Do Forestry Innovation Ecosystems Contribute to the Green Economy?—Based on SBM-GML and Two-Way Fixed Effect Models

Abstract

Although the importance of forestry has been widely recognized, research on the impact of forestry innovation ecosystems on green economic development is still in its infancy, and existing research is limited. This paper focuses on 31 provinces in China and conducts an empirical analysis using data from 2012 to 2021. It provides empirical insights on how forestry innovation ecosystems affect green economic growth, thus addressing the gap of existing research. The findings are as follows: (1) Forestry innovation ecosystems have a significant positive effect on green total factor productivity. (2) Mechanism analysis reveals a significant positive effect of forestry innovation ecosystems on the technical efficiency index, but no significant effect on technical progress. (3) Heterogeneity analysis indicates that in eastern and central regions, forestry innovation ecosystems do not significantly impact green total factor productivity or the paths of technical efficiency and technical progress. However, in western regions, forestry innovation ecosystems have a significant effect of green total factor productivity and technical efficiency, while technical progress remains unaffected. In conclusion, the forestry innovation ecosystems play a significant role in promoting green economic development, particularly in the western regions of China. The potential contributions of this research are twofold: (1) This study addresses this gap by providing empirical insights on how forestry innovation ecosystems influence green economic growth. (2) This paper also investigates the mechanisms by which forestry innovation ecosystems impact green development through various types of innovation, offering practical references for stakeholders involved in forestry and sustainable development. Differentiated policies should be assigned to address regional disparities, with increased support for forestry innovation in the west and the encouragement of green technology investments in the east and center. Additionally, promoting technical progress through policy innovation, international cooperation, and enhanced intellectual property protection is essential to maximizing the benefits of forestry innovation ecosystems.

1. Introduction

In the context of globalization, the synergistic relationship between economic development and ecological environmental protection has become an extremely urgent need and an unstoppable trend [1]. Globally, climate change is intensifying at an alarming rate, leading to increasingly frequent extreme weather events. The problem of resource scarcity is becoming increasingly acute, with many important natural resources facing the crisis of depletion; furthermore, ecosystem degradation is worsening, with a sharp decline in biodiversity and a serious disruption of the ecological balance [2]. These intertwined problems constitute a great threat to the sustainable development of human society. In response, the whole human society is urgently seeking for a sustainable development model that balances economic growth with effective ecological environmental protection and restoration.

Green economic development emphasizes reducing resource consumption and environmental pollution while enhancing eco-efficiency, all within the context of fostering economic growth. This approach aims to achieve the coordinated development of the economy, society, and environment. Among the indicators for measuring green economic development, green total factor productivity stands out as it accounts for traditional factor inputs such as capital and labor, as well as factors like energy consumption and environmental pollution. Consequently, green total factor productivity offers a more comprehensive and accurate assessment for the quality and sustainability of green economic development [3].

As one of the most important ecosystems on earth, forestry not only provides human beings with material resources such as timber and forest products, but forests also plays an irreplaceable role in regulating the climate, maintaining soil and water, and preserving biodiversity [4]. However, the traditional forestry development model has long prioritized timber extraction and economic benefits, often at the expense of ecological protection and sustainable resource management. This approach has led to the overexploitation of forest resources and the disruption of the ecological balance [5].

The Forestry Innovation Ecosystem, as a crucial element and the driving force for green economic development, has garnered significant attention from academics and policymakers. Its potential in resource utilization, ecological services enhancement, and industrial innovation is noteworthy. An innovation ecosystem refers to a cohesive system formed by the interaction of various innovation entities pursuing shared goals within a specific regional environment [6]. Similar to an agricultural innovation ecosystem, forestry innovation ecosystems describe how various forestry innovation entities interact with materials, energy, and information in a specific innovation environment. This interaction fosters the invention and creation of new knowledge, varieties, and technologies, which are then applied and disseminated throughout the forestry production process [7]. China has rich forest resources, but the spatial distribution is not uniform. The relatively young stand age for Chinese forests is mostly due to the large proportion of newly planted forests (0–40 years old), which are more prevailing in south China. Older forests (stand age > 60 years old) are more frequently found in the east Qinghai–Tibetan Plateau and the central mountain areas of west and northeast China, where human activities are less intensive [8,9].

Additionally, China’s commitment to the “dual-carbon” goal and the pursuit of high-quality development have highlighted the importance of innovative ecosystems as a new engine for advancing the green development of forestry. China’s dual carbon goal means that China’s carbon emissions will peak before 2030 and achieve carbon neutrality by 2060 [10].

This paper posits that the impact of the forestry innovation system on green total factor productivity may be operated through four mechanisms: (1) Technological Innovation: Innovations in precision planting, intelligent irrigation, and monitoring technologies allow for the real-time monitoring of forest health and resource changes [11,12]. These technologies provide a basis for scientific management and decision making, thereby enhancing the efficiency of forestry production and management, directly boosting green total factor productivity. (2) Market Innovation: market innovation facilitates the balance of supply and demand for green products and services through innovative market mechanisms, thereby improving resource utilization efficiency and environmental protection [13]. (3) Institutional Innovation: Institutional innovation is a crucial component of the forestry innovation ecosystems. Improving management systems provides institutional support for the green development of forestry. Forest management models based on the value of ecosystem services emphasize the ecological functions of forests [14]. (4) Management Innovation: Management innovation improves resource utilization efficiency by implementing refined management and scientific planning [15]. This is achieved by introducing advanced management tools and methods, enabling more scientific and reasonable management decisions. In addition, management innovation promotes integrated management models that organically combine forestry production, ecological protection, and social development. However, research on forestry innovation ecosystems and their impact on green economic development is still in its infancy, with relatively few studies available. Numerous issues remain unexplored and require further investigation.

Although there has been considerable research on the synergistic relationship between green economic development and ecological protection, studies specifically addressing the role of forestry innovation ecosystems in influencing green total factor productivity remain scarce. The existing literature primarily focuses on macro-level green technologies and policy innovations in fostering economic growth, often overlooking the critical role that forestry ecosystems play in advancing the green transition. This study aims to fill this research gap by investigating how forestry innovation ecosystems, through multiple pathways such as technological innovation, market mechanisms, institutional innovation, and management innovation, enhance green total factor productivity. Furthermore, it explores the regional heterogeneity in these effects across China’s eastern, central, and western regions. By conducting an empirical analysis using data from 31 Chinese provinces between 2012 and 2021, this paper not only reveals the contribution of forestry innovation ecosystems to green economic growth, but also provides valuable insights into the formulation of regionally tailored policies that can support sustainable forestry development.

The potential contributions of this paper are as follows: (1) As an integral part of the ecosystem and large-scale agriculture, the importance of the forestry system in green economic development is self-evident. However, research specifically focused on forestry ecological innovation systems remains relatively scarce, and this study aims to fill this gap. (2) This paper discusses the underlying mechanisms by which the forestry ecosystem impacts green economic development from various perspectives, including technological innovation, market innovation, institutional innovation, and management innovation. Through empirical analysis, this paper aims to provide valuable references for policymakers, forestry practitioners, and farmers to support the green development of forestry.

The structure of this paper is arranged as follows: the second section is the literature review, the third section covers the theoretical analysis and research hypotheses, the fourth section discusses the data, the fifth section presents the empirical results analysis, and the sixth section provides conclusions and policy recommendations.

2. Literature Review

2.1. Innovation Ecosystems in Forestry

Research on innovation ecosystems originated from the ecosystem theory. In the 1990s, with the rise of the information technology revolution, business ecosystems took shape. Moore [16] was the first to integrate ecological thinking with business systems and proposed the concept of a business ecosystem; Adner [17] was the first to introduce the concept of an innovation ecosystem as a coherent, customer-facing solution composed of a company’s products through collaborative arrangements; Walrave et al. [18] defined innovation ecosystems as networks with interdependent subjects co-creating value through complementary resources or capabilities; and Granstrand et al. [6] defined innovation ecosystems as an evolutionary collection of subjects, behaviors, institutions, and other elements. Over time, innovation ecosystems have continued to evolve and develop, becoming a new innovation paradigm. Their hierarchical structure includes macro, meso, and micro levels, i.e., a national innovation ecosystem, industrial innovation ecosystem, and enterprise innovation ecosystem [19].

Since the 1980s, agricultural innovation management has shifted from the traditional “input-output” or “upstream-midstream-downstream” single-linear innovation paradigm to the systemic innovation paradigm of a chain–loop interaction. Influenced by the concepts of a national innovation system and regional innovation system, an agricultural innovation system has emerged from early systems like the agricultural research system and agricultural knowledge information system, and has since become an important research paradigm of agricultural innovation management [20]. Scholars mainly focused on the concepts [11], structure [21], and mechanisms [22] of the agricultural innovation system, resulting in significant practical applications [23]. A few scholars have attempted to introduce the innovation ecosystem paradigm into the field of agricultural innovation management research, and have conducted some preliminary explorations of agricultural innovation ecosystems [24]. However, the breadth and depth of the research need to be further expanded.

As an integral part of agriculture, forestry innovation ecosystems are closely related to agricultural innovation systems while exhibiting unique characteristics in key areas. The forestry innovation ecosystems are a complex and dynamic entity comprising various interconnected elements and actors within the forestry sector. This system encompasses not only the forest resource itself, but also innovations in technology, management, institutions, and markets [7].

From a structural perspective, the forestry innovation ecosystems include diverse actors such as forestry enterprises, scientific research institutions, government departments, financial institutions, and forest farmers [25]. Forestry enterprises, as the core drivers of innovation, advance the industry by developing new technologies and launching new products. Scientific research institutions play a critical role in knowledge innovation and technical support, providing the theoretical foundations and cutting-edge technologies for forestry innovation. Government agencies facilitate a conducive innovation environment through relevant policies and regulations, guiding and standardizing forestry innovation activities [26]. Financial institutions offer the necessary financial support to ensure the smooth implementation of innovation projects. Forest farmers, as direct participants in forestry production, contribute to forestry innovation through their practical experiences and feedback on their needs [8].

In terms of operational mechanisms, the forestry innovation ecosystems are characterized by synergy and interactions. Various actors form a close cooperation network through information exchange, resource sharing, and collaborative research and development. Technological innovations are continuously disseminated and diffused within this network, promoting the transformation and application of innovative outcomes. Additionally, market demand acts as a crucial driver for innovation, compelling each actor to adjust their innovation strategies to meet the diverse demands of consumers for forestry products and services [15].

2.2. Green Development of the Economy

Green development represents a significant theoretical advancement in humanity’s pursuit of sustainable development, systematically integrating the concepts of green growth, green welfare, and green wealth. Its aim is to achieve the coordinated development of resources and the environment alongside socio-economic activities, thereby maximizing the comprehensive benefits of the economy, society, and ecology [27]. China’s high-quality economic development emphasizes innovation, coordination, greenness, openness, and sharing, with green development being a crucial component. This focus is evident in the harmonious growth of the economy and the protection of ecological and environmental resources. High-quality economic development requires not only efficient economic operations, but also robust sustainability. In the context of the “dual-carbon” goal, advancing the transformation and upgrading of China’s green economic development significantly contributes to the high-quality development of the nation’s economy [28].

In terms of the concepts of green development, it is regarded as a comprehensive development concept aimed at realizing the benign interaction and organic integration of economic growth and environmental protection [29]. This approach covers the reduction of resource consumption and environmental pollution, while also emphasizing the restoration and protection of ecosystems. It advocates for the optimization, transformation and upgrading of the economic structure to achieve economic, social, and environmental sustainability [30]. As far as the connotation of green development is concerned, green development is a sustainable development model that protects the environment and promotes social progress alongside economic growth. Combining the concept of green development with regional innovation and coordinated development is crucial for high-quality growth and the sustainable development of the regional economy. The primary objectives of the ecological civilization construction are to conserve resources, protect the environment, and preserve the ecology [31].

In terms of factors affecting the green development of the economy: Technological innovation is a key driver, with the continuous emergence and wide application of advanced green technologies, such as clean energy, energy-saving and emission reduction, and resource recycling and reuse technology, significantly improving production efficiency and reducing environmental costs [32]. The growing environmental awareness among consumers have led to an increasing demand for green products and services, prompting enterprises to adjust their production strategies and increase investment in green innovation [33]. In addition, policies and regulations play a guiding and motivating role. Governments have created a favorable policy environment for green economic development by setting strict environmental standards, implementing green tax policies and providing green subsidies and incentives [34]; at the same time, international cooperation is becoming increasingly important in promoting green economic development. Countries are working together to promote the green transformation of the global economy by jointly addressing climate change, improving the quality of their products and services, and sharing green technologies and experiences [35].

In terms of the impact of green economic development, at the macro level, green economic development enhances the country’s overall competitiveness, attracts green investment, and promotes the transition of the industrial structure to high-end, intelligent, and green practices. At the meso level, green economic development promotes industrial agglomeration and synergy, forming green industrial chains and industrial clusters that improve the overall resource utilization efficiency and environmental performance of industries [36]. At the micro level, enterprises can reduce operating costs, increase product added value, enhance market competitiveness, and establish a good corporate image by implementing green development strategies [37]. However, the green development of the economy also faces many challenges. The R&D and application of green technologies encounter technical bottlenecks and high costs. Some enterprises may face operational pressure in the short term due to green transformation [38]. There is a regional imbalance in green development: economically less developed regions lack the capital, technology, and talents in green development. In addition, the formulation and implementation of green development policies need to be more scientifically grounded to avoid uncertainties and misalignments that could adversely affect the green development of the economy [39].

3. Theoretical Analysis

3.1. The Impact of Forestry Innovation Ecosystems on Green Total Factor Productivity

As one of the meso-level innovation ecosystems, the forestry innovation ecosystems play a significant role in forestry innovation and economic development. On one hand, the forestry innovation ecosystems integrate advanced technologies and management models to optimize resource allocation and improve the resource utilization efficiency. Precision forestry technology utilizes remote sensing and geographic information systems to monitor forest resources in real-time, optimizing harvesting and planting strategies to enhance productivity per unit area. These technologies can identify the most suitable areas for planting and the optimal times for harvesting, thereby maximizing the resource utilization efficiency and reducing unnecessary resource waste and environmental damage [40].

The forestry innovation ecosystems can also optimize forestry production processes through the application of environmental technologies and efficient management models, reducing resource waste and environmental pollution. For example, biological control technologies replace traditional chemical pesticides, reducing soil and water pollution while enhancing the quality and yield of forest products [26]. Environmentally friendly forestry machinery and production processes, such as low-emission logging machinery and green production lines, can reduce pollutant emissions, thereby protecting the ecological environment. Additionally, efficient management models, such as refined management and intelligent control, improve production efficiency and reduce operational costs [41].

On the other hand, research and innovation within the forestry innovation ecosystems are core drivers of technological advancements. By increasing investment in green technology research and development, research institutions and enterprises can develop more efficient and environmentally friendly forestry technologies and equipment. The application of information technologies, such as big data analytics and the Internet of Things (IoT) optimizes forestry management, improving production efficiency and resource utilization rates. Through research and innovation, the forestry innovation ecosystems continuously enhance the overall technological level, driving technical progress [40].

The forestry innovation ecosystems promote the dissemination and sharing of technological knowledge by establishing industry–academia–research collaboration platforms and technology exchange mechanisms. For instance, collaboration between research institutions and enterprises facilitates the development and commercialization of new technologies, allowing for rapid technology transfer and application. Events such as technology exchange meetings, seminars, and training sessions further support the dissemination and sharing of technological knowledge, enabling more enterprises and practitioners to master advanced technologies and management experiences. This widespread dissemination and application of knowledge enhances the industry’s overall technological level and innovation capacity [42].

Furthermore, the forestry innovation ecosystems create a synergistic innovation mechanism by integrating resources and encouraging collaborative innovation. By introducing new ideas and technologies from fields like information technology, biotechnology, and materials science, the forestry innovation ecosystems enhance R&D efficiency, reduce innovation costs, and accelerate the processes of technological innovation and application. This synergistic approach drives rapid technological advancements [43] and contributes to the realization of a circular economy [25,26]. Cellulose, wood fibers, and many new materials made from them have high recyclability potential.

Based on the above analysis, this paper proposes the following hypotheses:

H1: 

The forestry innovation ecosystems increase green total factor productivity, thereby promoting green economic development.

H2: 

The forestry innovation ecosystems improve technical efficiency, thereby promoting green economic development.

H3: 

The forestry innovation ecosystems promote technical progress, thereby promoting green economic development.

3.2. Mechanistic Analysis

As a complex system integrating various innovative elements, the forestry innovation ecosystems theoretically contribute to green total factor productivity in several ways.

3.2.1. Impact of Forestry Innovation Ecosystems on Technological Innovation

Technological innovation serves as the core driving force of the forestry innovation ecosystems, significantly improving the efficiency and effectiveness of forestry production. Specifically, new forestry planting technologies, such as precision planting and intelligent irrigation, allow for precise adjustments based on soil, climate, and tree growth conditions, resulting in a faster tree growth speed, enhanced quality, and increased timber yield per unit area. For instance, intelligent irrigation systems automatically adjust the irrigation levels in response to weather forecasts and soil moisture, ensuring optimal growth conditions for trees. Additionally, innovations in forestry monitoring technologies, such as satellite remote sensing and drones, improve the accuracy and timeliness of forest resource monitoring. These technologies provide the real-time monitoring of forest health and resource changes, supporting scientific management and decision making. As a result, these technological innovations have a direct and positive impact on green total factor productivity [11,12].

Based on the above analysis, this paper proposes the following hypothesis:

H4: 

The forestry innovation ecosystems increase green total factor productivity through technological innovation, thereby promoting green economic development.

3.2.2. Impact of Forestry Innovation Ecosystems on Market Innovation

Market innovation plays a driving role in balancing the supply and demand for green products and services through innovative business models and market mechanisms, thereby improving the resource utilization efficiency and environmental protection levels. For instance, the promotion of sustainable forest certification systems ensures the legality and sustainability of timber sources, enhancing the competitiveness of forestry products in the international market [25]. Additionally, developing the under-forest economy, such as planting medicinal herbs and breeding forest poultry, allows for a multi-level utilization of land and resources, thereby increasing the overall benefits of forestry. Moreover, introducing ecological compensation mechanisms incentivizes forest farmers and forestry enterprises to adopt environmental protection measures. These mechanisms encourage forest ecosystem protection and restoration, ultimately enhancing the health of forest ecosystems [13].

Based on the above analysis, this paper proposes the following hypothesis:

H5: 

The forestry innovation ecosystems increase green total factor productivity through market innovation, thereby promoting green economic development.

3.2.3. Impact of Forestry Innovation Ecosystems on Institutional Innovation

Institutional innovation, as a crucial component of the forestry innovation ecosystems, provides institutional support for the green development of forestry. Forest management models based on the value of ecosystem services go beyond timber production to the ecological functions of forests, such as water conservation, soil retention, and carbon storage. The application of digital management tools, such as forestry resource databases and management information systems, enhances management accuracy and timeliness, while also reducing costs. Furthermore, the application of management information systems helps digitalize and informatize forestry production and management, improving work efficiency and management levels [14].

Based on the above analysis, this paper proposes the following hypothesis:

H6: 

The forestry innovation ecosystems increase green total factor productivity through institutional innovation, thereby promoting green economic development.

3.2.4. Impact of Forestry Innovation Ecosystems on Management Innovation

The management innovation is one of the key pathways to enhancing the efficiency and effectiveness of the forestry innovation ecosystems. By employing refined management and scientific planning, management innovation introduces advanced management tools and methods, such as big data analysis and artificial intelligence, to better analyze and predict the dynamic changes in forestry resources. This enables more informed and rational management decisions, improving resource utilization efficiency. Additionally, management innovation promotes integrated management models, synergizing forestry production, ecological protection, and social development. For example, by integrating community development needs and combining forestry with tourism, culture, and other industries, management innovations enhance the comprehensive benefits and sustainability of forestry operations [15].

Based on the above analysis, this paper proposes the following hypothesis:

H7: 

The forestry innovation ecosystems increase green total factor productivity through management innovation, thereby promoting green economic development.

On the basis of theoretical analyses, this paper proposes a research idea as shown in Figure 1.

4. Empirical Design

4.1. Sample Selection and Data Sources

Given the data availability, this paper focuses on 31 provinces in China as the research subjects using data from 2012 to 2021 for analysis. The variables are categorized into explanatory, explained, and control variables, with basic descriptive statistics performed on the main variables to ensure the scientific rigor.

Data related to green total factor productivity, the primary explanatory variable, are obtained from the China Statistical Yearbook and the China Energy Statistical Yearbook. The measurement data for other explanatory variables such as forestry R&D innovation, the forestry policy drive, forestry industry operation, forestry social participation, and natural environment are mainly sourced from the China Statistical Yearbook, China Science and Technology Statistical Yearbook, China Forestry and Grassland Statistical Yearbook, China Environment Statistical Yearbook, and the Wisdom Sprout Global Patent Database (PatSnap). Control variables—such as the economic development level, local financial expenditures on agriculture, forestry, and water affairs, forestry wage level, and urbanization level—primarily come from the China Statistical Yearbook and the China Forestry and Grassland Statistical Yearbook.

A total of 34 provinces in China are selected as study samples. Considering the distinct political and economic environments of the special administrative regions and the availability of data, Hong Kong, Macao, and Taiwan are excluded from our analysis. Consequently, panel data from 31 provinces over 10 consecutive years from 2012 to 2021 are selected as the data source for this study.

4.2. Variable Definition and Descriptive Statistics

4.2.1. Dependent Variable

We use the Slack-based Measure-Global Malmquist–Luenberger (SBM-GML) productivity index to calculate the green total factor productivity of forestry and treat it as the dependent variable. The SBM-GML method evaluates the production efficiency and forecasts development trends by identifying potential areas for improvement, measuring the utilization rates of inputs and outputs in the production process, and suggesting measures to enhance the production efficiency [44,45].

Currently, Malmquist–Luenberger indices and Global Malmquist–Luenberger indices based on the directional distance function are widely used to measure green total factor productivity at the regional or industry levels [44,46]. The Global Malmquist–Luenberger index considers the entire reference set, effectively addressing issues such as infeasible linear programming solutions and “technological regression”, while also maintaining the properties of transitivity and cyclical accumulation.

The non-radial SBM model [47] with undesirable outputs is formulated as follows: let N represent the number of decision units, each with M inputs, desirable outputs, and undesirable outputs, where ${\mathrm{s}}_{\mathrm{i}\mathrm{d}}^{-}$, ${\mathrm{s}}_{{\mathrm{r}}_1\mathrm{d}}^{\mathrm{g}+}$, and ${\mathrm{s}}_{{\mathrm{r}}_2\mathrm{d}}^{\mathrm{b}-}$ represent the i-th input, the r1-th desirable output, and the r2-th undesirable output of DMUj, respectively.

The explanatory variable in this paper is green total factor productivity, which is measured according to the SBM-GML model, and its indicator system is shown in

Table 1. Measurement system of green total factor productivity.

Indicators	Sub-Indicators	Basic Indicators	Specific Evaluation Indicators	Unit
Input	Labor	Labor input	Employment at the end of the year	ten thousand
	Capital	Capital input	Value of fixed assets at the end of the year	billion
	Intermediate input	Water input	Total water consumption	billion tons
		Energy input	Total energy consumption	million tons of standard coal
Output	Expected output	Economic output	Value added by industry	billion
	Unexpected output	Environmental output	Industrial SO2 emissions	tons
		Environmental output	Industrial wastewater discharge	ten thousand tons

. This model consists of an objective function and multiple constraints.

$$\begin{array}{c}\mathsf{\rho }=\min {\displaystyle \frac{1-\frac{1}{\mathrm{M}}{\sum }_{\mathrm{i}=1}^{\mathrm{M}}\frac{{\mathrm{s}}_{\mathrm{i}\mathrm{d}}^{-}}{{\mathrm{x}}_{\mathrm{i}\mathrm{d}}}}{1+\frac{1}{{\mathrm{S}}_1+{\mathrm{S}}_2}\left({\sum }_{{\mathrm{r}}_1=1}^{{\mathrm{S}}_1}\frac{{\mathrm{s}}_{{\mathrm{r}}_1\mathrm{d}}^{\mathrm{g}+}}{{\mathrm{y}}_{{\mathrm{r}}_1\mathrm{d}}^{\mathrm{g}}}+{\sum }_{{\mathrm{r}}_2=1}^{{\mathrm{S}}_2}\frac{{\mathrm{s}}_{{\mathrm{r}}_2\mathrm{d}}^{\mathrm{b}-}}{{\mathrm{y}}_{{\mathrm{r}}_2\mathrm{d}}^{\mathrm{b}}}\right)}}\\ \mathrm{s}.\mathrm{t}.{\mathrm{x}}_{\mathrm{i}\mathrm{d}}={\displaystyle \sum _{\mathrm{j}=1}^{\mathrm{N}}}{\mathsf{\lambda }}_{\mathrm{j}\mathrm{d}}{\mathrm{x}}_{\mathrm{i}\mathrm{j}}+{\mathrm{s}}_{\mathrm{i}\mathrm{d}}^{-},\\ {\mathrm{y}}_{{\mathrm{r}}_1\mathrm{d}}^{\mathrm{g}}={\displaystyle \sum _{\mathrm{j}=1}^{\mathrm{N}}}{\mathsf{\lambda }}_{\mathrm{j}\mathrm{d}}{\mathrm{y}}_{{\mathrm{r}}_1\mathrm{j}}^{\mathrm{g}}-{\mathrm{s}}_{{\mathrm{r}}_1\mathrm{d}}^{\mathrm{g}+},\\ {\mathrm{y}}_{{\mathrm{r}}_2\mathrm{d}}^{\mathrm{b}}={\displaystyle \sum _{\mathrm{j}=1}^{\mathrm{N}}}{\mathsf{\lambda }}_{\mathrm{j}\mathrm{d}}{\mathrm{y}}_{{\mathrm{r}}_2\mathrm{j}}^{\mathrm{b}}+{\mathrm{s}}_{{\mathrm{r}}_2\mathrm{d}}^{\mathrm{b}-},\\ {\displaystyle \sum _{\mathrm{j}=1}^{\mathrm{N}}}{\mathsf{\lambda }}_{\mathrm{j}\mathrm{d}}=1\\ \mathsf{\lambda }\ge 0,{\mathrm{s}}_{\mathrm{i}\mathrm{d}}^{-}\ge 0,{\mathrm{s}}_{{\mathrm{r}}_1\mathrm{d}}^{\mathrm{g}}\ge 0,{\mathrm{s}}_{{\mathrm{r}}_2\mathrm{d}}^{\mathrm{b}}\ge 0\\ \mathrm{i}=\mathrm{1,2},\dots ,\mathrm{M};{\mathrm{r}}_1=\mathrm{1,2},\dots ,{\mathrm{S}}_1;{\mathrm{r}}_2=\mathrm{1,2},\dots ,{\mathrm{S}}_2;\mathrm{j}=\mathrm{1,2},\dots ,\mathrm{N}\end{array}$$

(1)

where:

$\mathsf{\rho }$ is both the efficiency score and the distance function ${\mathrm{D}}^{\mathrm{t}}({\mathrm{x}}_{\mathrm{i}\mathrm{d}}^{\mathrm{t}},{\mathrm{y}}_{{\mathrm{r}}_1\mathrm{d}}^{\mathrm{t}},{\mathrm{y}}_{{\mathrm{r}}_2\mathrm{d}}^{\mathrm{t}})$.

${\mathrm{s}}_{\mathrm{i}\mathrm{d}}^{-}$, ${\mathrm{s}}_{{\mathrm{r}}_1\mathrm{d}}^{\mathrm{g}+}$, and ${\mathrm{s}}_{{\mathrm{r}}_2\mathrm{d}}^{\mathrm{b}-}$ are the inputs, desirable output, and undesirable output slack variables.

${\mathrm{x}}_{\mathrm{i}\mathrm{j}}$, ${\mathrm{y}}_{{\mathrm{r}}_1\mathrm{j}}^{\mathrm{g}}$, and ${\mathrm{y}}_{{\mathrm{r}}_2\mathrm{j}}^{\mathrm{b}}$ are the inputs, desirable output, and undesirable output of the DMU j.

${\mathrm{x}}_{\mathrm{i}\mathrm{d}}$, ${\mathrm{y}}_{\mathrm{r}\mathrm{d}}$, and ${\mathrm{y}}_{{\mathrm{r}}_2\mathrm{d}}^{\mathrm{b}}$ are the inputs, desirable output, and undesirable output vectors of the DMU being evaluated.

${\mathsf{\lambda }}_{\mathrm{j}\mathrm{d}}$ is the vector of intensity variables.

In the SBM model, the efficiency score $\mathsf{\rho }$ ranges from 0 to 1. A higher $\mathsf{\rho }$ indicates greater production efficiency. When $\mathsf{\rho }$ = 1, it signifies that the decision-making unit is operating on the production frontier.

We use the GML method to further calculate green total factor productivity based on the production efficiency calculated by SBM [46,48]. The specific formula for GML is shown below.

$$\mathrm{G}\mathrm{M}{\mathrm{L}}_{\mathrm{t}}^{\mathrm{t}+1}={\displaystyle \frac{1+{\mathrm{D}}^{\mathrm{G}}\left({\mathrm{x}}_{\mathrm{i}\mathrm{d}}^{\mathrm{t}},{\mathrm{y}}_{{\mathrm{r}}_1\mathrm{d}}^{\mathrm{t}},{\mathrm{y}}_{{\mathrm{r}}_2\mathrm{d}}^{\mathrm{t}}\right)}{1+{\mathrm{D}}^{\mathrm{G}}\left({\mathrm{x}}_{\mathrm{i}\mathrm{d}}^{\mathrm{t}+1},{\mathrm{y}}_{{\mathrm{r}}_1\mathrm{d}}^{\mathrm{t}+1},{\mathrm{y}}_{{\mathrm{r}}_2\mathrm{d}}^{\mathrm{t}+1}\right)}}$$

(2)

In the formula, ${\mathrm{D}}^{\mathrm{G}}\left({\mathrm{x}}_{\mathrm{i}\mathrm{d}}^{\mathrm{t}},{\mathrm{y}}_{{\mathrm{r}}_1\mathrm{d}}^{\mathrm{t}},{\mathrm{y}}_{{\mathrm{r}}_2\mathrm{d}}^{\mathrm{t}}\right)$ is the global directional distance function, which is derived from the global production possibility set (the set of all production possibilities across all periods). If $\mathrm{G}\mathrm{M}{\mathrm{L}}_{\mathrm{t}}^{\mathrm{t}+1}$ > 1, it indicates that efficiency has increased during the observation period. A value of $\mathrm{G}\mathrm{M}{\mathrm{L}}_{\mathrm{t}}^{\mathrm{t}+1}$ = 1 signifies no change, while a value of $\mathrm{G}\mathrm{M}{\mathrm{L}}_{\mathrm{t}}^{\mathrm{t}+1}$ < 1 suggests a decline in efficiency, requiring corresponding measures for improvement.

The GML index can be further decomposed into a technical progress index and technical efficiency index [46,48], representing technical progress and technical efficiency, respectively. The decomposition expression is as follows:

$$\begin{array}{c}\mathrm{G}\mathrm{M}{\mathrm{L}}_{\mathrm{t}}^{\mathrm{t}+1}={\displaystyle \frac{1+{\mathrm{D}}^{\mathrm{t}}\left({\mathrm{x}}_{\mathrm{i}\mathrm{d}}^{\mathrm{t}},{\mathrm{y}}_{{\mathrm{r}}_1\mathrm{d}}^{\mathrm{t}},{\mathrm{y}}_{{\mathrm{r}}_2\mathrm{d}}^{\mathrm{t}}\right)}{1+{\mathrm{D}}^{\mathrm{t}+1}\left({\mathrm{x}}_{\mathrm{i}\mathrm{d}}^{\mathrm{t}+1},{\mathrm{y}}_{{\mathrm{r}}_1\mathrm{d}}^{\mathrm{t}+1},{\mathrm{y}}_{{\mathrm{r}}_2\mathrm{d}}^{\mathrm{t}+1}\right)}}\times \left[{\displaystyle \frac{\frac{1+{\mathrm{D}}^{\mathrm{G}}\left({\mathrm{x}}_{\mathrm{i}\mathrm{d}}^{\mathrm{t}},{\mathrm{y}}_{{\mathrm{r}}_1\mathrm{d}}^{\mathrm{t}},{\mathrm{y}}_{{\mathrm{r}}_2\mathrm{d}}^{\mathrm{t}}\right)}{1+{\mathrm{D}}^{\mathrm{t}}\left({\mathrm{x}}_{\mathrm{i}\mathrm{d}}^{\mathrm{t}},{\mathrm{y}}_{{\mathrm{r}}_1\mathrm{d}}^{\mathrm{t}},{\mathrm{y}}_{{\mathrm{r}}_2\mathrm{d}}^{\mathrm{t}}\right)}}{\frac{1+{\mathrm{D}}^{\mathrm{G}}\left({\mathrm{x}}_{\mathrm{i}\mathrm{d}}^{\mathrm{t}+1},{\mathrm{y}}_{{\mathrm{r}}_1\mathrm{d}}^{\mathrm{t}+1},{\mathrm{y}}_{{\mathrm{r}}_2\mathrm{d}}^{\mathrm{t}+1}\right)}{1+{\mathrm{D}}^{\mathrm{t}+1}\left({\mathrm{x}}_{\mathrm{i}\mathrm{d}}^{\mathrm{t}+1},{\mathrm{y}}_{{\mathrm{r}}_1\mathrm{d}}^{\mathrm{t}+1},{\mathrm{y}}_{{\mathrm{r}}_2\mathrm{d}}^{\mathrm{t}+1}\right)}}}\right]\\ =\mathrm{E}{\mathrm{C}}_{\mathrm{t}}^{\mathrm{t}+1}\times \mathrm{T}{\mathrm{C}}_{\mathrm{t}}^{\mathrm{t}+1}\end{array}$$

(3)

The technical efficiency change reflects variations in economies of scale due to institutional innovations and accumulated experience, while the technical progress change captures innovations in production technology and processes. Technical efficiency greater than 1 or less than 1 indicates an improvement or deterioration in the green technical efficiency. Similarly, technical progress greater than 1 or less than 1 indicates progress or regression in green technology.

The indicators used in this paper to calculate green total factor productivity are shown in Table 1.

4.2.2. Core Explanatory Variable

The core explanatory variable of this paper is forestry innovation ecosystems, which is measured utilizing a system consisting of five key variables [49,50]: forestry R&D innovation, the forestry policy drive, forestry industry operation, forestry social participation, and natural environment, as shown in

Table 2. Measurement system of forestry innovation ecosystems.

Indicators	Subdivision Indicators	Specific Evaluation Indicators	Unit
Technological innovation	Forestry R&D Innovation	Number of forestry patents granted in the year	-
Market innovation	Forestry policy driven	Level of environmental regulation in forestry = investment in industrial pollution control/total output value of forestry’s secondary industry	-
Institutional innovation	Forestry Industry Operations	Actual investment in forestry and grassland completed during the year	ten thousands CNY
Management innovation	Social participation in forestry	Number of visitors to the Science and Technology Museum during the year	ten thousand
	environment	Forest cover for the year	%

.

In this study, we refer to Liu’s [49] research results and take the number of granted patents as an indicator to measure the ability of scientific and technological innovation. Therefore, the number of forestry patents granted in the year is selected as the index to evaluate the level of forestry technological innovation.

According to Ye’s [50] research, this study selected the level of forestry environmental regulation as an index to evaluate forestry market innovation. In the case of low environmental regulation intensity, forest-related enterprises usually adopt similar strategies to other industries based on a cost–benefit analysis, resulting in the gradual migration of these enterprises to areas with fewer environmental regulations. However, when the intensity of environmental regulations is high, due to the continuous increase of end management costs, forest-related enterprises are under pressure to promote forestry green technology innovation and adopt more environmentally friendly and innovative technologies. The application of these technologies can not only reduce the cost of previous forestry innovations, but also bring additional net benefits as the results of green innovation in forestry are translated.

With reference to Liu’s [49] research results, this study takes the completed amount of forestry fixed asset investment as an indicator to measure the intensity of policy implementation. Therefore, this study selects the completed amount of forestry fixed assets investment as the index to evaluate institutional innovation.

According to Ye’s [50] research, this study selected the number of visitors to science and technology museums in that year as one of the indicators to evaluate forestry management innovation. Social participation plays a key role in the use and dissemination of knowledge resources. In the highly competitive market, consumer demand can effectively guide forestry enterprises to develop in the direction of green innovation. The public provides feedback to the government on green innovation policies through the expression of public opinion, which is also a form of social participation. The regulatory pressure of citizens’ environmental awareness can effectively complement environmental regulatory measures, and the public’s demand for emerging environmental technologies can significantly promote the vitality of the green innovation market.

According to Ye’s [50] research, this study selected the forest coverage rate as one of the indicators to evaluate forestry management innovation. The forest coverage rate is an important index to reflect the status of forestry resources in various regions. In the context of promoting energy conservation and emission reduction, the decline in forest coverage often reflects inadequate management.

We first standardize these variables to eliminate the differences in the scale and value range of different variables, allowing for a uniform comparison and analysis. Following the standardization, we calculate the average value to construct the forestry innovation ecosystems. The index constructed in this way provides a comprehensive assessment of the forestry innovation ecosystems’ status.

4.2.3. Control Variable and Mechanism Variable

Control Variables: Based on the research theme, this study selects the logarithm of the regional Gross Domestic Product (lnGDP), average wage, urbanization rate, and agricultural, forestry, and water affairs expenditure as the core control variables to comprehensively account for external factors influencing the relationship between forestry innovation ecosystems and green total factor productivity. The specific rationale for each control variable is as follows.

• Logarithm of Regional Gross Domestic Product.

This indicator is used to measure the overall economic scale and development level of a region. Economic development has profound implications for the innovation capacity and resource allocation in various industries, including the forestry sector. Higher Gross Domestic Product levels typically indicate greater financial and human resources available to support innovation activities, thereby facilitating the construction and optimization of forestry innovation ecosystems [51].

2.

Average Wage.

The average wage level reflects the supply–demand relationship in the regional labor market and labor costs, and it indirectly indicates the local residents’ living standards and economic vitality. Higher average wages are generally associated with a stronger consumption capacity. Regions with high economic activity and a strong consumption capacity often have a higher demand for green products and sustainable development, which, in turn, stimulates green technology innovation and promotes the Forestry Innovation Ecosystem towards a more environmentally friendly and efficient direction [52].

3.

Urbanization Rate.

The urbanization process is a crucial factor influencing forestry innovation ecosystems and green total factor productivity. As the urbanization rate increases, the population tends to concentrate in urban areas, which can lead to changes in the demand structure for timber and other forestry products. Moreover, urbanization is often accompanied by shifts in land use patterns and resource allocation, significantly impacting the spatial distribution of forestry resources and the ecological environment’s carrying capacity. Thus, controlling for the urbanization rate helps to eliminate the potential effects of regional socioeconomic structural transformation on the development of forestry innovation ecosystems [53].

4.

Agricultural, Forestry, and Water Affairs Expenditure.

Government expenditure in agricultural, forestry, and water affairs directly drives the development of the forestry sector. Such expenditures typically encompass areas such as forestry infrastructure construction, scientific and technological innovation, and ecological protection, which serve as important material guarantees for the establishment of forestry innovation ecosystems. By including agricultural, forestry, and water affairs expenditure as a control variable, this study can effectively exclude the potential interference of government investments in the forestry and related sectors, thereby providing a more accurate evaluation of the independent impact of forestry innovation ecosystems on green total factor productivity [54].

In summary, the selection of these core control variables allows the study to systematically explore the actual impact mechanism of forestry innovation ecosystems on green total factor productivity, providing a more reliable foundation for subsequent policy formulation.

Mechanism variables: based on the discussion in Section 3.2, we selected technological innovation, market innovation, institutional innovation, and management innovation as variables for the mechanism analysis.

4.2.4. Descriptive Statistics

As shown in

Table 3. Descriptive statistics.

Variable	Obs	Mean	Std. Dev.	Min	Max
Green total factor productivity	310	1.013	0.132	0.608	1.928
Forestry innovation ecosystems	310	−0.016	0.532	−0.855	1.743
LnGDP	310	9.78	0.996	6.566	11.734
Average wage (CNY)	310	64,168.839	34,419.481	19,893	207,948
Urbanization rate (%)	310	59.28	12.767	22.87	89.6
Expenditure on agriculture, forestry, and water affairs (ten thousand CNY)	310	576.009	276.915	100.98	1339.36
Technical efficiency	310	0.993	0.147	0.436	1.853
Technical progress	310	1.047	0.21	0.492	2.561
Technological innovation	310	0.000	1	−0.968	4.935
Market innovation	310	0.000	1	−0.832	5.978
Institutional innovation	310	0.000	1	−0.174	11.378
Management innovation	319	0.000	0.761	−1.309	2.332

, the descriptive statistics for all variables are presented as follows.

The mean value of green total factor productivity is 1.013, with a standard deviation of 0.132. The minimum value of 0.608 and the maximum value of 1.928 reflect a wide range of productivity levels, suggesting the existence of significant disparities in green productivity among provinces. The mean value of forestry innovation ecosystems is −0.016, with a standard deviation of 0.532. The values range from a minimum of −0.855 to a maximum of 1.743, demonstrating substantial heterogeneity among provinces in terms of forestry innovation capabilities.

The mean value of lnGDP is 9.78, with a standard deviation of 0.996. The minimum and maximum values are 6.566 and 11.734, respectively, indicating considerable differences in the economic scale of different provinces over the sample period. The mean value of the average wage is CNY 64,168, with a relatively high standard deviation of CNY 34,419, reflecting significant variations in the labor cost and income levels across the provinces. The minimum value is CNY 19,893, while the maximum value is CNY 207,948, indicating large disparities in wage levels between different regions. The mean urbanization rate is 59.28%, with a standard deviation of 12.77%. The urbanization rate varies substantially between provinces, with a minimum value of 22.87% and a maximum value of 89.6%, reflecting significant regional differences in urbanization levels. The mean value of expenditure on agriculture, forestry, and water affairs is CNY 5,760,009, with a standard deviation of CNY 2,769,150. The minimum and maximum values are CNY 1,009,800 and CNY 13,393,600, respectively, indicating a considerable range in government spending on agriculture, forestry, and water affairs across different provinces.

The mean value of technical efficiency is 0.993, with a standard deviation of 0.147. The minimum and maximum values are 0.436 and 1.853, indicating significant differences in technical efficiency across regions. The mean value of the technical progress index is 1.047, with a standard deviation of 0.21. The values range from a minimum of 0.492 to a maximum of 2.561, suggesting considerable variability in the extent of technical progress achieved by different regions.

The mean value of technological innovation is 0.000, with a standard deviation of 1. The values range from a minimum of −0.968 to a maximum of 4.935, indicating substantial differences in technological innovation levels among provinces, with some regions displaying significant innovation advantages. The mean value of market innovation is 0.000, with a standard deviation of 1. The minimum and maximum values are −0.832 and 5.978, respectively, demonstrating a wide range in market innovation capabilities among provinces, suggesting that some provinces have achieved notable advancements in market innovation. The mean value of institutional innovation is 0.000, with a standard deviation of 1. The values range from a minimum of −0.174 to a maximum of 11.378, reflecting substantial heterogeneity in the institutional innovation levels across provinces. Finally, the mean value of management innovation is 0.000, with a standard deviation of 0.761. The minimum value is −1.309, and the maximum value is 2.332, indicating variability in the management innovation performance among provinces, with some regions performing significantly better in management innovation than others.

Overall, the descriptive statistics reveal considerable heterogeneity among provinces in terms of economic, social, and innovation-related factors, highlighting the complexity of the relationship between forestry innovation ecosystems and green total factor productivity.

Two-way fixed effects models contain two dimensions of fixed effects: individual fixed effects and time fixed effects [55]. Individual fixed effects account for unobservable individual heterogeneity, which refers to stable, time-invariant differences across individuals. Time-fixed effects control for common temporal trends or macro-environmental changes that affect all individuals simultaneously [56]. By incorporating both types of fixed effects, the two-way fixed effects model effectively addresses potential omitted variable bias, improving the accuracy and validity of the estimation. This model makes the assessment of the relationship between the independent and dependent variables more reliable and precise. Therefore, this paper applies the two-way fixed effects model for empirical analyses.

$${GTFP}_{it}={\beta }_0+{\beta }_1{FIE}_{it}+\beta \sum controls+{\mu }_{it}+{\delta }_{it}+{\epsilon }_{it}$$

(4)

where FIE represents forestry innovation ecosystems and GTFP represents green total factor productivity.

5. Results

5.1. Main Test Results

Table 4. Two-way fixed effect regression results.

Variable	(1)	(2)	(3)
Forestry innovation ecosystems	0.074 ** (0.032)	0.088 ** (0.040)	−0.050 (0.055)
Constant	−0.587 (0.861)	0.424 (1.083)	−0.293 (1.481)
Controls variable	Yes	Yes	Yes
Time-fixed effect	Yes	Yes	Yes
Region-fixed effect	Yes	Yes	Yes
Observations	310	310	310

Notes: Standard errors are indicated in parentheses. *** p < 0.01, ** p < 0.05, * p < 0.1.

reports the regression results of the two-way fixed effects model. Column (1) presents the empirical regression result for Hypothesis 1, showing a significant effect of forestry innovation ecosystems on green total factor productivity with a coefficient of 0.074. This indicates that a one-unit increase in forestry innovation ecosystems increases green total factor productivity by 0.074, thus validating Hypothesis 1. Columns (2) and (3) present the empirical results for Hypotheses 2 and 3. Column (2) demonstrates a significant contribution of forestry innovation ecosystems to the technical efficiency, with a coefficient of 0.088, meaning that a one-unit increase in forestry innovation ecosystems raises technical efficiency by 0.088, thereby supporting Hypothesis 2. However, Column (3) shows no significant effect of forestry innovation ecosystems on technical progress, invalidating Hypothesis 3.

5.2. Mechanism Analysis

Table 5. Two-way fixed effect regression results for mechanism analysis.

Variable	(1)	(2)	(3)	(4)
Technological innovation	0.015 (0.016)
Market innovation		0.009 (0.015)
Institutional innovation			0.020 ** (0.009)
Management innovation				0.014 (0.027)
Constant	−0.669 (0.878)	−0.841 (0.862)	−0.912 (0.857)	−0.749 (0.878)
Control variable	Yes	Yes	Yes	Yes
Time-fixed effect	Yes	Yes	Yes	Yes
Region-fixed effect	Yes	Yes	Yes	Yes
Observations	310	310	310	310

Notes: Standard errors are indicated in parentheses. *** p < 0.01, ** p < 0.05, * p < 0.1.

reports the results of the mechanism analysis. Columns 1 to 4 present the bidirectional fixed effects regression results for the impact of technological innovation, market innovation, institutional innovation, and management innovation on green total factor productivity. It can be observed that technological innovation, market innovation, and management innovation do not have a significant impact on green total factor productivity. In contrast, institutional innovation exhibits a significant positive effect on green total factor productivity, with a marginal coefficient of 0.020. This suggests that for every unit increase in institutional innovation, green total factor productivity increases by 0.020 units. Therefore, Hypotheses 4, 5, and 7 are not supported, while Hypothesis 6 is confirmed.

5.3. Heterogeneity Analysis

Table 6. Two-way fixed effect regression results in the eastern region.

Variable	(1)	(2)	(3)
Forestry innovation ecosystems	0.057 (0.055)	0.051 (0.060)	−0.054 (0.093)
Constant	−1.816 (1.581)	−0.971 (1.745)	−0.390 (2.673)
Controls variable	Yes	Yes	Yes
Time-fixed effect	Yes	Yes	Yes
Region-fixed effect	Yes	Yes	Yes
Observations	110	110	110

Notes: Standard errors are indicated in parentheses. *** p < 0.01, ** p < 0.05, * p < 0.1.

reports the regression results for the two-way fixed effects model in the eastern region. The findings indicate that in the eastern region, forestry innovation ecosystems do not have a significant effect on green total factor productivity. Additionally, forestry innovation ecosystems do not significantly impact the two mechanism paths: technical efficiency and technical progress.

Table 7. Two-way fixed effect regression results in the central region.

Variable	(1)	(2)	(3)
Forestry innovation ecosystems	−0.024 (0.118)	−0.011 (0.109)	−0.099 (0.129)
Constant	−0.175 (1.790)	−0.564 (1.645)	1.396 (1.954)
Controls variable	Yes	Yes	Yes
Time-fixed effect	Yes	Yes	Yes
Region-fixed effect	Yes	Yes	Yes
Observations	80	80	80

Notes: Standard errors are indicated in parentheses. *** p < 0.01, ** p < 0.05, * p < 0.1.

presents the regression results for the two-way fixed effects model specific to the central region. The results indicate that in this region, forestry innovation ecosystems do not exert a significant effect on green total factor productivity. Furthermore, forestry innovation ecosystems do not significantly impact the two mechanism paths: technical efficiency and technical progress.

Table 8. Two-way fixed effect regression results in the western region.

Variable	(1)	(2)	(3)
Forestry innovation ecosystems	0.100 ** (0.049)	0.162 * (0.084)	−0.092 (0.106)
Constant	−0.999 (1.274)	1.778 (2.170)	−1.104 (2.736)
Controls variable	Yes	Yes	Yes
Time-fixed effect	Yes	Yes	Yes
Region-fixed effect	Yes	Yes	Yes
Observations	120	120	120

Notes: Standard errors are indicated in parentheses. *** p < 0.01, ** p < 0.05, * p < 0.1.

reports the results of the two-way fixed effects regressions for the western region. Columns (1) through (3) show the regression results for green total factor productivity, technical efficiency, and technical progress. In the western region, forestry innovation ecosystems have a significant effect on green total factor productivity, with a marginal coefficient of 0.1. This indicates that a one-unit increase in forestry innovation ecosystems leads to a 0.1-unit increase in green total factor productivity. Additionally, forestry innovation ecosystems significantly affect technical efficiency, with a marginal coefficient of 0.162, suggesting that a one-unit increase in forestry innovation ecosystems results in a 0.162-unit increase in technical efficiency. However, forestry innovation ecosystems do not have a significant effect on technical progress.

In summary, the impact of forestry innovation ecosystems on green total factor productivity exhibits significant regional disparities. In western regions, forestry innovation ecosystems have a notable positive effect on green total factor productivity, while in eastern and central regions, this impact is not statistically significant. This discrepancy may be closely related to the spatial distribution of forests in China. The relatively young stand age for Chinese forests is mostly due to the large proportion of newly planted forests (0–40 years old), which are more prevailing in south China. Older forests (stand age > 60 years old) are more frequently found in the east Qinghai–Tibetan Plateau and the central mountain areas of west and northeast China, where human activities are less intensive [9]. This age difference in forest resources might be one of the reasons why forestry innovation ecosystems have a less significant impact on green total factor productivity in the eastern and central regions.

5.4. Robustness Check

To ensure the robustness of the conclusions, we employed two methods: replacing the forestry innovation ecosystems variable and modifying the sample period. First, we re-estimated the model by substituting the forestry innovation ecosystems variable with its mean value through normalization [57]. The results, as shown in

Table 9. Robustness test results using the replacement of forestry innovation ecosystems.

Variable	(1)	(2)	(3)
Forestry innovation ecosystems	0.074 ** (0.032)	0.088 ** (0.040)	−0.050 (0.055)
Constant	−0.587 (0.861)	0.424 (1.083)	−0.293 (1.481)
Controls variable	Yes	Yes	Yes
Time-fixed effect	Yes	Yes	Yes
Region-fixed effect	Yes	Yes	Yes
Observations	310	310	310

Notes: Standard errors are indicated in parentheses. *** p < 0.01, ** p < 0.05, * p < 0.1.

, indicate a significant effect of forestry innovation ecosystems on green total factor productivity, confirming that Hypothesis 1 passes the robustness test. Additionally, forestry innovation ecosystems significantly affect technical efficiency, meaning that Hypothesis 2 passes the robustness test. However, forestry innovation ecosystems do not have a significant effect on technical progress, invalidating Hypothesis 3. The primary findings of the robustness test are consistent with the results of the main test.

We present the robustness test results based on the modified sample period in

Table 10. Robustness test results for changing sample period range.

Variable	(1)	(2)	(3)
Forestry innovation ecosystems	0.155 ** (0.054)	0.205 *** (0.073)	−0.140 (0.100)
Constant	1.952 (1.383)	2.425 (1.884)	0.079 (2.559)
Controls variable	Yes	Yes	Yes
Time-fixed effect	Yes	Yes	Yes
Region-fixed effect	Yes	Yes	Yes
Observations	217	217	217

Notes: Standard errors are indicated in parentheses. *** p < 0.01, ** p < 0.05, * p < 0.1.

. Considering that splitting the sample into two separate time periods would lead to insufficient sample sizes, we chose to use data from 2012 to 2018 for the robustness test. As shown in Table 10, the impact of forestry innovation ecosystems on green total factor productivity remains significant, indicating that Hypothesis 1 passes the robustness test. Moreover, forestry innovation ecosystems have a significant effect on technical efficiency, thereby confirming Hypothesis 2. However, the impact of forestry innovation ecosystems on technical progress is not significant, rendering Hypothesis 3 invalid. Overall, the main findings of the robustness test are consistent with those of the primary analysis, reinforcing the reliability of the conclusions.

In summary, the results of these robustness tests demonstrate that the conclusions of this study are robust and reliable.

6. Discussion

This study empirically supports the positive impact of forestry innovation ecosystems on green economic development, especially in boosting green total factor productivity in western regions. This finding has important implications for promoting green economic transformation across different regions in China.

First, the results from the regional heterogeneity analysis indicate that the forestry innovation ecosystems have a significant positive effect on green total factor productivity and technical efficiency in the western regions. This suggests that the potential for ecological protection and green development in these areas has not been fully tapped. In contrast, the eastern and central regions may have already achieved a higher level of green development in other aspects, resulting in the relatively smaller marginal contribution of the forestry innovation ecosystems. This can be compared with the research results of Liu et al. (2023) [49], who took the influence of the economic development level into special consideration when discussing the coordinated development of forestry. Due to the higher level of economic development in the eastern region, the forestry development level in the eastern region is also more prominent in the study of Liu et al. Therefore, policymakers should adopt region-specific approaches in promoting nationwide green economic transformation, particularly by enhancing support for the construction of forestry innovation ecosystems in the western regions to fully leverage their potential in green economic development.

Second, the mechanism analysis shows that although technological innovation, market innovation, and management innovation do not demonstrate a significant impact on green total factor productivity, institutional innovation has a significant positive effect. This highlights the importance of institutional development within the innovation ecosystem in driving the growth of green total factor productivity. In the study of Liu et al. [49], it was also found that the factors affecting the coordinated development of forestry were the government’s financial capacity, market environment, scientific and technological innovation capacity, economic development level, and policy implementation. Policy formulation should, therefore, focus more on optimizing and improving the institutional environment by designing and implementing effective policies that incentivize innovation activities in the forestry sector, thereby enhancing the industry’s overall green development level.

Lastly, although the forestry innovation ecosystems do not show a significant impact on technical progress, this does not imply that forestry innovation is irrelevant to technological advancement. A possible explanation is that the effects of forestry innovation on technical progress may require a longer time to manifest or may be more effective in specific technological domains. Kevin et al. [58] mentioned these possible barriers to innovation in the forestry sector: (i) weak networks blocking the capacity development of new forest owners; (ii) infrastructural problems blocking the reach and effectiveness of “knowledge networks”; and (iii) rigid institutional structures and policy blocking co-innovation and sustainable transitions. Therefore, future research can continue to investigate the long-term impact of the forestry innovation ecosystems on technical progress and explore their potential mechanisms.

Overall, this study provides new perspectives and evidence on the role of forestry innovation ecosystems in promoting green economic development and offers directions for future policy formulation and research. Advancing the development of forestry innovation ecosystems not only contributes to regional green economic development, but also aligns with global sustainable development goals.

7. Conclusions and Policy Recommendations

7.1. Conclusions

In the context of globalization, the synergistic relationship between economic development and ecological environmental protection has become an urgent need of the times, as well as an inevitable trend. Green economic development emphasizes the coordination of the economy, society, and the environment by reducing resource consumption, lowering environmental pollution, and improving eco-efficiency alongside economic growth. Forestry, as one of the most important ecosystems on earth, not only supplies material resources such as timber and forest products, but also plays an irreplaceable role in regulating the climate, conserving soil and water, preserving biodiversity, etc. As a key element and driving force behind green development, forestry innovative ecosystems have demonstrated great potential in optimizing resource use, enhancing ecological services, fostering industries’ innovations, etc. However, research on the impact of forestry innovation ecosystems on green economic development is still in the primary stage, with limited studies available. This paper bridges the gap by analyzing data from 31 provinces in China from 2012–2021, employing a two-way fixed effects empirical model. The research findings are as follows.

(1)

There is a significant effect of forestry innovation ecosystems on green total factor productivity, with a coefficient of 0.074, indicating that a one-unit increase in forestry innovation ecosystems can increase green total factor productivity by 0.074.

(2)

A mechanism analysis finds a significant promotion effect of forestry innovation ecosystems on technical efficiency, and its coefficient is 0.088, signifying that a one-unit increase in forestry innovation ecosystems boosts the technical efficiency increase by 0.088. However, there is no significant effect of forestry innovation ecosystems on technical progress.

(3)

A heterogeneity analysis finds that in the eastern and central regions, forestry innovation ecosystems do not have a significant effect on green total factor productivity or the two mechanism paths of technical efficiency and technical progress. However, in the western region, forestry innovation ecosystems significantly affect green total factor productivity, with a marginal coefficient of 0.1, which means that a one-unit increase in forestry innovation ecosystems increases green total factor productivity by 0.1. Forestry innovation ecosystems also have a significant effect on technical efficiency, with a marginal coefficient of 0.162, meaning that a one-unit increase in forestry innovation ecosystems enhances the technical efficiency by 0.162, but without a significant effect on TFP and technical progress.

7.2. Policy Recommendations

Based on the empirical results, this paper puts forward the following policy recommendations.

• The government should further increase the support for the forestry innovation ecosystems and promote its development by increasing funds for forestry-related scientific research and technological development. This includes setting up a special fund to encourage enterprises and scientific research organizations to invest in the research, application, and development of green technologies. The government should formulate and implement a series of policies to optimize the forestry production process, enhance efficient resource use, strengthen technical training and upgrade skills for forestry practitioners, and promote the transfer and diffusion of technological knowledge to improve the technical level and industry efficiency as a whole.
• Given the regional differences, differentiated policies should be formulated. In the western region, increased input and support to the forestry innovation ecosystems are crucial in them further playing their role in promoting green development. In the eastern and central regions, policy guidance should focus on encouraging enterprises and research institutes to invest more in green technology research, application, and development to enhance the overall effectiveness of the forestry innovation ecosystems.
• The government should promote the role of forestry innovation ecosystems in technical progress through innovative policies. It should establish special scientific research funds for new forestry-related technologies, facilitating international cooperation to introduce advanced green technologies and management experience from abroad, and promoting technical exchanges and personnel training. In addition, by formulating and implementing intellectual property protection policies, the government should incentivize enterprises and research institutes to invest more in protecting technological innovations.