Spatial Divergence of Forestry Green Total Factor Productivity in China Under the Constraint of Carbon Emissions

Abstract

In the dual-carbon context, forestry green total factor productivity (FGTFP) serves as a key indicator of the quality and efficiency of forestry development. Based on New Economic Geography Theory, this study explores FGTFP and its spatial divergence under the constraint of carbon emissions. We analyzed panel data from 30 Chinese provinces between 2004 and 2022. The Directional Distance Function (DDF) model was applied to measure FGTFP, and the Global Malmquist–Luenberger (GML) model was applied to measure FGTFP’s decomposition index. The Dagum Gini coefficient was employed to analyze the degree of spatial divergence of FGTFP and identify its sources. Using Porter’s model and Sustainable Development Theory, the geo-detector was applied to examine the driving factors of FGTFP and its decomposition index. The study’s findings indicate that (1) FGTFP in China generally trended upward from 2004 to 2022, with significant heterogeneity observed at both interprovincial and regional levels; (2) Technological Improvement (TI) was the primary driver of FGTFP growth in the eastern, northeastern and central regions, while Efficiency Change (EC) was the key driver in the western region; (3) FGTFP exhibited distinct spatial divergence patterns in China, with hypervariable density as the primary source, followed by interregional differentiation, and regional differentiation contributing the least; and (4) green energy transition factors consistently showed a significant “two-factor enhancement effect” and a “non-linear enhancement trend”, while external environmental factors exhibited strong interaction effects but demonstrated a “non-linear weakening trend”. Therefore, it is essential to promote the green transformation of production modes, facilitate structural adjustments and upgrades in the forestry industry, enhance regional collaboration, and advance the “dual enhancement” of technological progress and efficiency. Additionally, leveraging regional comparative advantages will promote coordinated development.

1. Introduction

In 2022, General Secretary Xi Jinping emphasized that forests are not only reservoirs of water, money and grain but also serve as “carbon reservoirs”. Carbon storage is a key component of forest ecosystem services, with forests serving as the primary “carbon reservoir” of terrestrial ecosystems. According to the results of the ninth national forest resources inventory, the carbon reserves of national forest vegetation amount to 9.186 million tons. Forestry development is closely linked to achieving carbon peaking and carbon neutrality. Maximizing the carbon storage function of forests is essential for meeting the Intended Nationally Determined Contributions (INDCs) [1]. The “14th Five-Year” Forestry and Grassland Protection and Development Plan emphasizes that forestry provides not only tangible forest products, such as timber, but also significant intangible ecological value, which is vital for the sustainable development of the economy and society. To meet the requirements of high-quality economic development in the new era, the forestry development model should be transformed to enhance quality, efficiency and growth momentum [2]. Additionally, investing in advanced production factors, such as green production technologies, will facilitate the green transformation of the forestry sector.

In the context of increasing concerns about global warming and environmental protection, the goal of reaching carbon peak and achieving carbon neutrality has gained global consensus. As the largest developing country, China is committed to achieving carbon peak and carbon neutrality in the shortest time globally, while advancing these actions in a steady, orderly and gradual manner. The 20th National Congress of the Communist Party of China suggests “accelerating the green transformation of the development model” and “actively and steadily advancing carbon peaking and carbon neutrality”. The State Council’s 2030 Peak Carbon Action Program emphasizes the need to “solidly promote peak carbon actions”. The challenging task of “achieving carbon peak” has created more urgent requirements for promoting high-quality forestry development and maximizing the role of forests as a “carbon reservoirs”. Meanwhile, the differentiation in resource endowments and industrial development strategies across regions in China has become more pronounced, leading to greater regional disparities in FGTFP. Under the dual-carbon context, analyzing the spatial and temporal variation in FGTFP and its driving factors under the constraint of carbon emissions will assist in understanding the path toward high-quality forestry development. It will also promote coordinated regional development and inform future forestry policies to achieve China’s dual-carbon targets.

Forestry total factor productivity (FTFP) is a key indicator of the quality and efficiency of forestry development. Scholars have approached the study of FTFP from various perspectives, including industrial specialization [3], ecological vulnerability [4], territorial networks [5], industrial agglomeration [6], sustainable development [7] and factor allocation [8]. From the methodological perspective, Cao et al. innovatively applied the three-stage DEA model to assess FTFP and employed the fractional regression model to investigate the differential effects of forestry subsidies and forestry regulatory policies on FTFP [9]. Shi et al. employed the stochastic frontier model (SFA) combined with the Malmquist index to measure FTFP [10]. From the perspective of measurement indicators, Huang et al. measured FTFP based on the three main input factors (land, capital and labor) and the primary output factor (forestry income). The study found significant absolute and conditional β convergence of household FTFP in the northwest region [11]. Shah et al. also selected land, capital and labor as input factors, and, considering multiple objectives, including economic, ecological and social benefits, used three output factors (forestry output value, forest stock volume and timber output) [12]. Liu et al. used three output factors (gross output value, forest land renovation area and value added of forestry) to measure the FTFP in Guangdong Province [13]. Existing research on FTFP is abundant in both research perspectives and measurement methods.

With the rapid global economic development and the increasing frequency of environmental pollution, green total factor productivity (GTFP) has gained significant academic attention. Currently, research on GTFP primarily focuses on urban [14], agriculture [15], industry [16] and other fields. As the forestry industry continues to develop, limited forestry resources have become a major constraint on the high-quality growth of the sector. Meanwhile, GTFP in the forestry sector has shown gradual improvements. Based on the concept of sustainable development, FGTFP accounts for negative externalities, such as energy consumption and carbon stock reduction, and more comprehensively reflects resource utilization efficiency in high-quality forestry development. Liu et al. incorporated the emissions of wastewater, soot and solid waste from the forestry industry into an accounting index system using the “non-desired output” super-efficiency SBM–Malmquist model and analyzed the spatial and temporal divergence, along with the influencing factors, using the Terre index and Tobit regression model [17]. Chen et al. also considered forestry waste gas, wastewater and waste materials as non-desired outputs, calculated FGTFP for 30 provinces in China using the Non-Radial Directional Distance Function (NDDF) model, and used a spatial Dobbin model to analyzed the impact of the digital economy on FGTFP [18]. In general, scholars have considered various types of waste as “non-desired outputs” of FGTFP and assessed the key factors influencing it.

However, few scholars have explored the spatial analysis of FGTFP. In economics, spatial analysis often involves the theory of spatial autocorrelation, which posits that, in geospatial terms, the economic behavior or characteristics of a region are influenced by neighboring regions. Lv et al. observed a significant strengthening of spatial positive autocorrelation in FGTFP, with high–high (H-H) aggregation in the central and eastern regions and low–low (L-L) aggregation in the western and northeastern regions from 2014 to 2019 [19]. Building on previous studies, this paper further analyzes the spatial divergence of FGTFP to more accurately identify the driving factors behind it. In this study, the concept of “spatial” encompasses not only geographic proximity but also serves as a key tool for analyzing differences in regional economic performance, while “divergence” refers to disparities arising during regional development. The concept of “spatial differentiation” is explored within the framework of New Economic Geography Theory. In 1991, Krugman published the seminal paper “Increasing Return and Geography of Economics” in the Journal of Political Economy, where he introduced the theory of New Economic Geography [20]. Krugman systematically developed the theoretical framework of “center-periphery”, analyzed the formation of industrial agglomeration and offered a dynamic perspective on the concept of “spatial divergence”. The core concept of spatial divergence in the theory of New Economic Geography is that the uneven development of geographical spaces arises from the interaction of geographic location, economies of scale, natural resources, transportation costs, technology diffusion and path dependence. Therefore, New Economic Geography provides a framework to deeply explore how natural resources promote forestry production, how forestry resources interact with economies of scale and how the diffusion of green technology occurs unevenly, thereby offering a comprehensive understanding of the underlying causes of spatial divergence in FGTFP.

Drawing on New Economic Geography Theory, the spatial effects of green energy transition factors and external environmental factors on FGTFP enhancement have been further explored through the Porter effect [21]. In the context of the green energy transition, spatial interactions between policy incentives, green technology innovation and industrial agglomeration drive the differentiated development of FGTFP. Specifically, the government promotes the sustainable use of green energy through policy incentives, such as carbon taxes and green technology research and development subsidies, to enhance FGTFP. Green energy transformation depends on green technology to generate spatial knowledge spillovers, break technological barriers and foster cross-regional innovation networks. Green technology and green industry drive regional industrial agglomeration through spatial interactions, reshape the green production pattern of forestry with geographic location as a core element, ensure the parallel development of industrial competitiveness and environmental sustainability, and promote FGTFP growth. According to Sustainable Development Theory, ecosystem resilience and resistance contribute to the rational and sustainable use of forest resources, enhancing ecosystem stability and carbon sink capacity. Forest resources also provide an ecological resource supply advantage for forestry production. The level of regional economic development offers a strong market space for the green transformation of the forestry industry, while the energy-intensive labor structure reflects the efficiency of energy use under the constraint of carbon emissions. Attracting green investment and skilled personnel and optimizing factor allocation efficiency can jointly promote the green development of forestry and further enhance FGTFP. Figure 1 presents a diagram of the theoretical analysis pathway.

Exploring the spatial divergence of FGTFP is crucial for understanding regional variations, helping the government optimize resource allocation and providing a basis for developing targeted green strategies. However, there are relatively few studies on the spatial divergence of FGTFP and its drivers, compared to more developed research on the spatial divergence of FTFP. Several scholars have employed VAR Granger causality tests [5], spatial autocorrelation analysis [6,9], the Terre index [7], the Kruskal–Wallis test [12], the kernel density function [18] and the Gini coefficient [22] to examine the spatial divergence characteristics of FTFP. However, these methods cannot pinpoint the specific sources of spatial heterogeneity or the distribution of sub-samples. The Dagum Gini coefficient can systematically reveal the extent and sources of regional differences, addressing the issue of cross overlap between sample data [23]. The Dagum Gini coefficient has been used by numerous scholars to analyze regional divergence in GTFP [24], low-carbon total factor productivity [25], water resource total factor productivity [26] and agricultural total factor productivity [27]. However, its application in the forestry field remains limited. Regarding the driving factors of FTFP, several scholars have applied econometric methods, including the system generalized method of moments (SYS-GMM) [3], the Tobit regression model [4] and the spatial Durbin model [28]. These studies focused on factors such as technological progress bias [8], climate [10], human capital [11] and economic development [29]. However, they did not address the influence of spatial divergence in FTFP. The geo-detector is a statistical method used to analyze spatial divergence and its driving factors [30] that is capable of identifying key factors that influence the spatial divergence of FTFP. Compared to the Tobit model, geo-detector model interaction detection results exhibit lower bias and higher validity [31]. This method has been progressively applied to study spatial divergence and its driving factors in agricultural and social sciences, including agricultural low-carbon productivity [32], grain production efficiency [33], production efficiency in processing tomatoes [34] and digital production efficiency in the manufacturing industry [35].

In summary, the existing literature primarily focuses on the overall situation and driving factors of FTFP. However, research on the driving factors of FGTFP, considering the constraint of carbon emissions from a multi-framework perspective, is somewhat insufficient. In response to this gap, this paper introduces two innovations: First, considering the ecological role of forests as “carbon reservoirs”, deforestation reduces carbon stocks and diminishes their carbon storage capacity. Additionally, carbon dioxide emissions are included as a “non-desired output” in the accounting system for FGTFP growth. The DDF-GML model was applied to quantitatively assess FGTFP across 30 provinces in China. Secondly, Porter’s model and Sustainable Development Theory were applied to reveal the driving factors of FGTFP and its index decomposition, focusing on the constraint of carbon emissions at the green energy transition and external environmental levels.

2. Materials and Methods

2.1. Construction of the Indicator System and Sources of Driving Factors

2.1.1. Input–Output Factors

According to the Theory of Factors of Production, land, labor and capital are the three traditional factors of production [36]. Thus, these factors were selected as input factors: forest area as a proxy for land, number of employees in the forestry sector at year-end as a proxy for labor and investment in forestry fixed assets as a proxy for capital. Desired outputs are represented by the total output value of the forestry industry and forest stock, reflecting economic and ecological benefits, respectively. Carbon emissions are a major contributor to global warming. As global concern over climate change intensifies, reducing greenhouse gas emissions has become a shared global goal. China also promotes the development of green, low-carbon forestry to support environmentally sustainable development. By quantifying the negative externalities of forestry production activities, we can assess their sustainability and align with the global goal of combating climate change and with China’s green, low-carbon development strategy. In this paper, carbon dioxide emissions from forestry production were used to reflect undesired outputs. These emissions were calculated using the carbon emission coefficients for various energy types from the 2006 IPCC Guidelines for National Greenhouse Gas Inventories (

Table 1. Input and output indicators.

Index	Variables		Unit
Input	Forest area		hm2
	Number of employees in the forestry sector at year-end		People
	Investment in forestry fixed assets		CNY ten thousand
Output	Desired output	Total output value of the forestry industry	CNY ten thousand
		Forest stock	Ten thousand m3
	Non-desired output	Carbon dioxide emissions from forestry production	Ten thousand tCO2e

).

2.1.2. Driving Factors

Based on the Porter effect, FGTFP, EC and TI are influenced by both the green energy transition factors and the external environmental factors, particularly under the constraint of carbon emissions [21]. In terms of the green energy transition, appropriate technological regulation can enhance innovation, thereby boosting FGTFP. Therefore, energy consumption, clean energy structure and forestry industry structure were selected as the key influencing factors. Specifically, clean energy structure is defined as the ratio of clean energy consumption to total energy consumption, while forestry industry structure is defined as the ratio of forestry tertiary industry output to the total forest and grass output. In terms of the external environment, the coordinated development of nature, economy and society is crucial for sustaining the growth of FGTFP. Therefore, forest coverage, GDP per capita and the number of employees in energy-intensive industries were identified as key influencing factors (

Table 2. Driving factors of FGTFP in China under the constraint of carbon emissions.

Factor	Variables		Abbreviation	Unit
Green energy transition factors	Energy consumption		X1	Ten thousand tons
	Clean energy structure		X2	%
	Forestry industry structure		X3	%
External environmental factors	Natural	Forest coverage	X4	%
	Economic	GDP per capita	X5	CNY/people
	Social	Number of employees in energy-intensive industries	X6	Ten thousand people

).

2.1.3. Data Sources

Based on data availability, this study selected 30 provinces (excluding Tibet, Hong Kong, Macao and Taiwan) as samples from 2004 to 2022. The data were sourced from the China Statistical Yearbook, the China Forestry and Grassland Statistical Yearbook (formerly the China Forestry Statistical Yearbook), the China Environmental Statistical Yearbook, the China Rural Statistical Yearbook and the 60 Years Statistical Yearbook of New China’s Agricultural Industry. Missing data for individual provinces were addressed through interpolation and simple linear trend extrapolation.

2.2. Methods

2.2.1. DDF-GML Models with Non-Desired Outputs

When assessing FGTFP, the traditional DEA model cannot measure the efficiency of non-desired outputs. It also fails to align with the real development process of forestry and does not adequately account for the slackness of inputs and outputs [36]. In response, this study incorporated the impact of non-desired outputs by combining the global technology frontier, which includes these outputs, with a non-radial, non-angular slack-variable-based directional distance function (DDF) to more precisely measure FGTFP [37,38]. DDF can be expressed as the following linear programming model:

$${\mathrm{D}}^{\mathrm{G}}\left({\mathrm{X}}^{\mathrm{t}},{\mathrm{Y}}^{\mathrm{t}},{\mathrm{B}}^{\mathrm{t}},{\mathrm{g}}^{\mathrm{t}}\right)=\mathrm{M}\mathrm{a}\mathrm{x}\mathsf{\beta }$$

(1)

$$\mathrm{s}.\mathrm{t}.\left\{\begin{array}{c}{\displaystyle \sum _{\mathrm{t}=1}^{\mathrm{T}}}{\displaystyle \sum _{\mathrm{j}=1}^{\mathrm{n}}}{\mathsf{\lambda }}_{\mathrm{j}}^{\mathrm{t}}{\mathrm{X}}_{\mathrm{i}\mathrm{j}}^{\mathrm{t}}\le {\mathrm{X}}_{\mathrm{i}0}^{\mathrm{k}}-\mathrm{\beta }{\mathrm{g}}_{\mathrm{i}0}^{\mathrm{k}\mathrm{X}},\mathrm{i}=\mathrm{1,2},\cdots ,\mathrm{m}\\ {\displaystyle \sum _{\mathrm{t}=1}^{\mathrm{T}}}{\displaystyle \sum _{\mathrm{j}=1}^{\mathrm{n}}}{\mathsf{\lambda }}_{\mathrm{j}}^{\mathrm{t}}{\mathrm{Y}}_{\mathrm{r}\mathrm{j}}^{\mathrm{t}}\ge {\mathrm{Y}}_{\mathrm{r}0}^{\mathrm{k}}+\mathrm{\beta }{\mathrm{g}}_{\mathrm{r}0}^{\mathrm{k}\mathrm{Y}},\mathrm{r}=\mathrm{1,2},\cdots ,\mathrm{s}\\ {\displaystyle \sum _{\mathrm{t}=1}^{\mathrm{T}}}{\displaystyle \sum _{\mathrm{j}=1}^{\mathrm{n}}}{\mathsf{\lambda }}_{\mathrm{j}}^{\mathrm{t}}{\mathrm{B}}_{\mathrm{l}\mathrm{j}}^{\mathrm{t}}={\mathrm{B}}_{\mathrm{l}0}^{\mathrm{k}}+\mathrm{\beta }{\mathrm{g}}_{\mathrm{l}0}^{\mathrm{k}\mathrm{B}},\mathrm{l}=\mathrm{1,2},\cdots ,\mathrm{z}\\ {\mathsf{\lambda }}_{\mathrm{j}}\le 0, \mathrm{j}=\mathrm{1,2},\cdots ,\mathrm{n}, \mathrm{t}=\mathrm{1,2},\cdots ,\mathrm{l}\end{array}\right.$$

$${\mathrm{D}}^{\mathrm{t}}\left({\mathrm{X}}^{\mathrm{t}},{\mathrm{Y}}^{\mathrm{t}},{\mathrm{B}}^{\mathrm{t}},{\mathrm{g}}^{\mathrm{t}}\right)=\mathrm{M}\mathrm{a}\mathrm{x}\mathsf{\alpha }$$

(2)

$$\mathrm{s}.\mathrm{t}.\left\{\begin{array}{c}{\displaystyle \sum _{\mathrm{j}=1}^{\mathrm{n}}}{\mathsf{\lambda }}_{\mathrm{j}}^{\mathrm{t}}{\mathrm{X}}_{\mathrm{i}\mathrm{j}}^{\mathrm{t}}\le {\mathrm{X}}_{\mathrm{i}0}^{\mathrm{t}}-\mathrm{\alpha }{\mathrm{g}}_{\mathrm{i}0}^{\mathrm{t}\mathrm{X}},\mathrm{i}=\mathrm{1,2},\cdots ,\mathrm{m}\\ {\displaystyle \sum _{\mathrm{j}=1}^{\mathrm{n}}}{\mathsf{\lambda }}_{\mathrm{j}}^{\mathrm{t}}{\mathrm{Y}}_{\mathrm{r}\mathrm{j}}^{\mathrm{t}}\ge {\mathrm{Y}}_{\mathrm{r}0}^{\mathrm{t}}+\mathrm{\alpha }{\mathrm{g}}_{\mathrm{r}0}^{\mathrm{t}\mathrm{Y}},\mathrm{r}=\mathrm{1,2},\cdots ,\mathrm{s}\\ {\displaystyle \sum _{\mathrm{j}=1}^{\mathrm{n}}}{\mathsf{\lambda }}_{\mathrm{j}}^{\mathrm{t}}{\mathrm{B}}_{\mathrm{l}\mathrm{j}}^{\mathrm{t}}={\mathrm{B}}_{\mathrm{l}0}^{\mathrm{t}}+\mathrm{\alpha }{\mathrm{g}}_{\mathrm{l}0}^{\mathrm{t}\mathrm{B}},\mathrm{l}=\mathrm{1,2},\cdots ,\mathrm{z}\\ {\mathsf{\lambda }}_{\mathrm{j}}\ge 0, \mathrm{j}=\mathrm{1,2},\cdots ,\mathrm{n}, \mathrm{t}=\mathrm{1,2},\cdots ,\mathrm{l}\end{array}\right.$$

where ${\mathrm{D}}^{\mathrm{G}}\left({\mathrm{X}}^{\mathrm{t}},{\mathrm{Y}}^{\mathrm{t}},{\mathrm{B}}^{\mathrm{t}},{\mathrm{g}}^{\mathrm{t}}\right)$ is the global DDF in period t and ${\mathrm{D}}^{\mathrm{t}}\left({\mathrm{X}}^{\mathrm{t}},{\mathrm{Y}}^{\mathrm{t}},{\mathrm{B}}^{\mathrm{t}},{\mathrm{g}}^{\mathrm{t}}\right)$ is the current DDF in period t.

Since the Malmquist–Luenberger index has not resolved the unsolvable problem or the potential acyclic issue arising from intertemporal efficiency measurement, this paper adopted the Global Malmquist–Luenberger productivity index (GMLPI), as suggested by Oh [39], to measure FGTFP. The formula is as follows:

$$\begin{array}{l}\mathrm{G}\mathrm{M}\mathrm{L}\mathrm{P}\mathrm{I}\left({\mathrm{X}}^{\mathrm{t}},{\mathrm{Y}}^{\mathrm{t}},{\mathrm{g}}^{\mathrm{t}},{\mathrm{X}}^{\mathrm{t}+1},{\mathrm{Y}}^{\mathrm{t}+1},{\mathrm{g}}^{\mathrm{t}+1}\right)={\displaystyle \frac{1+{\mathrm{D}}^{\mathrm{G}}\left({\mathrm{X}}^{\mathrm{t}},{\mathrm{Y}}^{\mathrm{t}},{\mathrm{g}}^{\mathrm{t}}\right)}{1+{\mathrm{D}}^{\mathrm{G}}\left({\mathrm{X}}^{\mathrm{t}+1},{\mathrm{Y}}^{\mathrm{t}+1},{\mathrm{g}}^{\mathrm{t}+1}\right)}}\\ ={\displaystyle \frac{1+{\mathrm{D}}^{\mathrm{t}}\left({\mathrm{X}}^{\mathrm{t}},{\mathrm{Y}}^{\mathrm{t}},{\mathrm{g}}^{\mathrm{t}}\right)}{1+{\mathrm{D}}^{\mathrm{t}+1}\left({\mathrm{X}}^{\mathrm{t}+1},{\mathrm{Y}}^{\mathrm{t}+1},{\mathrm{g}}^{\mathrm{t}+1}\right)}}\times \left[{\displaystyle \frac{\left(1+{\mathrm{D}}^{\mathrm{C}}\left({\mathrm{X}}^{\mathrm{t}},{\mathrm{Y}}^{\mathrm{t}},{\mathrm{g}}^{\mathrm{t}}\right)\right)/\left(1+{\mathrm{D}}^{\mathrm{t}}\left({\mathrm{X}}^{\mathrm{t}},{\mathrm{Y}}^{\mathrm{t}},{\mathrm{g}}^{\mathrm{t}}\right)\right)}{\left(1+{\mathrm{D}}^{\mathrm{C}}\left({\mathrm{X}}^{\mathrm{t}+1},{\mathrm{Y}}^{\mathrm{t}+1},{\mathrm{g}}^{\mathrm{t}+1}\right)\right)/\left(1+{\mathrm{D}}^{\mathrm{t}+1}\left({\mathrm{X}}^{\mathrm{t}+1},{\mathrm{Y}}^{\mathrm{t}+1},{\mathrm{g}}^{\mathrm{t}+1}\right)\right)}}\right]={\mathrm{E}\mathrm{C}}^{\mathrm{t},\mathrm{t}+1}\times {\mathrm{T}\mathrm{I}}^{\mathrm{t},\mathrm{t}+1}\end{array}$$

(3)

where ${\mathrm{X}}^{\mathrm{t}}$ represents inputs at time t, ${\mathrm{Y}}^{\mathrm{t}}$ represents outputs at time t, ${\mathrm{g}}^{\mathrm{t}}$ represents non-desired outputs at time t, ${\mathrm{E}\mathrm{C}}^{\mathrm{t},\mathrm{t}+1}$ denotes the change in efficiency from period t to $\mathrm{t}+1$, and ${\mathrm{T}\mathrm{I}}^{\mathrm{t},\mathrm{t}+1}$ denotes the degree of Technological Improvement away from the best-practice frontier from period t to $\mathrm{t}+1$. An EC greater than 1 denotes an increase in efficiency, a TI greater than 1 denotes technological progress and a GMLPI greater than 1 denotes FGTFP growth.

2.2.2. Dagum Gini Coefficient

To examine the spatial divergence and sources of FGTFP in China, this study employed the Gini coefficient decomposition method proposed by Dagum [23]. According to this method, the overall spatial divergence $\mathrm{G}$, is composed of region divergence ${\mathrm{G}}_{\mathrm{w}}$; interregional divergence ${\mathrm{G}}_{\mathrm{n}\mathrm{b}}$; and hypervariance density ${\mathrm{G}}_{\mathrm{t}}$, with the relationship $\mathrm{G}={\mathrm{G}}_{\mathrm{w}}+{\mathrm{G}}_{\mathrm{n}\mathrm{b}}+{\mathrm{G}}_{\mathrm{t}}$. The formula for the Gini coefficient is as follows:

$$\mathrm{G}={\displaystyle \frac{\sum _{\mathrm{j}=1}^{\mathrm{k}}\sum _{\mathrm{h}=1}^{\mathrm{k}}\sum _{\mathrm{i}=1}^{{\mathrm{n}}_{\mathrm{j}}}\sum _{\mathrm{r}=1}^{{\mathrm{n}}_{\mathrm{h}}}\left|{\mathrm{y}}_{\mathrm{j}\mathrm{i}}-{\mathrm{y}}_{\mathrm{h}\mathrm{r}}\right|}{2{\mathrm{n}}^2\overline{\mathrm{y}}}}$$

(4)

where k represents the number of regions, n is the total number of provinces, ${\mathrm{n}}_{\mathrm{j}}\left({\mathrm{n}}_{\mathrm{h}}\right)$ is the number of provinces in the region j(h), ${\mathrm{y}}_{\mathrm{j}\mathrm{i}}\left({\mathrm{y}}_{\mathrm{h}\mathrm{r}}\right)$ is the FGTFP of the provinces i(r) in the region j(h) and $\overline{\mathrm{y}}$ is the mean value of the FGTFP of all provinces. A larger Gini coefficient indicates a higher degree of spatial divergence of FGTFP. For the specific decomposition formula, refer to the study by Dagum [23].

2.2.3. Geo-Detector

From the perspective of spatial divergence, the geo-detector was employed to assess the relationship between geographic objects. The core assumption is that the spatial distributions of explanatory and dependent variables are similar in a study if an explanatory variable significantly influences a dependent variable [40]. The four primary detectors include factor detection, interaction detection, risk detection and ecological detection. In this study, interaction detection was employed to analyze the driving factors of FGTFP, EC and TI. In interaction detection, the q-value indicates the magnitude of an influencing factor’s explanatory power, with values ranging from 0 to 1. A larger q-value suggests greater explanatory power, while a smaller q-value indicates less explanatory power. The formula is as follows:

$$\mathrm{q}=1-{\displaystyle \frac{\sum _{\mathrm{h}=1}^{\mathrm{L}}{\mathrm{N}}_{\mathrm{h}}{{\mathsf{\sigma }}_{\mathrm{h}}}^2}{\mathrm{N}{\mathsf{\sigma }}^2}}$$

(5)

where $\mathrm{h}=1,...,\mathrm{L}$ is the stratification or partition of variables X, $\mathrm{N}\left({\mathrm{N}}_{\mathrm{h}}\right)$ is the sample size of the region as a whole (stratum h) and ${\mathsf{\sigma }}^2\left({{\mathsf{\sigma }}_{\mathrm{h}}}^2\right)$ is the variance of FGTFP for the region as a whole (stratum h).

3. Results

3.1. Analysis of FGTFP

As shown in

Table 3. FGTFP and its decomposition in China from 2004 to 2022.

Year	FGTFP	EC	TI	East	Northeast	Central	West
2004–2005	0.970	0.994	0.978	0.936	0.974	1.001	0.984
2005–2006	1.005	0.996	1.010	1.032	0.991	0.996	0.990
2006–2007	0.991	0.961	1.036	0.977	1.007	1.002	0.994
2007–2008	1.007	1.041	0.970	1.023	0.983	1.005	0.999
2008–2009	0.994	1.012	0.988	0.971	1.015	1.034	0.987
2009–2010	1.027	0.990	1.038	1.087	0.994	0.999	0.997
2010–2011	1.042	1.076	0.989	1.059	1.009	1.074	1.019
2011–2012	0.999	1.010	1.000	1.018	0.989	0.978	0.995
2012–2013	1.030	1.014	1.024	1.044	1.034	1.022	1.021
2013–2014	1.020	0.994	1.029	1.035	1.001	1.026	1.010
2014–2015	1.013	1.005	1.009	1.000	1.007	1.023	1.020
2015–2016	1.011	0.995	1.017	1.012	0.992	1.032	1.004
2016–2017	1.014	0.995	1.028	1.016	1.010	1.035	1.002
2017–2018	1.040	1.005	1.039	1.042	1.040	1.042	1.038
2018–2019	1.106	1.032	1.073	1.080	1.031	1.090	1.160
2019–2020	1.002	0.993	1.012	1.022	1.018	0.973	0.995
2020–2021	0.948	0.973	0.974	0.970	0.950	0.989	0.904
2021–2022	1.017	0.982	1.037	1.015	1.053	1.012	1.013
Average	1.013	1.004	1.014	1.019	1.005	1.018	1.007

, FGTFP in China shows an overall upward trend from 2004 to 2022, though the growth rate is modest, with an average annual growth rate of 0.3%. The average annual growth rate of TI during the same period is 0.4%, while EC remains dynamically stable throughout.

As shown in Figure 2, the trends of TI and FGTFP are generally consistent; TI is the primary driver of FGTFP growth. Specifically, 2006, 2007 and 2008 mark the turning points of positive growth for EC, TI and FGTFP, respectively. This may be attributed to the growing prominence of issues arising from the 2003 reform of the collective forest rights system, such as the abnormal concentration of forest rights and the phenomenon of “the cultivator losing his own land”, which delayed the expected effects of the reform. The largest increase in EC occurred in 2009–2010, reaching 8.7%. This was likely due to the long production cycle in forestry and the gradual impact of policies introduced in the Eleventh Five-Year Plan after 2008.

These policies reflect the country’s increased investment in forestry and its focus on promoting the technologization and informatization of the sector, thereby enhancing production efficiency. After 2014, TI increased steadily, reflecting the continued adoption of low-carbon technologies under the constraint of carbon emissions. However, in 2020–2021, TI dropped to 0.974, likely due to the negative impacts of natural disasters, such as low-temperature rain, snow and ice, as well as the economic crisis. FGTFP declined significantly from 2018 to 2021, likely due to the worsening negative externalities at forest production sites and increasing conflicts in the forest rights trading market. These factors led to lower resource utilization efficiency, which in turn impacted FGTFP. Since 2021, China’s forestry industry has steadily moved towards the goal of high-quality development, with the green development model under the constraint of carbon emissions gradually showing its positive effects.

At the interprovincial level, Jiangsu exhibits the highest average FGTFP, with TI as the primary driver of its growth. Following Jiangsu, Zhejiang has the next highest level of FGTFP, with TI being the main source of its growth. Jiangsu has actively advanced the reform of collective forest rights; enhanced support for the forestry science and technology sector; fostered the integration of industry, academia, research and utilization; and guided the integrated development of primary, secondary and tertiary industries in forestry. Zhejiang has implemented the “Pilot Project on Ecological Protection and Restoration of Mountains, Waters, Forests, Fields, Lakes and Grasses”, reinforced green regulations and mobilized public participation to promote sustainable forestry development. Tianjin exhibits the lowest FGTFP, due to its economic structure, which is dominated by heavy industry. This lack of financial support and market incentives hampers large-scale, intensive forestry operations, thereby affecting the improvement of FGTFP. At the regional level, FGTFP varies across regions, with the lowest observed in the northeast, followed by the western, central and eastern regions, in that order. The average annual growth rates of FGTFP were 0.5%, 0.2% and 0.1% in the eastern, western and central regions, respectively, while the growth rate in the northeastern region declined. In the eastern region, the high level of forestry economic development, combined with the gradual implementation of low-carbon development strategies, has accelerated the green transformation of the forestry industry. FGTFP in the central region has remained stable, likely due to the stimulation of forestry development potential from the deepening of collective forest area reform in the southern part of the region. Despite having a foundation of forest resources, the northeastern region faces challenges in institutional reform and slow technological progress, which have hindered the growth of FGTFP (

Table 4. FGTFP and its decomposition in different regions from 2004 to 2022.

Regions	FGTFP	Rankings	EC	Rankings	TI	Rankings
Beijing	1.006	20	0.994	29	1.014	16
Tianjin	0.990	30	0.994	30	1.011	19
Hebei	1.015	16	1.001	14	1.024	5
Shanxi	1.005	21	1.021	2	1.016	12
Inner Mongolia	0.992	29	0.995	28	0.999	30
Liaoning	1.010	18	1.005	9	1.013	18
Jilin	1.002	25	1.000	16	1.002	27
Heilongjiang	1.004	22	1.001	15	1.005	21
Shanghai	1.028	4	1.000	16	1.028	4
Jiangsu	1.036	1	1.000	16	1.036	1
Zhejiang	1.030	2	1.000	16	1.030	3
Anhui	1.028	3	1.017	4	1.016	13
Fujian	1.015	15	1.000	16	1.015	14
Jiangxi	1.023	7	1.006	7	1.019	9
Shandong	1.026	5	0.997	27	1.031	2
Henan	1.016	13	1.006	8	1.017	11
Hubei	1.019	11	1.008	6	1.013	17
Hunan	1.019	10	1.005	10	1.015	15
Guangdong	1.023	6	1.000	24	1.024	6
Guangxi	1.021	8	1.000	16	1.021	7
Hainan	1.018	12	0.999	25	1.021	8
Chongqing	1.014	17	1.014	5	1.005	22
Sichuan	1.004	23	1.000	16	1.004	23
Guizhou	1.016	14	1.019	3	1.004	24
Yunnan	1.000	27	1.000	16	1.000	29
Shaanxi	1.003	24	1.001	13	1.002	26
Gansu	1.000	28	0.997	26	1.003	25
Qinghai	1.001	26	1.002	12	1.002	28
Ningxia	1.019	9	1.028	1	1.018	10
Xinjiang	1.009	19	1.004	11	1.007	20
Eastern average	1.019	-	0.998	-	1.024	-
Northeastern average	1.005	-	1.002	-	1.007	-
Central average	1.018	-	1.010	-	1.016	-
Western average	1.007	-	1.006	-	1.006	-
National average	1.013	-	1.004	-	1.014	-

).

As shown in the trend graph in Figure 3, FGTFP has steadily increased across all regions, particularly from 2018 to 2019. This may be attributed to the inclusion of ecological civilization construction in the “Five-in-one” layout at the 18th National Congress of the Communist Party of China, which highlighted the importance of forestry in ecological civilization development, leading to increased regional support for forestry. Notably, FGTFP in the western region increased significantly during 2017 to 2018.

Regarding growth driving factors (Table 4), TI is the primary driver of FGTFP growth in the eastern, northeastern and central regions, contributing to varying degrees in each area. The average value of TI is 1.014, compared to 1.004 for EC, indicating that technological optimization and upgrading are the most critical methods for improving total factor productivity in forestry under the constraint of carbon emissions. Due to the low level of basic technology in the western region, both EC and TI contribute to promoting FGTFP, highlighting the need to improve the efficiency of large-scale forestry operations in the region.

3.2. Spatial Divergence and Sources of FGTFP

The Dagum Gini coefficient and its decomposition method were applied to measure total differences, regional differences, interregional differences and contribution rates of FGTFP in China from 2005 to 2022 (

Table 5. Gini coefficient and contribution rate of FGTFP in China from 2005 to 2022.

Year	Total	Regional Gini Coefficient				Interregional Gini Coefficient						Contribution (%)
		East (E)	Central (C)	West (W)	Northeast (N)	E-C	E-W	E-N	C-W	C-N	W-N	Regional	Interregional	Hypervariable Density
2005	0.028	0.056	0.002	0.013	0.013	0.036	0.039	0.040	0.009	0.014	0.015	28.767	50.674	20.559
2006	0.021	0.044	0.005	0.007	0.004	0.029	0.031	0.031	0.007	0.006	0.006	28.769	43.778	27.453
2007	0.013	0.024	0.001	0.006	0.012	0.015	0.017	0.021	0.005	0.009	0.011	27.633	45.531	26.836
2008	0.012	0.021	0.003	0.004	0.010	0.013	0.015	0.021	0.004	0.011	0.010	27.008	57.786	15.206
2009	0.028	0.048	0.003	0.017	0.003	0.032	0.037	0.029	0.025	0.009	0.017	27.198	44.382	28.420
2010	0.030	0.064	0.005	0.003	0.006	0.045	0.046	0.047	0.005	0.006	0.005	27.163	66.194	6.642
2011	0.035	0.037	0.047	0.021	0.014	0.048	0.036	0.035	0.040	0.039	0.019	25.475	35.862	38.664
2012	0.031	0.024	0.051	0.028	0.005	0.040	0.029	0.021	0.043	0.036	0.019	27.130	25.813	47.057
2013	0.014	0.014	0.009	0.013	0.004	0.015	0.017	0.012	0.011	0.009	0.012	27.217	39.306	33.477
2014	0.017	0.026	0.012	0.009	0.002	0.020	0.021	0.020	0.013	0.013	0.006	27.960	39.669	32.371
2015	0.017	0.028	0.008	0.011	0.006	0.021	0.022	0.019	0.010	0.010	0.010	28.784	30.557	40.659
2016	0.012	0.013	0.010	0.007	0.005	0.015	0.011	0.012	0.015	0.020	0.008	23.853	53.059	23.089
2017	0.017	0.024	0.011	0.009	0.007	0.020	0.019	0.017	0.018	0.015	0.009	26.055	37.990	35.955
2018	0.020	0.028	0.016	0.017	0.003	0.023	0.023	0.019	0.017	0.012	0.012	30.193	4.160	65.647
2019	0.065	0.062	0.015	0.089	0.026	0.045	0.080	0.051	0.063	0.035	0.072	30.885	32.230	36.885
2020	0.039	0.064	0.032	0.016	0.011	0.057	0.047	0.043	0.027	0.029	0.016	27.521	25.787	46.692
2021	0.085	0.135	0.016	0.072	0.007	0.096	0.114	0.094	0.053	0.021	0.050	29.859	23.238	46.902
2022	0.035	0.066	0.007	0.018	0.015	0.041	0.045	0.048	0.014	0.021	0.025	29.319	11.650	59.031
Average	0.029	0.043	0.014	0.020	0.009	0.034	0.036	0.032	0.021	0.018	0.018	27.822	37.093	35.086

). This analysis aimed to reveal the sources of the spatial divergence of FGTFP in China.

From 2005 to 2022, the total Gini coefficient of FGTFP in China increased from 0.028 to 0.035. Regarding the regional Gini coefficient, the eastern region exhibited the greatest spatial divergence, with an average Gini coefficient of 0.043 during the examination period. This was followed by the western region (0.02), the central region (0.014) and the northeastern region (0.009), which displayed smaller degrees of spatial divergence.

As shown in Figure 4, the spatial divergence of FGTFP in China generally follows a fluctuating upward trend, followed by a gradual decline. Notably, the Gini coefficient fluctuates most in the eastern region, reaching a peak of 0.135 in 2021, while changes in the northeastern region are smaller. This indicates that FGTFP in China exhibited clear spatial divergence during the examination period. The overall regional variation trend of FGTFP shows consistency, beginning with a significant widening trend in 2008, followed by a narrowing phase, a stable period from 2013 to 2018, and then a gradual decline after a sharp rise in 2019. The spatial divergence in the western and eastern regions showed a more obvious expansion trend. This may be attributed to the 2015 issuance of the state-owned Forestry Farm Reform Programme and the state-owned Forestry Reform Guidelines by the CPC Central Committee and the State Council. Additionally, with the strong east wind of deepening the reform of the ecological civilization system, the total output value of the national forestry industry exceeded CNY 7 trillion for the first time in 2017, so the spatial divergence expanded significantly.

According to the interregional Gini coefficients in Table 5 and Figure 5, the greatest divergence during the study period occurred between the eastern and western regions, with an average Gini coefficient of 0.036. This was followed by the eastern and central regions (0.034) and the eastern and northeastern regions (0.032). The smallest spatial divergence was observed between the western and northeastern regions and the central and northeastern regions, with an average Gini coefficient of 0.018. In terms of trends, spatial divergence of FGTFP across regions showed fluctuating increases followed by a slight decline. The Gini coefficient between the eastern and western regions increased the most, peaking at 0.114 in 2021, while the Gini coefficient between the central and northeastern regions showed the least increase.

Figure 6 illustrates the evolution of the contributions of various sources of spatial divergence in FGTFP. The contribution rate of interregional spatial divergence is the largest, followed by the hypervariable density contribution rate. The contribution rate of regional spatial divergence is the smallest. Regional spatial divergence measures FGTFP regional disparities across provinces within the four regions, interregional spatial divergence assesses the gap between high and low FGTFP average levels across regions, and hypervariable density gauges the degree of cross-cluster overlap of outliers among regions [40].

In terms of change magnitude, the hypervariable density contribution rate increases annually, and the contribution rate of interregional spatial divergence decreases. This indicates that, with the development of the forestry industry, the primary source of spatial divergence in FGTFP has shifted from interregional to hypervariable density. This shift suggests an increasing overlap in FGTFP among regions.

3.3. Driving Factors of Spatial Divergence of FGTFP in China

Based on the studies by Li et al. [41] and Yu et al. [42], FGTFP, EC and TI in China experienced a significant turnaround in 2011. Consequently, four time points (2005, 2011, 2017 and 2022) were selected. The numerical variables of the driving factors were transformed into discrete variables using the K-means clustering method. The influence of each factor on the spatial divergence of FGTFP was calculated separately using GeoDetector2015 interaction detection. The results of the interaction detection for each factor are shown in Figure 7.

Overall, the interaction effects between green energy transition factors and external environmental factors show significant differentiation. The interaction effect between green energy transition factors on the dependent variable has a higher q-value than the single-factor effect. Additionally, the two-by-two interaction among the three factors (energy consumption, clean energy structure and forestry industry structure) are stronger than any single-factor effect. This interaction consistently demonstrates a “two-factor enhancement effect” and a “non-linear enhancement trend”, reflecting the synergy of policies and the rational allocation of resources. External environmental factors exert a strong interaction effect on the dependent variable but exhibit a “non-linear weakening trend”. Over time, the impact of these factors on the dependent variable shows a marginal diminishing effect, slightly weakening their influence.

In 2005, the interaction between clean energy structure and forestry industry structure had the strongest impact on all variables. The enactment of the Renewable Energy Law in China has accelerated the development of clean energy. The optimization of the clean energy structure involves the widespread application of energy technology innovations in forestry production, which drives the transformation and upgrading of the industrial structure to meet the demand for more efficient and cleaner energy. In 2011, the interaction between energy consumption and clean energy structure had the strongest impact on all variables. China’s economy was experiencing rapid growth, accompanied by a significant rise in energy demand. Concurrently, the application of clean energy technologies, such as hydroelectric power generation and biomass utilization, in forestry production became more widespread, helping to reduce dependence on traditional fossil fuels. In 2017 and 2022, energy consumption ∩ clean energy structure had the greatest impact on EC, and clean energy structure ∩ forestry industry structure had the greatest impact on TI. The continuous advancement of clean energy technologies has significantly improved energy utilization efficiency. The transformation of the energy structure has driven technological advancements, significantly improving the automation and intelligence of forestry production. Moreover, the optimization of the clean energy structure has facilitated key breakthroughs in renewable energy technologies, smart grids and energy storage. These breakthroughs have become the driving force behind technological innovation and industrial upgrading.

4. Discussion

This paper examines the spatial divergence and driving factors of FGTFP under the constraint of carbon emissions, transitioning from qualitative to quantitative analysis, with the aim of identifying pathways for enhancing FGTFP.

(1) This paper measured the FGTFP and its decomposition index in China from 2004 to 2022 using the DDF-GML models. The study found that, over the time period from 2004 to 2022, FGTFP in China generally showed an upward trend. The results of this study align with previous research on the changes in FTFP in China [12], suggesting that the industry is steadily progressing toward the goal of high-quality development. At the provincial level, FGTFP shows significant heterogeneity. Jiangsu has the highest growth rate, followed by Zhejiang, while Tianjin has the lowest growth rate. At the regional level, FGTFP exhibits a decreasing trend from east–central–west, with the northeastern region exhibiting the lowest value. This aligns with numerous related studies both domestically and internationally, further confirming that the highest levels of FGTFP are found in the eastern region [43]. Regarding driving factors of growth, TI is the primary source of FGTFP growth in the east, northeast and central regions, while EC is the main driver in the western region. In contrast to previous studies, which used the Malmquist index to decompose the growth drivers of FTFP into technological progress (TP) and technical efficiency (TE) [4], this paper employed the GML index to decompose FGTFP into EC and TI, with the goal of clarifying the respective roles of technology and efficiency.

(2) This paper used the Dagum Gini coefficient to show that FGTFP in China exhibits significant spatial divergence characteristics. ① At the regional level, the degree of spatial divergence of FGTFP is expanding across the eastern, central, western and northeastern regions. ② At the interregional level, the greatest degree of differentiation can be observed between the east and west, while a smallest degree of differentiation can be observed between the central and northeast regions. Spatial divergence between regions shows a fluctuating increasing trend, with the biggest increase observed between the west and northeastern regions. ③ The primary source of spatial divergence in FGTFP is hypervariable density, which continues to exhibit dynamic growth, followed by the contribution of interregional spatial divergence, with regional spatial divergence contributing the least. The results of this study align with those on GTFP in the National High-Tech Zones, both indicating that the source of regional divergence is primarily hypervariable density [44]. Scholars typically employ the Dagum Gini coefficient to examine the sources of regional differences in GTFP across sectors such as agriculture [45], logistics [46] and the marine economy [47]. However, this paper extended the analysis of regional differences in GTFP to forestry, providing a foundation for subsequent investigation into the factors influencing the spatial divergence of FGTFP. Additionally, the paper recommends that future high-quality forestry development should prioritize the coordination of resources across regions.

(3) This paper used the geo-detector-based interaction factor detector to identify significant divergence in the internal interaction effects between green energy transition factors and external environmental category factors. The impact of green energy transition factors on the dependent variable consistently showed a “two-factor enhancement effect” and a “non-linear enhancement trend”, while the effect of external environmental factors showed a strong interaction but exhibited a “non-linear weakening trend”. Overall, during the study period, the interaction between clean energy structure and forestry industry structure, as well as the interaction between energy consumption and clean energy structure, had a strong impact on the variables. A comparison with previous studies revealed that scholars have used the spatial Durbin model [48] and panel Tobit model [49] to analyze the factors influencing GTFP. In contrast, this paper employed a geo-detector to examine the factors influencing the spatial divergence of FGTFP from a spatial perspective. Regarding research focus, prior studies have analyzed the impacts of climate change [50], the digital economy [51] and policy implementation [52] on FTFP, typically exploring the influencing factors in isolation. In contrast, this paper utilized an interactive factor detector to investigate the combined effects of green energy transition factors and external environmental factors on FGTFP within the context of carbon emission constraints.

This paper has the following limitations: First, the study focused on the macro-provincial scale. Refining the research scale through field research and interviews could improve the accuracy of policy implementation. Second, due to limitations regarding statistical data, the system of input–output variables requires further improvement, and the methodology for accounting for carbon emissions needs to be enhanced.

5. Conclusions

FGTFP is a key indicator of the quality and efficiency of forestry development. Assessing FGTFP helps identify pathways for high-quality forestry development and promotes the green transformation of the forestry development model. This paper examines the degree of spatial and temporal divergence of FGTFP and its drivers under the constraint of carbon emissions using the DDF-GML models, the Dagum Gini coefficient and the geo-detector. The results show that FGTFP in China generally trended upward from 2004 to 2022. TI was the primary driver of FGTFP growth in the eastern, northeastern and central regions, while EC was the key driver in the western region. FGTFP exhibited distinct spatial divergence patterns in China, with hypervariable density as the primary source. Green energy transition factors consistently showed a significant “two-factor enhancement effect” and a “non-linear enhancement trend”, while external environmental factors exhibited strong interaction effects but demonstrated a “non-linear weakening trend”. The conclusions of this paper are very policy-oriented and are intended to inform future forestry policy development and the realization of China’s dual-carbon goals. In the process of policy implementation, the following aspects should be noted:

(1) To address the regional divergence in the growth of FGTFP across China, local governments should enhance technological innovation support in the western region, promote green technology development and application, and improve technological advancement. In the eastern, northeastern and central regions, the focus should be on precision management to enhance efficiency. In the context of reforms to the collective forest rights system and the strategy for regional coordinated development, each region should create shared and collaborative development plans tailored to their resource endowments, promoting the formation of a “two-wheel-driven” Chinese development model focused on technology and efficiency.

(2) Considering the significant spatial divergence of FGTFP in China, local governments should establish a collaborative forestry network linking the four major regions and promote joint research on key common production technologies. Simultaneously, they should establish a regional compensation mechanism for forestry carbon trading and implement the “integration of forestry factor markets” initiative. Utilizing digital technology, a national dynamic forestry resources monitoring platform has been established to enable intelligent resource sharing. Differentiated policies have been designed to enable a regional division of labor in the innovation chain, spanning basic research, technology development and achievement transformation, while fostering a gradient collaborative innovation system.

(3) The influence of external environmental factors on FGTFP exhibits a “non-linear weakening trend”. Therefore, local governments should promote the coordinated development of nature–economy–society coupling, establish a two-way management system for “ecologically sensitive areas and economic development zones”, and facilitate the virtuous cycle of ecological protection and economic development; introduce policies to implement regional ecological replenishment mechanisms, accelerate the development of forest carbon trading markets and promote the realization of the value of forest ecological products; adjust and improve the regional policy system, promote the efficient flow and concentration of resources, stimulate industrial upgrading momentum, and accelerate the establishment of a power system for high-quality forestry development; and establish a three-tier collaborative governance platform of “government–enterprise–community” to enhance public participation and create a favorable policy environment for high-quality forestry development.