Facilitation or Inhibition? Aging Rural Labor Force and Forestry Economic Resilience: Based on the Perspective of Production Factors

Abstract

Globally, the accelerating aging of the rural labor force is profoundly impacting the economic resilience of the labor-intensive forestry sector. However, the intrinsic connection between the two has not been fully understood and requires further exploration. As the most populous nation globally and a top producer, trader, and consumer of forest products, China stands out as a perfect case study for this issue. Based on this, this study utilizes panel data from 30 provinces in China from 2012 to 2022 and employs a dual machine learning model to empirically examine the impact and mechanisms of rural labor force aging on forestry economic resilience from the perspective of production factors. The findings indicate: (1) overall, the increase in rural labor force aging significantly inhibits forestry economic resilience; (2) rural labor force aging enhances forestry economic resilience by promoting large-scale forest land management, driving forestry technological innovation, and increasing government capital investment; it also inhibits forestry economic resilience by reducing educational human capital and health human capital; (3) the rural force aging exerts a marked adverse effect on the resilience of the forestry economy in the eastern and central regions, major grain-producing areas, and major grain-consuming areas. Based on this, this study proposes policy recommendations in three areas: building a flexible and diversified labor supply and replacement system, exploring a “scale and technology” integration path suited to national conditions, and implementing differentiated regional strategies. The aim is to provide a reference for government departments in formulating strategies to enhance the resilience of the forestry economy in the era of aging.

1. Introduction

Forestry economics, as the core safeguard for global ecological security and a key pillar of the green economy, exerts a vital function in maintaining global carbon cycle balance and biodiversity conservation. It also supports the livelihoods of billions of people and global economic resilience, making it a common agenda for countries striving to achieve sustainable development goals. However, under the compound pressures of a restructuring global economic landscape, frequent extreme weather events, and tightening ecological constraints, the resilience of forestry economics faces multiple uncertainties and challenges. In this context, conducting systematic research on forestry economic resilience is not only a theoretical necessity for breaking through industrial development bottlenecks and enhancing its risk resistance capabilities, but also a practical need for stabilizing farmers’ livelihoods and promoting regional economic coordination. It holds profound theoretical value and practical significance for establishing a sustainable development system for the forestry industry and achieving synergistic benefits between economic and ecological outcomes.

From a theoretical perspective, forestry economic resilience refers to the process by which the forestry economic system self-recovers, adjusts to changes, and redefines its growth trajectory after being subjected to uncertain shocks [1]. Being a typically labor-intensive sector, forestry development depends far more heavily on labor resources compared to other industries. Nevertheless, the tendency toward an aging labor force has grown especially prominent in recent years, emerging as a global social phenomenon [2]: In developed countries like Japan, as of January 2024, the share of people aged 65 and over neared 30%, posing a threat to the continuity of traditional forestry production models. In developing countries like India and Brazil, meanwhile, rapid urbanization has driven young and middle-aged rural workers to migrate to cities [3]. This migration has caused the rural labor force to age structurally and brought about significant shifts in the labor supply scenario for labor-intensive sectors like forestry [4]. Changes in the labor force will inevitably pose potential risks to the stable operation of the forestry economic system.

China, the world’s most populous nation, is also the globe’s top producer, trader, and consumer of forest products. This unique “double first” status makes it an ideal case study for exploring the relationship between rural labor force aging and forestry economic resilience. On the one hand, as rural urbanization progresses in China, the rural labor force structure is showing a clear trend toward aging. Based on figures from the seventh national census, the share of rural residents aged 65 and over stood at 17.72% in 2020, 6.61 percentage points higher than that in urban areas [5]. Ten years ago, in the sixth census, these figures were 10.06% and 2.26 percentage points, respectively. As a labor-intensive industry, forestry is particularly sensitive to shifts in the labor force structure. On the other hand, China is rich in forestry resources, has a complete industrial system, and engages in large-scale trade of forest products. The dynamic changes in the resilience of its forestry economy exert a notable influence on the global forestry economy.

Against this backdrop, this research uses panel data from 30 Chinese provinces spanning 2012 to 2022, along with a dual machine learning model, to empirically explore the mechanism by which rural labor force aging affects forestry economic resilience. Its purpose is to offer insights for formulating strategies to strengthen forestry economic resilience in the era of labor force aging.

2. Literature Review

Reviewing existing research, the academic community has explored rural labor force aging and forestry economic resilience from multiple disciplinary perspectives, including ecology, geography, and economics. In terms of research content, studies have covered both macro-level trend analyses [6,7] and micro-level mechanism explorations [8,9].

Regarding rural labor force aging, numerous studies have focused on its impact on agricultural production systems. From an economic perspective, scholars generally believe that rural labor force aging exerts significant adverse impacts on agricultural green total factor productivity [10], land transfer efficiency [11], and production efficiency [12] through pathways such as reduced labor supply [8], rigid skill structures [13], and insufficient capital investment [14]. However, recent research has also sparked controversy. Some scholars point out that aging may promote the popularization of agricultural mechanization through a “backward-forcing mechanism” or lead to the concentration of land in new types of operating entities, thereby indirectly enhancing scale efficiency [15,16]. Additionally, Several researches have explored the potential effect of aging on the transformation of agricultural production methods, such as promoting the adjustment of agricultural production toward simplification and intelligence, as well as accelerating the improvement of the agricultural socialized service system. These perspectives provide a multi-faceted comprehension of the connection between aging and forestry development, which also has labor-intensive characteristics. However, direct discussions on the association between rural labor force aging and forestry economic resilience are still in their infancy. Two indirect logical chains have been established: first, the foundational impact of labor force aging on forestry production. Scholars primarily employ traditional regression analysis methods, finding that labor force aging leads to “aging and low-skilled” forestry workers, a change that directly reduces efficiency in afforestation, nurturing, logging, and processing, providing micro-level evidence for analyzing the impact of rural aging on forestry economic resilience [17]. Second, research has focused on the transmission mechanisms involving intermediate factors. Some studies have proposed, based on case studies, that extending the industrial chain (such as developing forest-tourism integration and forest-based economy models) can reduce reliance on a single product. Other studies have called for the establishment of a “warning–response–recovery” mechanism across the entire chain [18], providing practical guidance for enhancing the resilience of the forestry economy. Additionally, scholars have adopted methods such as constructing indicator systems, model simulations, and scenario analyses [19] to interpret the meaning of forestry economic resilience. Related studies typically highlight four key dimensions: resistance, adaptability, recovery, and transformation capacity. For example, Sicheng An et al. constructed a forestry economic resilience indicator system based on the pressure–state–response (PSR) framework [20]. Yali Mu et al. evaluated forestry economic resilience from three perspectives: resistance, adaptability, and transformation capacity [1]. However, forestry has both natural ecological and socio-economic characteristics, and its level of economic resilience is affected not only by natural elements like temperature, rainfall, and soil quality but also closely linked to a range of social variables [21,22]. Therefore, the economic resilience level of the forestry sector is essentially a comprehensive reflection of the combined effects of natural environmental conditions and socio-economic factors.

In summary, the existing literature has established a firm basis for this study, yet certain limitations remain: (1) the interpretation of the connotation of forestry economic resilience has not fully incorporated the characteristics of the forestry industry, and the evaluation indicator system lacks specificity. (2) Research on the association between rural labor force aging and forestry economic resilience is still in its preliminary stages, and existing studies often treat labor force aging as an independent variable, neglecting the interdependent effects of production factors, making it difficult to uncover the underlying logic of the impact. (3) In empirical analysis, research methods primarily rely on traditional econometric models such as the two-way fixed-effects model and OLS regression. These models have limited ability to capture nonlinear relationships and high-dimensional interaction effects between variables, and are susceptible to endogeneity issues. In fields like forestry economics, which are influenced by multiple complex factors, the robustness and generalizability of results are therefore insufficient. (4) Research cases exhibit strong regional limitations, with insufficient systematic studies on developing countries, thereby limiting their international reference value.

3. Research Hypotheses

There are significant differences between the rural forestry labor market and the urban formal employment market. In China’s rural forestry sector, which is primarily family based, there is no unified retirement system. Elderly workers, facing limited alternative incomes and inadequate pension coverage, often continue to participate in production even after their physical strength and skills have declined. These “low exit rate” and “low productivity” characteristics make the impact of labor force aging on the forestry economic system more complex. This may directly suppress the resilience of the forestry economy through declining labor quality, or indirectly generate intermediary effects through forcing scale expansion and technological innovation.

3.1. The Direct Impact of Rural Labor Force Aging on the Resilience of the Forestry Economy

The forestry economy encompasses the primary sector (primary production activities such as afforestation, forest management, and logging), the secondary sector (processing activities such as wood processing, human-made board manufacturing, and forest chemical product production), and the tertiary sector (service areas such as eco-tourism, forestry logistics, and carbon credit trading). Its sustainability is highly dependent on the quantity and quality of labor across the entire industrial chain. The aging of the rural labor force directly weakens this resilience in three ways. In the primary sector, the physical decline of older workers leads to reduced efficiency in high-intensity operations [23]. And more critically, when faced with shocks such as pest outbreaks or extreme weather, they are unable to promptly carry out tasks like seedling rescue or forest land protection, significantly reducing the system’s resistance capacity. In the secondary sector, the knowledge structure of the aging workforce is rigid, and their adaptation to precision processing equipment and environmental standards is slower than that of younger workers, leading to a decline in the quality stability of the processing process [24]. Quality fluctuations amplify the impact of market shocks and weaken the system’s recovery capacity. In the tertiary sector, the aging workforce has limited ability to use digital tools, making it difficult to respond to diversified market demands, resulting in insufficient market adaptability of the service industry and constraining the system’s transformation capacity. Therefore, this study proposes hypothesis 1: The deepening aging of the rural labor force will suppress the resilience of the forestry economy.

3.2. The Mediating Effect of Rural Labor Force Aging on Forestry Economic Resilience

3.2.1. Effects of Large-Scale Forest Land Management

• The impact of rural labor force aging on large-scale forest land management

Forest revenue is an important factor influencing changes in forest land ownership [25]. However, the aging of the rural labor force has, to a certain extent, driven the development of forest land toward large-scale and intensive management. From the perspective of income security, elderly forest farmers, due to declining physical strength and reduced risk tolerance, find it difficult to sustain long-term, high-investment forestry production. They tend to transfer their management rights to professional entities through forest land transfers, subleasing, or trusteeship [26], obtaining stable income through rent or profit-sharing. This choice of “transferring land to secure income” directly accelerates forest land consolidation. From the perspective of efficiency improvement, the aging of the labor force has led to rising labor costs. Traditional dispersed management models are no longer sustainable due to excessively high unit costs. New management entities with capital and technological advantages achieve economies of scale through mechanized operations, prompting elderly forest farmers to voluntarily exit the industry and driving forest land toward efficient management units [27].

2.

The impact of large-scale forest land management on the resilience of the forestry economy

The theory of economies of scope argues that if a region concentrates suppliers of labor, related services, raw materials, and sales channels required by a particular industry, it will have a competitive advantage over other regions in the sustainable development of that industry. This is often regarded as the theoretical basis for business entities to adopt diversified business strategies [28]. Based on this, large-scale forest land management is not only about expanding the area, but also about diversifying the content and methods of management across the forestry’s primary, secondary, and tertiary industries. This diversification enhances the resilience of the forestry economy. In the primary sector, large-scale operations can achieve high-quality seedling cultivation, standardized planting, and centralized management. This enhances the stability of primary products such as timber and forest-grown crops and provides reliable raw materials for downstream processing industries. In the secondary sector, large-scale forest land operations facilitate the unified planning and efficient allocation of core elements such as infrastructure, technical capabilities, machinery, and market channels, effectively reducing unit costs and creating internal “synergy effects.” Additionally, this operational structure enhances the external connectivity of operators, making it easier to access financial support, government subsidies, and policy resources, thereby providing stable support for responding to emergencies. In the tertiary sector, large-scale operators can diversify their revenue streams by expanding into forest-based economies, ecological tourism, and carbon trading, reducing reliance on single products and effectively mitigating market and climate risks, thereby achieving both economic and ecological benefits [27].

Therefore, this study proposes hypothesis 2: Rural labor force aging will drive large-scale forestland management and thereby enhance the resilience of the forestry economy.

3.2.2. Labor Supply Quality Effects

• The impact of rural labor force aging on labor supply quality

Rural labor force aging reduces labor supply quality in terms of physical fitness and overall capabilities. On the one hand, as forest farmers age, their physical functions gradually deteriorate, bones become fragile, muscle strength weakens, and sensory abilities decline [29]. This makes it difficult for them to work as efficiently and sustainably as younger laborers when facing high-intensity, long-duration agricultural labor. Additionally, the deterioration of sensory functions such as vision and hearing reduces their commitment to production, weakening work motivation and productivity. On the other hand, elderly forest farmers generally have lower educational attainment and poorer understanding and acceptance of new technologies, making it difficult for them to adapt to the demands of forestry modernization [30]. Furthermore, elderly forest farmers have long relied on traditional experience, with rigid thinking patterns, and are reluctant to try new approaches or solutions. When faced with complex production issues or sudden risks, they lack the flexibility and innovation needed to respond effectively. These factors collectively constrain improvements in forestry labor quality and production efficiency.

2.

The impact of labor supply quality on the resilience of the forestry economy

A decline in labor supply quality threatens the resilience of the forestry economy in two ways: weakened adaptability and insufficient innovation. On the one hand, lower labor quality means that the forestry system lacks sufficient response and recovery capabilities when faced with external shocks such as natural disasters, pests and diseases, and market fluctuations. Forest farmers with poor physical condition are prone to fatigue and injury during high-intensity forestry work, affecting efficiency and safety. Older workers, due to lower educational levels and weaker information-gathering capabilities, struggle to adjust production and management strategies in a timely manner in response to changes in weather, markets, or policies, leading to a significant weakening of the forestry system’s sensitivity and adaptability [31]. On the other hand, the decline in labor supply quality also limits technological progress and the upgrading of production methods in forestry. High-quality labor is a key driver of forestry modernization. However, the current decline in labor quality, coupled with low acceptance of new technologies and insufficient innovation capacity among forest farmers who rely on traditional experience for production. This constrains the forestry economic system’s ability to address long-term, structural risks. Therefore, the decline in labor supply quality not only weakens the forestry system’s external adaptability but also hinders internal innovation-driven development, severely impeding the enhancement of forestry economic resilience.

Thus, this study proposes hypothesis 3: Rural labor force aging reduces labor supply quality and thereby inhibits forestry economic resilience.

3.2.3. Government Capital Investment Effects

• The impact of rural labor force aging on government capital investment

To address the labor supply contradictions and resource allocation challenges posed by rural labor force aging, the government will increase capital investment in the forestry sector. On the one hand, the aging of the rural labor force has placed traditional forestry production models in a double bind of labor loss and declining production efficiency. The core logic of institutional change theory suggests that when the foundational conditions of a socio-economic system—such as factor endowments, technological conditions, and interest structures—undergo profound changes, existing institutional arrangements often become inefficient due to their inability to adapt to new development needs, thereby generating an intrinsic impetus for institutional adjustment and innovation [32]. The challenges posed by the aging of the rural labor force to traditional forestry production fundamentally stem from the changes in the structure of the labor force within the socio-economic environment, which challenge the existing production systems and management models in the forestry sector. Therefore, the government promotes the modernization and scaling up of forestry production methods through financial support and policy guidance. For example, policies incentivize land transfers to encourage young people or enterprises to enter forestry production, thereby enhancing the intensification of forestry operations. On the other hand, from the perspective of public choice theory, government capital investment is both a direct response to the aging of the labor force and a manifestation of safeguarding the public interest. Public choice theory holds that in a market economy, public goods are difficult to supply effectively through market mechanisms due to their non-competitive and non-exclusive characteristics. As the representative of public interests, the government has a responsibility to compensate for market failures through public policies and resource investments to ensure the improvement of overall social welfare [33]. The aging of the rural labor force has led to a general decline in the production capacity and income levels of forest farmers. This has exerted a notable influence on the labor-intensive forestry industry, resulting in inefficient utilization of forest land resources and weak management. In addition, forestry has the attributes of a public good and is related to the public interest of society [34]. When market forces are unable to maintain the effective allocation of forestry resources and the performance of public functions due to changes in the labor force structure, government capital investment becomes a necessary means of intervention. Therefore, government capital investment can both alleviate the shortage of forestry labor and ensure the effective utilization of forest resources and the sustained performance of ecological functions.

2.

The impact of government capital investment on forestry economic resilience

Increasing government capital investment in the forestry industry can enhance its risk-resilience and sustainable development capabilities. On the one hand, traditional forestry production is inefficient and lacks sustainability due to the aging of the labor force and climate change. According to institutional change theory, government capital investment can be viewed as an institutional arrangement. Through fiscal and policy guidance, it promotes technological and management upgrades in forestry, reduces reliance on traditional labor, and improves resource allocation efficiency. This enhances the forestry system’s resilience to natural disasters and market fluctuations. On the other hand, given the ecological externalities and public goods attributes of forestry resources, their ecological functions cannot be effectively allocated through market mechanisms, necessitating government intervention through capital investment. By increasing capital investment, the government can ensure the sustained provision of ecological services and stabilize the foundation of the forestry industry. It can also promote the integrated development of forestry with agriculture, tourism, carbon trading, and other industries, and strengthen cross-industry resource integration and risk-resilience capabilities. These measures enhance the adaptability and stability of the national economic system on a broader scale.

Then, this study proposes hypothesis 4: Rural labor force aging will promote government capital investment and thereby enhance forestry economic resilience.

3.2.4. Effects of Forestry Technological Innovation

• The impact of rural labor force aging on forestry technological innovation

Rural labor force aging will influence forestry technological innovation from both the supply-side factor substitution and demand-side market perspectives. On the supply side, the factor substitution theory posits that when the supply of a particular production factor becomes scarce or its cost rises, economic entities will adjust the proportion of factor combinations, substituting with relatively abundant or lower-cost factors to achieve optimal resource allocation and sustained improvements in production efficiency [35]. This theory reveals the dynamic adaptive relationship between factors and emphasizes the importance of adjusting factor structures to respond to changes in production conditions. In the forestry sector, the aging of the rural labor force has led to a reduction in the quantity and quality of labor supply, directly increasing labor costs in forestry production [36]. To reduce costs and improve efficiency, forest farmers and forestry enterprises must increase their investment in technological innovation, using advanced technologies such as mechanization and automation to partially replace manual labor. Additionally, the preference of aging forest farmers for simple biochemical technologies also points to new directions for technological innovation. On the demand side, according to the induced innovation theory, the aging of the labor force leads to fluctuations in the quantity and quality of forest products, weakening market competitiveness. At the same time, consumers’ diverse demands for forest product quality and supply shortages jointly shape a new market landscape, forcing and driving forestry technological innovation.

2.

The impact of forestry technological innovation on forestry economic resilience

Forestry technological innovation enhances forestry economic resilience by reducing production costs and improving the industry’s risk resistance capabilities. On the one hand, based on the theory of factor substitution, technological innovation can replace traditional production factors, improve production efficiency, and reduce reliance on labor and natural resources [37]. For example, forest harvesting machinery and drone technology can significantly improve operational efficiency, while precision forestry technology can enhance the precision of resource utilization. This not only reduces costs but also enhances the industry’s ability to withstand external shocks and improves production automation and intelligence levels. On the other hand, from the perspective of dynamic capability theory, technological innovation enhances the adaptability and flexibility of the forestry industry. Dynamic capability emphasizes that industries must possess the ability to adapt flexibly and innovate to maintain competitiveness [38]. In the forestry sector, intelligent and precision management systems can monitor and adjust production methods and management strategies in real time to address evolving environmental demands. For example, big data and artificial intelligence technologies are capable of monitoring and forecasting climate conditions as well as the status of forest resources, enabling forestry to issue early warnings, adjust logging and planting plans, and mitigate losses from natural disasters.

Based on the above analysis, this study proposes hypothesis 5: The aging of the rural labor force will drive technological innovation, thus reducing its restrictive impact on the resilience of the forestry economy (Figure 1).

4. Research Design

4.1. Research Methods

To overcome the limitations of traditional econometric models in handling high-dimensional interaction effects and nonlinear relationships, this study employs a dual machine learning model to estimate the causal effects of rural labor force aging on forestry economic resilience. This method leverages machine learning algorithms to disentangle the complex associations between treatment variables and confounding factors, effectively mitigating endogeneity issues [39]. Given that the impact of rural labor force aging on forestry economic resilience is often nonlinear, and to enhance the reliability of the results, this study incorporates as many control variables as possible, which may trigger the “curse of dimensionality” issue. However, the dual machine learning model can effectively capture nonlinear relationships among variables and has advantages in addressing the “curse of dimensionality” issue, thereby yielding more reliable and accurate results [39]. Therefore, the dual machine learning model can better test the impact of rural labor force aging on forestry economic resilience. The specific implementation steps are as follows:

4.1.1. Core Algorithm and Model Selection

This study selected the random forest algorithm to predict and solve the main regression and auxiliary regression. The reason is that as an ensemble learning algorithm, the random forest automatically captures nonlinear relationships between independent and dependent variables, higher-order interaction effects, and complex associations among high-dimensional covariates through the combination of multiple decision trees [40]. Within a dual machine learning framework, the primary and auxiliary regressions must accurately fit conditional expectations. However, in reality, the relationship between rural labor force aging and forestry economic resilience may be influenced by nonlinear effects from multiple confounding factors. The flexibility of the random forest algorithm outperforms traditional linear models, thereby reducing model specification errors.

4.1.2. Model Implementation

This study used Stata 18 software to estimate treatment effects. During model estimation, fixed random seeds (seed = 42) were used to ensure reproducibility of results, and cross-validation folds were set to 5 to balance estimation bias and variance, thereby improving the stability and reliability of model estimation.

4.2. Model Construction

4.2.1. The Baseline Testing Model

The dual machine learning model for partial linear regression is set as follows:

$$G_{it}={\theta }_0{D}_{it}+Z(M_{it})+U_{it}$$

(1)

$$E(U_{it}|D_{it},X_{it})=0$$

(2)

Among these, $G_{it}$ represents the forestry economic resilience index of province $i$ in year $t$; $D_{it}$ denotes the degree of rural labor force aging in province $i$ in year $t$; ${\theta }_0$ is the key focus coefficient. $M_{it}$ represents the collection of control variables, which may include confounding variables that affect both $G_{it}$ and $D_{it}$ simultaneously. The specific functional form $Z(M_{it})$ is unknown, so machine learning algorithms must be employed to estimate $\stackrel{\wedge }{Z}(M_{it})$. $U_{it}$ is the error term whose conditional mean is 0. If machine learning algorithms are used to directly solve for $Z(M_{it})$ in the above model, the resulting estimated coefficient ${\stackrel{\wedge }{\theta }}_0$ is a regularized estimator, which is biased in finite samples. Thus, the auxiliary regression below is further established:

$$D_{it}=n(M_{it})+V_{it}$$

(3)

$$E(V_{it}|M_{it})=0$$

(4)

where the functional form $n(M_{it})$ is unknown, and $V_{it}$ is the error term with a conditional mean of 0. Therefore, this study uses machine learning algorithms to estimate $\stackrel{\wedge }{n}(M_{it})$, constructs residual estimates of ${\stackrel{\wedge }{V}}_{it}=D_{it}-\stackrel{\wedge }{n}(M_{it})$, and then uses the same algorithm to estimate $\stackrel{\wedge }{Z}(M_{it})$ in the main regression, obtaining $Y_{it}-\stackrel{\wedge }{Z}(M_{it})={\theta }_0{D}_{it}+U_{it}$. And treat ${\stackrel{\wedge }{V}}_{it}$ as an “instrumental variable” for $D_{it}$ in the regression, yielding the following coefficient estimates:

$${\stackrel{\vee }{\theta }}_0={({\displaystyle \frac{1}{n}}{\displaystyle \sum _{i\in I,t\in T}{\stackrel{\wedge }{V}}_{it}{D}_{it}})}^{-1}{\displaystyle \frac{1}{n}}{\displaystyle {\sum }_{i\in I,t\in T}{\stackrel{\wedge }{V}}_{it}}(Y_{it}-\stackrel{\wedge }{Z}(M_{it}))$$

(5)

At this point, the convergence speed of ${\stackrel{\vee }{\theta }}_0$ depends on the convergence speeds of $\stackrel{\wedge }{Z}(M_{it})$ and $\stackrel{\wedge }{n}(M_{it})$ toward $Z(M_{it})$ and $n(M_{it})$. The two machine learning estimates are beneficial in two ways: first, they help eliminate the influence of confusion variable set $D_{it}$ in disposal variable $X_{it}$; second, they accelerate the convergence speed of ${\theta }_0$, thereby obtaining accurate estimates with a limited number of samples. In addition, to improve the stability and reliability of the model estimates, this study uses a 5-fold cross-validation method to process the regression samples.

4.2.2. The Mechanism Testing Model

This study draws on Jiang Ting’s approach [41] to construct a two-step regression model to verify the impact of rural labor force aging on forestry economic resilience. And it further explains the mediating roles of forestry scale operations, labor supply quality, forestry capital investment, and forestry technological innovation in the relationship between rural labor force aging and forestry economic resilience. The model is as follows:

$$F_{it}={\theta }_0{D}_{it}+Z(M_{it})+U_{it}$$

(6)

$$H_{it}={\theta }_0{D}_{it}+Z(M_{it})+U_{it}$$

(7)

4.2.3. The Endogeneity Testing Model

This study further develops a double machine learning partial linear instrumental variable model for analysis [42], with $K_{it}$ as the $D_{it}$ instrumental variable.

$$I{M}_{it}={\theta }_0{D}_{it}+Z(M_{it})+U_{it}$$

(8)

$$K_{it}=n{M}_{it}+V_{it}$$

(9)

4.3. Variable Definitions

4.3.1. Dependent Variable: Forestry Economic Resilience

The term “resilience” originates from Latin and was initially used to describe the ability of individuals or systems to recover after experiencing shocks [43]. In its academic evolution, it has undergone three stages: engineering resilience, ecological resilience, and evolutionary resilience, focusing, respectively, on “restoring the original state,” “maintaining structural and functional stability,” and “exploring new development paths [44].” In line with the requirements of China’s high-quality economic development phase for “sustainability and innovation-driven growth,” the concept of resilience must transcend a short-term recovery perspective and also consider the optimization of development pathways under long-term disturbances [45]. Forestry, centered on the rational management and scientific administration of forest resources, requires economic resilience grounded in the dual nature of forestry: “short-term resilience” and “long-term evolution.” Therefore, attention must be paid to the system’s “interference resistance” in responding to short-term shocks, establishing an “impact-response” mechanism to ensure the “stable quantity and quality” of forest resources and maintain the foundation of the forestry economy. Meanwhile, ecological policies such as natural forest conservation are driving the transformation of the forestry industry structure. It is necessary to cultivate green industries such as forest tourism, strengthen technological empowerment, and promote the transition of forestry from “passive recovery” to “active evolution.” By leveraging long-term policy and market disturbances, new development paths can be explored to achieve the synergistic upgrading of economic, ecological, and social values. Therefore, the construction of a forestry economic resilience indicator system should consider incorporating multi-dimensional characteristics, including engineering resilience (short-term recovery), ecological resilience (structural stability), and evolutionary resilience (path innovation).

At present, the academic community lacks a standardized measurement system for the resilience of the forestry economy. This study systematically constructs a forestry economic resilience indicator system based on relevant definitions of forestry economic resilience, referencing existing indicators such as agricultural economic resilience and urban economic resilience. The indicator system covers three aspects: risk resistance capacity, adaptive adjustment capacity, and transformation and upgrading capacity. The entropy method is used for measurement. As shown in

Table 1. Evaluation index system of forestry economic resilience.

Primary Indicators	Secondary Indicators	Tertiary Indicators	Unit	Indicator Property
Risk resistance capacity	Production resilience	Forest land area	Ten thousand hectares	+
		Number of personnel in the forestry system at year-end	People	+
		Total forestry output value	Ten thousand yuan	+
	Development resilience	Forestry industrial structure (proportion of secondary and tertiary forestry industries’ output value to total output value)	%	+
		The production situation of major economic forest products in various regions	Ton	+
		The completed amount of forestry investment	Ten thousand yuan	+
Adaptation and adjustment capacity	Sustainable resilience	The number of forest parks	Individual	+
		Wetland resources situation	Hectare	+
		Artificial afforestation area in that year	Hectare	+
	Restorative resilience	Forest stock volume	Ten thousand cubic meters	+
		Forest coverage rate	%	+
		Forest disease Prevention and control rate	%	+
Transformation and upgrading capacity	Innovation resilience	The number of units providing forestry science and technology exchange and promotion services	Individual	+
		Investment in forestry science and technology education	Ten thousand yuan	+
		The quality of personnel in forestry workstations (number of people with college degrees or above/total number of long-term employees of the forestry station)	%	+
	Transformational resilience	The output value of forest undergrowth economy	Ten thousand yuan	+
		Output value of forestry tourism and leisure services	Ten thousand yuan	+
		Investment in ecological construction and protection was completed this year	Ten thousand yuan	+

, risk resilience encompasses three aspects: production resilience, development resilience, and sustainability resilience. Production resilience reflects the internal production capacity and output of the forestry sector, measured by three indicators: forest land area, the year-end number of personnel in the forestry system, and total forestry output value. Forest land area directly reflects the material foundation of forestry production and serves as the core carrier for resisting natural disasters. The year-end number of personnel in the forestry system indicates the stability of production factors. The total forestry output value directly measures the output effectiveness of production functions. Together, these three elements form a “resource–human resources–output” production resilience loop. Development resilience reflects the forestry sector’s ability to sustainably provide ecological services and promote socio-economic development, measured by three indicators: forestry industrial structure, production status of major economic forest products in various regions, and completed forestry investment. The structure of the forestry industry reflects the degree of diversification of system functions. the more balanced the structure, the stronger the ability to resist fluctuations in a single industry. The production status of major economic forest products in various regions reflects the stability of regional comparative advantages. The amount of forestry investment completed ensures the sustained improvement of production capacity. Together, these three factors resonate with the resilience requirements of “synergy between ecological services and economic development.” Sustainable resilience reflects the forestry sector’s ability to maintain stability and sustainable development in the face of external pressures and changes, measured by three indicators: number of forest parks, wetland resource status, and annual artificial afforestation area. The number of forest parks and wetland resources are core indicators of ecosystem stability, and their scale directly influences the forestry sector’s ability to withstand ecological shocks. The area of artificial afforestation in the current year reflects the sustainable logic of “using while restoring.” Recovery resilience denotes the capacity of forestry to bounce back after suffering impacts in the course of its development, measured by three indicators: forest stock volume, forest coverage rate, and forest pest control rate. Forest stock volume and forest coverage rate are key indicators for measuring the extent of forest resource recovery, with their growth directly reflecting the system’s ability to recover from disasters. The forest pest control rates reflect the capacity to proactively address biological shocks. Together, these three factors form the “resource stock–coverage–control efficiency” recovery chain. Innovation resilience refers to the diversity of collaboration and advancements in technology within forestry development, measured by three indicators: the number of forestry science and technology exchange and promotion service units, forestry science and technology education investment, and the quality of personnel at forestry work stations. The number of forestry science and technology exchange and promotion service units serves as a vehicle for technology diffusion, forestry science and technology education investment provides financial support for innovation. And the quality of personnel at forestry work stations reflects the innovative implementation capacity at the grassroots level. Together, these three elements form the “technology–funding–human resources” innovation support system. Transformative resilience refers to the ability to undergo transformation, change, and adjustment after facing shocks in forestry development, measured by three indicators: the economic output value of under-forest industries, the output value of forestry tourism and leisure services, and the investment completed in ecological construction and protection for the current year. The economic output value of under-forest industries and forestry tourism and leisure services reflects the transformation of the industrial structure from “traditional cultivation and breeding” to “diversified integration”. The investment completed in ecological construction and protection this year ensures the ecological bottom line during the transformation process. Together, these three elements support the evolutionary path of “economic–ecological” synergistic upgrading.

The entropy method can reduce human bias, fully utilize data characteristics, and make the evaluation results more in line with actual conditions [46]. Therefore, this study uses the entropy weight method to construct forestry economic resilience to avoid subjective bias in weight assignment. The specific formula is as follows:

(1)

For positive and negative indicators, respectively, use:

$${X^{\prime }}_{ij}={\displaystyle \frac{(X_{ij}-\min X_j)}{(Max{X}_j-\min X_j)}}$$

(10)

$${X^{\prime }}_{ij}={\displaystyle \frac{(max{X}_j-X_{ij})}{(max{X}_j-\min X_j)}}$$

(11)

(2)

Calculate the proportion of the $j$th indicator value in the $i$th year:

$$Y_{ij}={\displaystyle \frac{{X^{\prime }}_{ij}}{{\displaystyle \sum _{i=1}^m{X^{\prime }}_{ij}}}},i=1,2,\dots ,m; j=1,2,\dots n$$

(12)

(3)

Determine the information entropy of the $j$ th indicator:

$$e_j=-k{\displaystyle \sum _{i=1}^m(Y_{ij}\times \ln Y_{ij})}, j=1,2,\dots ,n$$

(13)

Let $K={\displaystyle \frac{1}{\ln m}}$, then $0\le e_j\le 1$, and when $Y_{ij}=0$, let $Y_{ij}\times \ln Y_{ij}=0$

(4)

Calculation of information entropy redundancy: $d_j=1-e_j, j=1,2,\dots ,n$

(5)

Determination of indicator weights:

$$W_j={\displaystyle \frac{d_j}{{\displaystyle \sum _{j=1}^n{d}_j}}}, j=1,2,\dots ,n$$

(14)

In the above formula, ${X^{\prime }}_{ij}$ and $X_{ij}$ represent the standardized results and original values of the $j$th indicator in the $i$th year, respectively; $\max X_j$ and $\min X_j$ represent the maximum and minimum values of the $j$th indicator across all years, respectively; denotes the total number of years measured; $m$ denotes the number of indicators used in the measurement; $n$ denotes the proportion of the $j$th indicator in the $i$th year relative to the total sum of the $j$th indicator over the years; $d_{ij}$denotes the information entropy redundancy (variability coefficient) of the jth indicator;$W_j$ denotes the weight value of the $j$th indicator.

Using the above method to determine the indicator weights, calculate the weights for each individual indicator. Based on the constructed comprehensive evaluation indicator system for forestry economic resilience, use the weighted function to obtain the forestry economic resilience index for each province. The calculation formula is as follows:

$$S_j={\displaystyle \sum _{j=1}^m({X^{\prime }}_{ij}·W_j)}$$

(15)

where ${X^{\prime }}_{ij}$ is the standardized value of the $j$th indicator in the $i$th year, $n=19$ is the number of selected indicators; $W_j$ represents the weight of the $j$th indicator; $S$ is the comprehensive evaluation value of the forestry economic resilience index.

4.3.2. Core Explanatory Variable: Rural Labor Force Aging

Due to the long working hours of Chinese farmers and the improvement of rural living standards and medical conditions, the life expectancy of the rural population has continued to increase, leading to a growing number of elderly people. At the same time, a large number of young and middle-aged rural laborers have migrated to urban areas for work, leaving the elderly labor force behind in rural areas to engage in agricultural production. Thus, this study adopts the method proposed by Yao Dongmin et al. [47] employing the ratio of rural residents aged 65 and over to the total rural population to gauge the extent of aging in the rural labor force.

4.3.3. Mediating Variable

Scaled-up forest land management. Land transfer is a prerequisite and foundation for scaled-up management. Through land transfers, scattered forest land can be consolidated into a certain scale of forest land resources, thus laying the groundwork for large-scale operations. The degree of large-scale forest land management is measured using the land transfer rate of each region in the current year, with the logarithm taken.

Labor supply quality. The quality of forestry labor supply is measured from two aspects: educational human capital and health human capital. Among these, educational human capital is measured by the proportion of rural residents aged 6 and over who have a high school education or higher. Health human capital is assessed by rural residents’ healthcare expenditure. Referring to Jiang Jian’s consideration that rural residents’ healthcare expenditure has a negative correlation with health human capital, the reciprocal of this value is taken [48].

Forestry capital investment. Following the approach of Shan Haojie (2008), the overall quantity of forestry fixed asset investment is used as the basis, and the perpetual inventory method is used to compute forestry capital investment [49]. Additionally, since China’s statistical data do not publish the fixed asset investment price index by province and industry, this study references Guo Pengfei’s use of the overall fixed asset investment price index for each province as a substitute in the specific calculation process [50].

Forestry technological innovation. The level of forestry technological innovation is gauged by the number of forestry patent applications disclosed annually per province, with its logarithm calculated.

4.3.4. Control Variables

To comprehensively analyze the impact of factors such as economic development and urbanization levels on the resilience of the forestry economy in the context of rural labor force aging, the following factors are selected as control variables based on relevant research. (1) Per capita regional GDP. In view of regional population disparities, local per capita GDP is adopted to assess economic development levels, while factoring in the impact of economic development on the development levels of forestry [51]. (2) Industrialization level. Since local economic development promotes increases in forestry industry employment and forestry economic growth [18], the ratio of industrial added value to GDP is employed to gauge the industrial structure. (3) Level of urbanization. This is measured by the percentage of the urban population in the total regional population (encompassing both agricultural and non-agricultural sectors) and serves to account for the influence of shifts in rural residents’ production and business activities on forestry development [52]. (4) Precipitation intensity. Changes in rainfall affect the severity of drought duration in a region, which in turn affects forest quality [53] and further impacts local forestry economic development. Considering regional area differences, the ratio of precipitation to provincial area was used as a measure [1]. Additionally, to mitigate data volatility and address heteroskedasticity issues, the two variables of per capita GDP and precipitation intensity are log transformed in the empirical analysis below. Partially missing data are filled using interpolation methods.

4.4. Data Sources

Due to extensive missing data for some variables, this study chose panel data from 30 provinces (excluding Tibet, Hong Kong, Macao, and Taiwan) spanning 2012 to 2022 as the sample. Forestry economic resilience sub-indicator data are sourced from the “China Forestry Statistical Yearbook” and the “China Forestry and Grassland Statistical Yearbook”. Data on rural labor force aging and economic indicators are sourced from the China Population and Employment Statistical Yearbook, the China Statistical Yearbook, and provincial statistical yearbooks (2012–2022). Forestry technology innovation data (patents) are sourced from the China Science and Technology Statistical Yearbook. Land transfer rate data are sourced from the China Rural Policy and Reform Statistical Annual Report. Missing data points were supplemented using linear interpolation to ensure the integrity of the dataset. The descriptive statistical results for each variable are presented in

Table 2. Descriptive statistical results of each variable.

Variable Type	Variable	Measurement Method	Mean	Std. Dev.	Min	Max
Dependent variable	Forestry economic resilience	Entropy weight method measure	0.211	0.0889	0.0309	0.456
Independent variable	Aging rural labor force	The population aged 65 and above in rural areas/the total population of rural areas	0.136	0.0452	0.0537	0.275
Mechanism variable	Large-scale operation of forestry	The land transfer rate is taken as the logarithm	3.387	0.555	1.309	4.512
	Educational human capital	The number of rural residents aged 6 and above with a high school education or above	0.103	0.0317	0.0456	0.239
	Health human capital	Expenditure on medical care and health for rural residents shall be taken as the reciprocal	0.107	0.0550	0.0372	0.354
	Forestry capital investment	Based on the total amount of fixed asset investment in forestry, the amount of capital input in forestry is calculated by using the perpetual inventory method	12.82	1.702	6.841	17.60
	Forestry technological innovation	Take the logarithm of the number of forestry patent applications disclosed in that year	7.096	1.137	3.497	9.587
Control variable	Per capita gross regional product	Per capita GDP	10.91	0.445	9.849	12.16
	Degree of industrialization	Industrial added value/GDP	0.321	0.0811	0.101	0.523
	Urbanization level	Urban population/total regional population	0.607	0.117	0.363	0.896
	Precipitation intensity	Precipitation/area of provinces and cities	3.975	1.531	−0.0994	8.726

.

5. Results and Analysis

As shown in Figure 2, this nuclear density plot depicts the aging of China’s rural labor force and the resilience of its forestry economy from 2012 to 2022. In 2012, the peak level of aging in China’s rural labor force was around 11%. Over time, the peak gradually shifted to the right, indicating that the aging of China’s rural labor force is becoming increasingly severe, and the gap in labor force aging between regions is gradually widening. The kernel density curve of China’s forestry economic resilience exhibits a distinct multi-peak pattern, indicating a phenomenon of multi-polar differentiation with significant regional disparities. The distribution center of gravity is overall shifted to the left, suggesting that the forestry economic resilience of various regions is relatively low. The vertical height of the peaks has been declining since reaching its peak in 2014. And the number of peaks has decreased, indicating that the kernel density is moving toward smaller values, with regional gaps narrowing and exhibiting dynamic convergence characteristics. The deepening aging of the rural labor force and the widening regional gaps are closely linked to the overall low resilience of the forestry economy and the dynamic convergence of regional differentiation. On the one hand, regions with higher aging rates face more pronounced labor constraints, resulting in more evident resilience shortcomings and exacerbating regional imbalances. On the other hand, the widespread deepening of aging may force some regions to alleviate labor shortages by introducing mechanized production, optimizing forestry management models, or attracting external labor through policy support. Such adaptive adjustments may to some extent resonate with the dynamic convergence of forestry economic resilience. This provides an intuitive real-world basis for exploring the influence of the former on the latter, and the underlying mechanisms between the two require further investigation.

5.1. Benchmark Regression Analysis

This study utilizes a dual machine learning model to investigate the effect of aging in the rural labor force on the resilience of the forestry economy. The regression findings presented in

Table 3. Baseline regression results.

Variable	(1)	(2)
	Forestry Economic Resilience	Forestry Economic Resilience
Aging rural labor force	−0.283 ***	−0.282 ***
	(−3.63)	(−3.71)
Control variable	Yes	Yes
Control the quadratic term of the variable	No	Yes
Year	Yes	Yes
Province	Yes	Yes
n	330	330

Note: *** indicate significance at the 1% level. The values in parentheses are robust standard errors.

. In Model (1), the regression coefficient for the rural labor force aging variable is significantly negative. From an economic perspective, this means that after controlling for province and time fixed effects, a 1% increase in the degree of rural labor force aging will reduce the resilience of the forestry economy by an average of 0.283 units. Building on Model (1), Model (2) further incorporates the quadratic terms of the control variables. The regression coefficient remains significantly negative, with negligible variation in magnitude. Hypothesis 1 is validated.

5.2. Robustness Tests

To further validate the reliability of the above research results, this study employs the following five methods for robustness tests. First, the sample splitting ratio of the dual machine learning mode is modified from the previous 1:4 to 1:2 and 1:7 to explore the potential impact of the sample splitting ratio on the research conclusions. Second, the sample size was reduced, and all variables in the baseline regression were trimmed at the 1% and 99% percentiles, as well as the 5% and 95% percentiles. Third, the study period was shortened to 2014–2022. Fourth, we replaced the machine learning algorithm, switching from the previously used the random forest algorithm for prediction to gradient boosting, to explore the potential impact of the prediction algorithm on the study conclusions. Fifth, we used a dual fixed-effects model to test the effectiveness of the dual machine learning model. The regression results indicate (

Table 4. Robustness test results (1).

Variable	(1)				(2)				(3)
	Change the Sample Segmentation Ratio				Reduce the Sample Size				Adjust the Research Sample
	1:2		1:7		Reduce the Tail by 1%		Reduce the Tail by 5%		Shorten the Sample Period to 2014–2022
Aging rural labor force	−0.324 ***	−0.319 ***	−0.304 ***	−0.264 *	−0.280 ***	−0.280 ***	−0.233 **	−0.233 **	−0.244 **	−0.246 **
	(−4.23)	(−4.16)	(−3.9)	(−2.75)	(−3.77)	(−3.62)	(−3.00)	(−2.97)	(−2.63)	(−2.6)
Control variable	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Control the quadratic term of the variable	No	Yes	No	Yes	No	Yes	No	Yes	No	Yes
Year	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Province	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes
n	330	330	330	330	330	330	330	330	270	270

Note: *, **, and *** indicate significance at the 10%, 5%, and 1% levels, respectively. The values in parentheses are robust standard errors.

and

Table 5. Robustness test results (2).

Variable	(1)		(2)
	Change the Algorithm Gradient Boosting		Switch to a Dual Fixed-Effects Model
Aging rural labor force	−0.065 **	−0.246 **	−0.664 ***	−0.633 ***
	(−2.46)	(−2.66)	(−5.05)	(−4.73)
Control variable	Yes	Yes	Yes	Yes
Control the quadratic term of the variable	No	Yes	No	Yes
Year	Yes	Yes	Yes	Yes
Province	Yes	Yes	Yes	Yes
n	330	330	330	330

Note: ** and *** indicate significance at the 5%, and 1% levels, respectively. The values in parentheses are robust standard errors.

) that rural labor force aging continues to significantly negatively suppress forestry economic resilience, and the benchmark regression results are robust. It is also important to note that the absolute value of the coefficient in the dual machine learning model (0.283) is smaller than that in the bidirectional fixed-effects model (0.664). This indicates that the traditional model overestimates the negative impact due to its failure to account for nonlinear relationships, thereby validating the corrective role of the machine learning model.

5.3. Endogeneity Testing

Endogeneity typically refers to a relationship where explanatory variables and response variables mutually influence each other, which may lead to biased estimation results. Omitted variable error, measurement error, reverse causality, self-selection bias, and sample selection bias are the primary causes of endogeneity issues. If not properly addressed, these issues may render empirical test results unreliable and lead to erroneous research conclusions. Therefore, it is necessary to discuss each potential endogeneity issue in this study. (1) Omitted variable error issue. In constructing the panel model, this study controlled for fixed effects including time and province, addressing the endogeneity issues associated with omitted variables that vary by province or over time. (2) Measurement error issues. The estimation of rural labor force aging as an explanatory variable is primarily based on sample surveys conducted in rural areas. The sample results are relatively independent and reliable, with a low likelihood of data contamination or recording errors, thereby accurately reflecting the level of rural labor force aging in the region. (3) Self-selection bias issue. The extent of changes in rural labor force aging is only slightly affected by regional policy adjustments, and regions have little autonomy to exert an impact on such changes. (4) Sample selection bias issue. This study utilizes panel data from 30 provinces in China, ensuring the comprehensiveness and validity of the data and research subjects. (5) Reverse causality issue. It is generally believed that rural labor force aging is a long-term process, and changes in forestry economic resilience may also be influenced by past labor force conditions. Therefore, the lagged rural labor force aging data are used as an explanatory variable to reflect the impact of past labor force conditions on the current resilience of the forestry economy. Additionally, using the lagged rural labor force aging variable can to some extent weaken the mutual influence between the explanatory variable and the dependent variable. Thereby, reducing the impact of endogeneity issues on the estimation results. It should be noted that the use of instrumental variables has already been addressed in the robustness tests. Building upon these robustness tests, this study takes a further step. It employs the impact of one-period-lagged rural labor force aging on forestry economic resilience as an endogeneity test for the influence of rural labor force aging on forestry economic resilience. As indicated in

Table 6. Results of the endogeneity test.

Variable	(1)	(2)
	Forestry Economic Resilience	Forestry Economic Resilience
The aging of the rural labor force lags behind the first phase	−0.230 **	−0.216 ***
	(−2.68)	(−2.47)
Control variable	Yes	Yes
Control the quadratic term of the variable	No	Yes
Year	Yes	Yes
Province	Yes	Yes
n	330	330

Note: **, and *** indicate significance at the 5%, and 1% levels, respectively. The values in parentheses are robust standard errors.

, the lagged rural labor force aging variable continues to have a negative effect on forestry economic resilience a significance level of at least 5%, which aligns with the outcomes of the baseline regression.

5.4. Heterogeneity Analysis

To further examine whether the impact of rural labor force aging on forestry economic resilience is consistent across different regions, China’s 30 provinces were divided into categories. These categories include eastern, central, and western regions (based on geographical location) as well as grain-producing regions, grain-producing and -consuming balanced regions, and grain-consuming regions (based on agricultural functional location). Corresponding empirical tests were then conducted. The findings of the tests are presented in

Table 7. Results of the geographical location heterogeneity test.

Variable	(1)		(2)		(3)
	Eastern		Central		Western
Aging rural labor force	−0.426 **	−0.426 **	−0.875 ***	−0.887 ***	−0.147	−0.137
	(−3.47)	(−3.37)	(−3.98)	(−3.99)	(−0.91)	(−0.81)
Control variable	Yes	Yes	Yes	Yes	Yes	Yes
Control the quadratic term of the variable	No	Yes	No	Yes	No	Yes
Year	Yes	Yes	Yes	Yes	Yes	Yes
Province	Yes	Yes	Yes	Yes	Yes	Yes
n	121	121	88	88	121	121

Note: **, and *** indicate significance at the 5%, and 1% levels, respectively. The values in parentheses are robust standard errors.

and

Table 8. Results of the functional localization heterogeneity test.

Variable	(1)		(2)		(3)
	Major Grain-Producing Areas		Grain Production and Marketing Balance Zone		Major Grain-Consuming Region
Aging rural labor force	−0.702 ***	−0.736 ***	−0.214	−0.233	−0.295 **	−0.305 **
	(−4.11)	(−4.14)	(−1.40)	(−1.59)	(−2.87)	(−2.83)
Control variable	Yes	Yes	Yes	Yes	Yes	Yes
Control the quadratic term of the variable	No	Yes	No	Yes	No	Yes
Year	Yes	Yes	Yes	Yes	Yes	Yes
Province	Yes	Yes	Yes	Yes	Yes	Yes
n	143	143	110	110	77	77

Note: **, and *** indicate significance at the 5%, and 1% levels, respectively. The values in parentheses are robust standard errors.

. The impact of rural labor force aging on forestry economic resilience is significantly negative in the eastern and central regions, as well as in grain-producing regions and grain-consuming regions. However, it is not significantly negative in the western region and grain-producing and -consuming balanced regions. The magnitude of the negative impact is greater in the central region than in the eastern region, and greater in grain-producing regions than in grain-consuming regions.

The reasons for this result may lie in the following: although the eastern region is economically more developed, its forestry industry is relatively small in scale, especially under limited land resources, making forestry production more reliant on labor. When young laborers migrate, the remaining elderly laborers struggle to effectively undertake labor-intensive tasks such as tree planting and logging. Additionally, an increasing number of laborers are flowing into other industries such as manufacturing and services, while new laborers cannot be promptly replenished into forestry production. This severely constrains the sustainable development of the forestry economy. In central regions, the forestry industry primarily relies on labor-intensive production such as forest harvesting and nursery cultivation, and the aging of the labor force directly leads to a decline in production efficiency. Furthermore, in comparison to the eastern region, the central region exhibits a lower level of agricultural modernization, characterized by inadequate investment in technology and capital. Therefore, the impact of labor outflow and aging on productivity cannot be effectively mitigated through modernization measures. The traditional labor-intensive forestry production model struggles to adapt to labor shortages, further reducing the resilience of the forestry economy. Although the western region possesses relatively abundant forestry resources, its economy and technology are relatively backward, transportation is inconvenient, and the forestry industry is small in scale and underdeveloped, primarily engaged in primary production with low reliance on labor. The weak industrial foundation means that the impact of an aging workforce on the forestry economy has not been significantly amplified.

In major grain-producing and grain-consuming regions, the forestry industry is not a primary economic pillar. These regions primarily rely on grain production, with forestry occupying a subordinate position in the local economy. Under such circumstances, the forestry industry struggles to secure adequate resource allocation, resulting in a lack of modern technological tools and financial support. This further leads to low production efficiency, inadequate management standards, and weak resilience against market fluctuations and natural disasters. Additionally, although agriculture and forestry are interdependent in grain-producing and grain-consuming regions, due to the priority given to grain production and the relatively low status of the forestry industry. When grain market conditions are favorable, the aging labor force may be more inclined to maintain traditional grain production models and overlook market opportunities in the forestry industry. When the forestry product market experiences fluctuations, they may struggle to adjust quickly, potentially causing losses in the forestry economy and reducing its resilience. In regions where grain production and consumption are balanced, the two are relatively in equilibrium, and the forestry economy may make up a moderate share of the overall economy. The impact of rural labor force aging on forestry economic resilience may vary depending on the scale and importance of the forestry industry within the region. If the forestry industry is small in scale or of low importance, the impact of aging may not be significant.

5.5. Mechanism Testing

Table 9. Results of the mechanism test.

Variable	(1)		(2)		(3)		(4)		(5)
	Large-Scale Forestry Operations		Educational Human Capital		Healthcare Human Capital		Forestry Capital Investment		Forestry Technological Innovation
Aging rural labor force	1.336 **	1.361 **	−0.148 ***	−0.151 ***	−0.212 ***	−0.209 ***	8.289 ***	8.303 ***	7.519 ***	7.611 ***
	(2.43)	(2.51)	(−3.86)	(−3.98)	(−5.49)	(−5.43)	(4.08)	(4.14)	(4.68)	(4.73)
Control variable	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Control the quadratic term of the variable	No	Yes	No	Yes	No	Yes	No	Yes	No	Yes
Year	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Province	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes
n	330	330	330	330	330	330	330	330	330	330

Note: **, and *** indicate significance at the 5%, and 1% levels, respectively. The values in parentheses are robust standard errors.

column (1) shows that the coefficient of the impact of rural labor force aging on the land area of large-scale forestry operations is 1.336, which is significantly positive within the 5% confidence interval. This indicates that as the rural labor force ages, some farmers transfer their forest land due to labor shortages, thereby promoting the concentration of forest land in the hands of large-scale operators. Large-scale forestry operations help improve resource allocation efficiency and management levels, enhance the ability to withstand market fluctuations and natural risks, and thereby increase the stability and resilience of the forestry economy. Hypothesis 2 is validated. Table 8, Column (2) shows that the impact of rural labor force aging on educational human capital is significantly negative at the 1% confidence level, with a coefficient of −0.148. This indicates that the outflow of young laborers inhibits the accumulation and development of educational human capital, constraining the intrinsic driving force for the sustainable development of the forestry economy and thereby suppressing its resilience. Column (3) of Table 8 shows that the impact of rural labor force aging on medical human capital is significantly negative at the 1% confidence level, with a coefficient of −0.212. This indicates that the aging of the labor force structure exacerbates the pressure on rural medical resources. However, due to the limited ability of elderly laborers to access and afford high-quality medical services, the level of medical human capital declines. This adversely affects the health protection capabilities and production efficiency of forestry workers. Thus, it can be seen that the aging of the rural labor force reduces the quality of labor supply from both educational human capital and medical human capital perspectives, thereby inhibiting forestry economic resilience, and hypothesis 3 is validated. Column (4) of Table 8 shows that the impact of rural labor force aging on government forestry capital investment is significantly positive within the 1% confidence interval. This indicates that as rural labor force aging continues to intensify, the shortage of forestry labor supply becomes increasingly evident. In turn, compels the government to increase its support for forestry to address the development bottlenecks caused by labor shortages. Government capital investment helps to enhance the modernization level and resource utilization efficiency of forestry, thereby strengthening the resilience and sustainable development capacity of the forestry economy. Hypothesis 4 is validated. Column (5) of Table 8 shows that the impact of rural labor force aging on forestry technological innovation is significantly positive at the 1% confidence level. This indicates that the aging of the rural labor force makes traditional labor-intensive forestry production methods unsustainable. To address labor shortages, forestry enterprises and farmers are more inclined to adopt intelligent equipment and efficient production technologies to improve labor efficiency and resource utilization rates. This shift has to some extent mitigated the negative impact of aging on the resilience of the forestry economy. Hypothesis 5 is validated.

6. Conclusions and Policy Recommendations

6.1. Research Conclusions and Discussion

Forestry development is inherently susceptible to natural disasters and other risks, and the increasingly severe phenomenon of population aging poses a significant challenge to the resilience of the forestry economy. Therefore, this study examines the impact of rural labor force aging on the resilience of the forestry economy and its underlying mechanisms from the perspective of production factors. Using panel data from 30 provinces in China from 2012 to 2022, a dual machine learning model is employed to analyze the effects and mechanisms at play. The findings of this study are consistent with the existing literature while expanding upon it, offering unique insights into the distinctive dynamics of the forestry sector and providing practical implications for a global context.

First, the core finding of this study, that “the aging of the rural labor force significantly inhibits the resilience of the forestry economy,” aligns with the academic consensus on the challenges posed by population aging to labor-intensive industries. Existing research indicates that labor force aging reduces labor supply, lowers production efficiency, and constrains industrial development [9,10,12]. This study validates the applicability of this pattern in the forestry sector, confirming the common vulnerabilities faced by labor-intensive primary industries in the face of demographic changes. In terms of methodological choices, this study employs a dual machine learning model to overcome the limitations of traditional fixed-effects or OLS methods [9,10], which struggle to handle the nonlinear relationships and high-dimensional interactions inherent in resilience research. By validating core results through robustness tests, we demonstrate the practicality of machine learning in enhancing causal inference in forestry economics, aligning with Athey et al.’s [39] advocacy for such methods in complex empirical settings.

Secondly, we further revealed its “double-edged sword” effect: although aging has driven positive adaptive changes, such as the expansion of forestry operations and technological innovation—which is consistent with the view in existing research that “aging may force efficiency improvements” [16]. For example, Lee et al. [54] found that developed economies such as Japan and Germany are utilizing robotics to maintain productivity amid aging populations and mitigate the negative impact of aging on total factor productivity growth, particularly in labor-intensive industries. However, these adaptive mechanisms are insufficient to offset the overall negative impact, particularly in terms of reduced human capital. This nuanced distinction deepens our understanding of the heterogeneity of population aging across industries: in the forestry sector, due to its ecological externalities and long production cycles, short-term labor constraints appear to outweigh long-term adaptive benefits. This distinction has been rarely emphasized in previous studies focused on labor-intensive industries like agriculture. Additionally, this study’s mechanism analysis integrates multi-dimensional pathways. Previous studies often focused on single associations, such as labor substitution, land transfers [15], and “harvesting efficiency” [17]. This study identifies interrelated channels involving forestry scale operations, human capital, government investment, and technological innovation. Our finding that population aging weakens educational and health human capital aligns with Liu et al.’s [13] observation that aging hinders technology adoption. Furthermore, “government capital investment” as an institutional response to labor shortages complements forestry capital accumulation research [50], emphasizing policy-driven adaptive mechanisms.

Finally, the heterogeneity analysis in this study indicates that the aging of the rural labor force has a significant negative impact on the resilience of the forestry economy in eastern and central regions, major grain-producing areas, and major grain-consuming areas, with the magnitude of the negative impact being greater in central regions than in eastern regions, and greater in major grain-producing areas than in major grain-consuming areas. This result may be attributed to differences in industrial structure, resource dependency, and policy support across regions. Essentially, regions with higher levels of economic development, faster industrial upgrading, and stronger resource substitution capabilities are better able to withstand the impact of labor force aging; conversely, regions reliant on traditional production and lacking sufficient resource substitution capabilities are more significantly affected. Cross-country comparisons also reveal structural disparities: for example, Burtless argues that there is little evidence of a negative link between aging and productivity in the United States [55]. However, the most significant negative impacts will occur in countries with rapidly aging labor forces, such as Spain, Italy, Portugal, Greece, and Ireland [56].

As a result, further reflection on the applicability of the findings of this study to other contexts can provide insights in multiple areas: From the perspective of a country’s stage of development, for those nations and regions similar to China’s central regions or major grain-producing areas—which are in an economic transition phase, have an industrial structure dominated by traditional labor-intensive industries, and possess limited technological substitution capabilities—it is essential to proactively plan and increase investments in forestry mechanization and intelligence when addressing the impact of rural labor force aging on the resilience of the forestry economy. Concurrently, enhancing skill training for elderly laborers is necessary to mitigate the negative effects of short-term labor constraints. For economically developed regions with rapid industrial upgrading, such as the United States, their experiences also indicate that by promoting industrial transformation toward technology-intensive sectors and improving policy support systems, they can effectively withstand the impacts of an aging workforce. From the perspective of industrial expansion, this study focuses on the forestry sector, which has ecological externalities and a long production cycle. Its conclusions also hold some reference value for other resource-dependent, long-cycle industries. These industries may also face situations where short-term labor constraints outweigh long-term adaptive benefits when confronting labor force aging. Therefore, they need to explore suitable labor replacement and industrial upgrading pathways tailored to their own industrial characteristics. At the policy-making level, policymakers should fully consider the local industrial structure, resource endowments, and technological levels when formulating differentiated policies for different regions and industries.

6.2. Policy Implications

Based on the research findings, this study offers the following policy implications.

First, establish a flexible and diversified labor supply and replacement system. The global forestry industry is widely affected by an aging workforce, and countries need to adopt tailored policies based on local conditions. Developing countries can draw on China’s targeted training and socialized service models to achieve intensive labor utilization through cooperatives and specialized companies. Developed countries can strengthen the combination of technological substitution and talent introduction to form a “technology–human resources” complementary system.

Second, explore a “scale and technology” integration path suited to national conditions. Countries should select their paths based on their resource endowments: countries with abundant resources but dispersed operations can draw on China’s experience with collective forest rights reform, focusing on improving supporting services for forest land transfers. This includes establishing professional assessment institutions, reducing transaction costs, and providing fiscal subsidies to encourage returning migrants to undergo training in large-scale operations. Countries with mature industries should focus on integrating “scaling up and digitalization,” implementing technology-adaptive policies, and building smart management platforms. For example, in forest areas with a high proportion of elderly laborers, develop user-friendly smart management tools and provide free skill training to ensure the inclusivity of technological innovations. Additionally, include common forestry-related injuries and illnesses in the scope of medical insurance reimbursement priorities and promote the use of mechanized auxiliary tools to reduce the labor intensity for elderly workers.

Third, implement differentiated regional strategies. For developing regions where the negative effects of aging are more pronounced, increase central government fiscal transfers to prioritize infrastructure construction for large-scale forest land management. Simultaneously, implement a “technology dissemination” program, deploy forestry science and technology specialists for on-site guidance, and prioritize the promotion of simplified, easy-to-operate production technologies in forest areas with a high concentration of elderly labor. High-income areas can leverage their economic advantages to pilot emerging industry models such as “forestry wellness,” utilizing ecological resources to develop wellness industries and provide flexible positions such as park management and tourist guidance for elderly laborers. Simultaneously, capital released through industrial upgrading can be reinvested into forestry technology research and development to reduce reliance on traditional labor. Major grain-producing regions should develop low-intensity industries suitable for elderly laborers, such as forest-based planting and farming, while ensuring grain production, and use grain subsidies to support forestry labor. Major grain-consuming regions should focus on “market linkage,” leveraging their consumer market advantages to establish e-commerce platforms for forestry products and stabilize forestry production expectations through contract farming.

6.3. Limitations and Prospects

This study is constrained by data availability, as the sample period does not cover the impact of extreme climate events, making it hard to precisely capture the dynamic changes in resilience during sudden risks. Future research might broaden the range of data collection, incorporate data related to extreme climate events into the analysis, and construct a dynamic assessment model. Supplementary microdata could be obtained through field surveys to strengthen empirical support.