowitz, M., & Thrän, D. (2016). Unlocking the energy potential of manure—an assessment of the biogas production potential at the farm level in Germany. Agriculture, 6(2), 20. Schulp, C. J., Nabuurs, G. J., & Verburg, P. H. (2008). Future carbon sequestration in Europe—effects of land use change. Agriculture, Ecosystems & Environment, 127(3-4), 251-264. Schumpeter, J. A. (1935). The analysis of economic change. The review of Economics and Statistics, 17(4), 2-10. Schütte, G. (2018). What kind of innovation policy does the bioeconomy need?. New Biotechnology, 40, 82-86. 132 Seppälä, J., Heinonen, T., Pukkala, T., Kilpeläinen, A., Mattila, T., Myllyviita, T., ... & Peltola, H. (2019). Effect of increased wood harvesting and utilization on required greenhouse gas displacement factors of wood-based products and fuels. Journal of Environmental Management, 247, 580-587. Sexton, R. J., (1986). Perspectives on the development of the economic theory of cooperatives. Canadian Journal of Agricultural Economics. 32, 423-436. Sharma, G. K., Khan, S. A., Shrivastava, M., Bhattacharyya, R., Sharma, A., Gupta, D. K., ... & Gupta, N. (2021). Circular economy fertilization: Phycoremediated algal biomass as biofertilizers for sustainable crop production. Journal of Environmental Management, 287, 112295. Sherwood, J. (2020). The significance of biomass in a circular economy. Bioresource Technology, 300, 122755. Sohag, K., Begum, R.A., Abdullah, S.M.S., Jaafar, M., 2015. Dynamics of energy use, technological innovation, economic growth and trade openness in Malaysia. Energy, 90, 1497-1507. Song, M., Peng, L., Shang, Y., & Zhao, X. (2022). Green technology progress and total factor productivity of resource-based enterprises: A perspective of technical compensation of environmental regulation. Technological Forecasting and Social Change, 174, 121276. Stoll, H R, & Whaley, R E. (1983). Transaction costs and the small firm effect. Journal of Financial Economics, 12(1), 57-79. Subramanian, A., & Qaim, M. (2009). Village-wide effects of agricultural biotechnology: the case of Bt cotton in India. World Development, 37(1), 256-267. Tayeh, H. N. A., Azaizeh, H., & Gerchman, Y. (2020). Circular economy in olive oil production–olive mill solid waste to ethanol and heavy metal sorbent using microwave pretreatment. Waste Management, 113, 321-328. Theuerl, S., Herrmann, C., Heiermann, M., Grundmann, P., Landwehr, N., Kreidenweis, U., & Prochnow, A. (2019). The future agricultural biogas plant in Germany: A vision. Energies, 12(3), 396. 133 Tian, Y., Jiang, G., Zhou, D., Ding, K., Su, S., Zhou, T., & Chen, D. (2019). Regional industrial transfer in the Jingjinji urban agglomeration, China: An analysis based on a new “transferring areaundertaking area-dynamic process” model. Journal of Cleaner Production, 235, 751-766. Tian, Y., Zhang, J., & He, Y. (2014). Research on spatial-temporal characteristics and driving factor of agricultural carbon emissions in China. Journal of Integrative Agriculture, 13(6), 1393-1403. Tiwari, T., Kaur, G. A., Singh, P. K., Balayan, S., Mishra, A., & Tiwari, A. (2024). Emerging bio-capture strategies for greenhouse gas reduction: Navigating challenges towards carbon neutrality. Science of The Total Environment, 172433. Troost C , Walter T , & Berger T . (2015). Climate, energy and environmental policies in agriculture: Simulating likely farmer responses in Southwest Germany. Land Use Policy, 46, 50-64. Ugur, M., Trushin, E., Solomon, E., & Guidi, F. (2016). R&D and productivity in OECD firms and industries: A hierarchical meta-regression analysis. Research Policy, 45(10), 2069-2086. Umweltbundesamt, 2019, URL: https://www.umweltbundesamt.de/daten/klima/treibhausgas-emissionenin-deutschland/kohlendioxid-emissionen#kohlendioxid-emissionen-im-vergleich-zu-anderentreibhausgasen. Van Lancker, J., Wauters, E., & Van Huylenbroeck, G. (2016). Managing innovation in the bioeconomy: An open innovation perspective. Biomass and Bioenergy, 90, 60-69. Vatn, A. (2001). Transaction Costs and Multifunctionality. Proceedings, OECD Workshop on Multifunctionality by the Directorate for Food Agriculture, and Fisheries, Paris, France. Vlaisavljevic, V., Medina, C. C., & Van Looy, B. (2020). The role of policies and the contribution of cluster agency in the development of biotech open innovation ecosystem. Technological Forecasting and Social Change, 155, 119987. Wallis, J. & North, D. (1986). Measuring the Transaction Sector in the American Economy: 1870-1970. In: Stanley L. Engerman and Robert E. Gallman, (eds.), Long-Term Factors in American Economic Growth. Chicago: University of Chicago Press, 95-162. 134 Wang, J., Rothausen, S. G., Conway, D., Zhang, L., Xiong, W., Holman, I. P., & Li, Y. (2012). China’s water–energy nexus: greenhouse-gas emissions from groundwater use for agriculture. Environmental Research Letters, 7(1), 014035. Wang, J., Wang, K., Shi, X., & Wei, Y. M. (2019). Spatial heterogeneity and driving forces of environmental productivity growth in China: would it help to switch pollutant discharge fees to environmental taxes?. Journal of Cleaner Production, 223, 36-44. Wang, L., Vo, X. V., Shahbaz, M., & Ak, A. (2020). Globalization and carbon emissions: Is there any role of agriculture value-added, financial development, and natural resource rent in the aftermath of COP21?. Journal of Environmental Management, 268, 110712. Wang, R. (2023). The modernization of harmonious coexistence between humanity and nature: Historical achievement, contradictory challenges and realization paths. Journal of Management World, 39(03), 19-30. (in Chinese) Wang, S., Chen, B., Sun, Z., Long, X., Xue, M., Yu, H., ... & Wang, Y. (2023). The booming non-food bioeconomy drives large share of global land-use emissions. Global Environmental Change, 83, 102760. Wang, Y., Yan, W., Ma, D., & Zhang, C. (2018). Carbon emissions and optimal scale of China's manufacturing agglomeration under heterogeneous environmental regulation. Journal of Cleaner Production, 176, 140-150. Weber, G., & Cabras, I. (2017). The transition of Germany's energy production, green economy, low-carbon economy, socio-environmental conflicts, and equitable society. Journal of Cleaner Production, 167, 1222-1231. Weiser, C., Zeller, V., Reinicke, F., Wagner, B., Majer, S., Vetter, A., & Thraen, D. (2014). Integrated assessment of sustainable cereal straw potential and different straw-based energy applications in Germany. Applied Energy, 114, 749-762.