Efficiency Evaluation and the Impact Factors of Sustainable Forestry Development in China: Adoption of Super-Efficiency Data Envelopment Analysis and Malmquist Index Methods

Abstract

Forestry is the underpinning of economic and environmental civilization for sustainable economic development. Forestry benefits ecosystems and local dwellings; thus, transforming and advancing forest products in a civilized society is critical to building a progressive community. This research aims to analyze the merits of forest products based on static, dynamic, and scale efficiency. It employed the super-efficient data envelopment analysis—Charnes, Cooper and Rhodes, and Malmquist index methods with an enhanced robustness check by applying the Stochastic Frontier Analysis. The results show that, first, from the perspective of static efficiency, the forestry industry’s operating performance is maintained at a relatively high level, with a minimal fluctuation range; however, the scale efficiency is observed at a deficient level. Limited by the scale efficiency results, the aggregate comprehensive technical efficiency was at a low standard. The average total factor productivity of dynamic efficiency for each company reached 1.029, indicating that the resource allocation of listed companies is relatively reasonable, operating efficiency is on the rise, and technical level changes are a substantial source of total factor productivity growth. Furthermore, the super efficiency DEA findings indicated that 11 of the most high-performance companies attained an efficiency value above criterion value one after re-ranking. The comprehensive efficiency value of Sun Paper reached 1.438 in 2019, Sophia 1.393, Hexing Packaging 1.383, Fujian Jinsen 1.326, Bunny 1.271, Meike Home Furnishing 1.235, Jingxing Paper 1.222, Zhongshun Jiezou 1.116, Meiying Sen 1.099, Chenming Paper 1.064, Da Ya Dekor 1.015. Second, the comprehensive efficiency value of Fujian Jinsen’s company has a more significant impact on the overall average efficiency of the forestry listed companies. After excluding Fujian Jinsen, the average comprehensive efficiency of 2018–2019 indicated an upward trend for the listed companies. This study suggests that policymakers should strengthen the forestry industry development plan and policy implementation regulations to focus on the forward-looking and guiding principles of industrial and technological innovation policies.

1. Introduction

Forestry is the foundation of ecological civilization construction for the sustainable development of economies [1,2,3,4,5]. Forestry enhances the environment and human habitation [2,6]. Thus, for economic growth and the realization of balanced forest development, the environment, and the global economy, it is imperative to comprehend the current state of the forestry industry and its accompanying resource utilization in China and accurately assess the forestry industry’s performance [7,8,9]. Forestry’s eco-friendly mechanism is the foundation of sustainable socioeconomic development [5,10]. As an essential industry, forestry significantly contributes to ecological construction, climate maintenance, and forest product supply for a sustainable environment [11,12,13]. Therefore, promoting a healthy forest is crucial for building a well-off society in an all-encompassing way [14,15]. The transformation of ecological advantages into economic advantages to achieve greening and ecological productivity has been critical for achieving economic growth [16,17]. China’s national forestry and grassland administration policies emphasize the need to ramp up nature reserves and ecological conservation sites to develop tourism resources, such as forestry, to promote forestry experience and recreation through a diversified and integrated forestry development model [18,19,20,21,22,23]. Prior research has revealed that the Chinese forestry industry has undergone an aggregate paradigm shift since the inception of the economic reform policy in 1978 and has facilitated development [24]. Hence, this study examines the operating performance and efficiency of the selected forestry enterprises and their corresponding contributions to the economy [18,22,23]. Operating performance is described by evaluating the results of enterprises with their relative assets employed to achieve target objectives [25,26,27]. Operating performance denotes an enterprise’s ability to utilize its available resources efficiently. In contrast, operating efficiency assesses the allocated resources based on the ratio of the output achieved from an enterprise to the input required to effectively run the enterprise [28,29]. This process helps improve the input-output ratio in determining operational efficiency [18,22,23]. Therefore, assessing operational efficiency helps enterprises minimize waste and yield quality products and services [30,31].

This study contributes to the existing body of knowledge by analyzing the paradigm shift in the forestry industry based on the economic reform policy established in China in 1978. Furthermore, it bridges the research gap created in this vital field through enhanced findings based on the super-efficiency data envelopment analysis (DEA) and Malmquist index models. Additionally, this study also considers stochastic frontier analysis (SFA) to check the robustness of the operating performance of the forestry industry in China [32,33]. The SFA model has been demonstrated as critical for checking robustness [32,33]. This research combines the super-efficiency DEA evaluation and Malmquist index methodologies to comprehensively evaluate the operational performance of listed forestry companies in China from static and dynamic perspectives. This study considers the dynamic efficiency of enterprises to examine the optimal investment with respect to time and optimal costs of enterprises in the short run [34,35]. Thus, based on time, dynamic efficiency is associated with improved productive and allocative efficiencies. These demonstrate the ability to develop better or new products and identify effective ways to provide services and produce goods. Static efficiencies are usually estimated for units that occur at equal periods [34,36].

Moreover, these standard models in connection to DEA were selected because of their inherent theoretical and empirical merits [37,38]. These models help analyze data based on constant returns to scale (CRS) and variable returns to scale (VRS) to achieve efficiency in estimation [37,38,39,40]. Again, this research employs super efficiency DEA, which helps to generate scores for ranks using the Banker, Charnes, and Cooper (BCC) and Charnes, Cooper, and Rhodes (CCR) models [37,38]. Thus, super-efficiency DEA permits the utilization of the traditional DEA software and guarantees feasible solutions for decision-making units (DMUs) [37,38,39,40].

The current study supports the notion that increasing the harvesting of wood products as a superior climate change mitigation measure is better than harvesting less for carbon storage [41]. Therefore, policymakers in China highly prioritize implementing ecological conservation, restoration, and management initiatives to mitigate increasing climate change and environmental hazards and support rural revitalization [42]. China has been implementing forestry development strategies while focusing on ecological construction, maintaining a sound ecology, and developing conservation policies. Low environmental hazards characterize forestry industries, which explains why enterprises adopt a responsible approach to ensure a sustainable environment [43,44,45]. Thus, the forestry industry demonstrates massive economic and environmental improvements in efficiency over time, indicating solid production outcomes [46]. Hence, the forestry industry plays a dual role in mitigating climate change and increasing regional economic output values [45,47]. For instance, the State Forestry Administration in China indicated that the total forestry output value in 2019 reached CNY 7.560 trillion, equivalent to USD 1.08 trillion, contributing to the country’s economic growth [48]. However, the industry was affected by the economic crisis engendered by the coronavirus disease 2019 (COVID-19) pandemic, which led to a decline in forestry output value [49]. Thus, approximately USD 160 million in funding is projected to be required globally to ensure sustainable forest management during the COVID-19 pandemic period [50]. However, recent statistics show improvement and recovery from the economic impact of the COVID-19 pandemic [51,52]. The continuous deepening of China’s opening up to the outside world has promoted cooperation in forestry related industries and transformed ecological products into productivity [53]. Thus, the advocacy for green infrastructure to ensure economic development has largely been emphasized in the academic literature [54,55]. However, a critical factor influences the transformation of green infrastructure to policy and practice guidance through planning and development: lack of solid decision adoption among practitioners regarding the implementation of the standard and high-quality green infrastructure within the built environment [55,56].

This study emphasizes the import and export of essential products in the forestry industry for analysis, as highlighted in this research [7]. The import products are logs, wood pulp, and paper products, while the export products include wood-based panels, wood, and bamboo furniture. The process of combining “input” and “output” or the “import” and “export” to achieve high-quality forestry policy standards for sustainable development is a significant issue confronting the development of forest products in China. Forestry development is conducive to optimizing and upgrading China’s industrial structure, transforming the economy, and ensuring sustainable economic growth. Forestry enterprises serve as representatives of the forestry industry and sources of product assetization. Thus, listed forestry companies directly impact the sustainable development of the forestry industry. Therefore, assessing the scientific performance of forestry companies is vital for determining development patterns in China.

Specifically, this study investigates the influencing factors and performance of forestry industries in China and the pathway of sustaining these industries to achieve economic development. In addition, this research may be considered a guide for forest stakeholders in their decision making in connection with alternative administrative practices. This study contributes to the existing literature in five ways. First, it adopts the super-efficiency DEA and Malmquist index methodologies as well as the SFA model to evaluate and analyze the operating performance of 32 listed forestry companies in China. Second, this study’s analysis was conducted from static and dynamic efficiency perspectives. Third, this study bridges the methodological and theoretical research gaps in analyzing the performance of Chinese forestry industries to discover companies’ growth patterns in resource allocation, utilization, and management. Fourth, this study provides three crucial findings that can assist stakeholders in decision making. Fifth, this research offers governmental and institutional policies at the macro and micro levels to curb operational performance issues. Furthermore, corresponding countermeasures to improve efficiency within forestry enterprises are outlined. Aside from the introduction, the rest of the study is organized as follows: Section 2 reveals the literature of the study; Section 3 indicates the methodology; Section 4 demonstrates the input and output indicators and data sources; Section 5 shows the empirical results; Section 6 represents the discussion of the study, and Section 7 reveals the conclusion, policy implication of the study, future scope, and limitations of the study.

2. Literature

In assessing the improvement and management of enterprises and forestry industries, many researchers have attached great importance and focused on evaluating and identifying the feasible factors in improving forestry management and related enterprises by adopting several indicators and methodologies for analysis. This study reveals the summary of gaps in the literature conducted on forestry industries and firms’ efficiency performance through countrywide analysis based on methods, variables used for analysis, and key findings from previous studies, as demonstrated in

Table 1. Comparative literature review and study gap summary analysis.

Literature	Region/Country	Methodology Used	Variables Analyzed	Key Related Findings
[1]	Congo	Contribution analysis approach	Timber organizations, government agencies, and NGOs	Contribution analysis is critical for tracing several causal pathways through a web of interactions for marginal and indirect contributions.
[3]	Central Italy	Conservation methods	Multi-span greenhouse	The conservation approach is suitable for sustainable building systems, energy efficiency, and reduction in energy consumption.
[4]	China	The extended economic model and space econometrics	Forestry products, Labor and Capital	The study revealed a statistically significant spatial correlation in China’s forestry products.
[6]	China	Terrestrial laser scanning and hemispherical photography	Forest canopy, woods	The findings revealed the ability of terrestrial-based algorithms to capture the leaf area index of forest stands at varying densities.
[7]	China	Entity value input-output models, Forest resource input-output model	Paper products, furniture, and other timber products	The study showed that the demand for forest products differs significantly across industries in China.
[12]	Pakistan	Interview, questionnaire survey, textbooks, and Internet materials	Forest products: timber and fuelwood	The study showed that forest is beneficial in controlling erosion, improving aesthetic beautification, and regulating temperature.
[27]	China	Grey relation analysis and DEA approach	Forest products	The findings revealed the average comprehensive efficiency of fourteen companies represented by 93% and 7% of waste resources.
[28]	Portugal	DEA and Malmquist’s index models	Technology firms of Madan Parque	DEA and Malmquist index models are relevant for measuring firms’ efficiency and productivity change.
[29]	China	Three-stage DEA and stochastic frontier analysis	Cultural and creative industries	Three-stage DEA and stochastic frontier methods are critical for analyzing enterprises’ efficiencies.
[57]	United Kingdom	Case study	Forest products, ecosystem, and services	The study revealed that improving our comprehension of human-modified tropical forest ecology with microclimate knowledge supports ecosystem conservation and restoration.
[58]	Sweden	Survey and triangulation approach	Forest products	Rehabilitation in forest management improves the mental and physical health of individuals.
[59]	China	Case study	Forest products, landscape, and ecosystem dimensions	Urban forestry is an approach to adapting natural resource management.
[60]	Tennessee	Harvesting and transportation survey	Forest products	The research findings reveal the variations in the input used for costing harvesting operations and the difficulty of harvesting systems.
[61]	United States of America (USA), Canada, Taiwan	Data envelopment analysis and stochastic frontier analysis	Forestry product data from Web of Science	The study revealed that forest industries are interested in competitive benchmarking, while forest management organizations focused on applied benchmarking for internal analysis.
[62]	United States of America	Network data envelopment analysis and slacked-based network model of super efficiency	Data from real commercial banks in the USA	Network DEA has the propensity to rank efficient DMUs than the conventional DEA.
[63]	China	Super efficiency slacked-based measure model	Forestry ecological efficiency, forestry fixed assets	The findings demonstrated China’s forestry ecological efficiency is low in the northeast but increased in the south of China.

. Based on the objective of this research, it is apparent that studies have been conducted on forestry product consumption and management at the international and domestic levels, which furnishes an excellent foundation for further studies to enhance the efficiency measurement of forestry industries in China. This research bridges the gap created and contributes to the existing research by employing super-efficiency DEA, Malmquist index, and SFA methodologies for analysis.

Furthermore, the embodiment of forest ecology is vital for the growth and functioning of the forest ecosystem, controlling the microclimate and water balance, and providing habitats for organisms [57]. Within a forest ecosystem, forest ecology helps establish the science of how organisms interact with the environment. Forest ecology and diversity play a key role in enhancing human activities, such as teaching and recreation in a forest environment [58]. Forestry resource management techniques have incorporated interdisciplinary, multifaceted, and tremendous technological advancements in controlling hazardous activities that endanger the environment [59,64]. Thus, the long-term supply of forest products and their measurement process, including timber and pulpwood, are critical components in determining the long-term profitability of forest operations and performance [60,65,66]. To reconcile environmental protection with economic development goals, policymakers have prioritized the ecological modernization concept of developing green infrastructure in strategic spatial plans as a potential for the growth of forestry enterprises while employing diverse methodologies for analysis [61,67,68,69,70].

Hence, extant studies have indicated the broad application of using different methodologies to measure the operating performance of listed companies at the industrial level: gray correlation, factor analysis, and DEA [62,71,72,73,74,75]. For instance, in determining the operational performance of forestry companies in China and the five critical factors affecting profitability, asset operation capacity, growth capacity, debt repayment capacity, and equity expansion capacity, the factor analysis and DEA methods were employed for analysis [76]. Additionally, factor analysis has been used to evaluate the comprehensive performance of 22 listed companies in China’s small and medium enterprises by ranking the companies’ performance based on average scores [62]. Furthermore, the super-efficiency DEA method has been utilized to evaluate management performance in listed logistics companies to counter the inherent limitations of the traditional DEA method [63]. Some studies have also applied the value-added economic approach to assess the performance of indigenous industries. The improved evaluation model supports their findings in determining the operating performance of listed companies in Shanghai and Shenzhen [77,78]. Recent tourism and forest assessment investigations have applied a new set of DEA approaches, including dynamic network data envelopment analysis and microdata [79,80].

Additionally, studies have applied parametric and non-parametric approaches to quantify productivity growth, efficiency, and outsourcing in manufacturing and service industries in the context of static, dynamic, and firm-specific modeling [5,33,81,82]. Their study revealed efficient methodologies for measuring productivity [83]. Again, research has demonstrated that integrating the SFA and DEA methodologies to measure enterprises’ total factor productivity (TFP) helps countercheck whether the findings obtained can be verified [84,85]. Nevertheless, these studies indicated unfavorable and uncoordinated industry development compared with development capacity and operational level, profitability and solvency, and insufficient debt financing capacity. It has also been established that employing the super efficiency DEA model with the Malmquist index methodology to analyze the overall operating performance of listed forestry companies from a dynamic and static perspective furnishes credible and accurate findings [26]. Again, the gray correlation and the DEA method were used to measure listed forestry companies’ input and output indicators to determine their performance [27]. The study’s findings reveal that forestry companies must improve their efficiency to remain competitive in a sluggish market environment and to reduce ineffective resource utilization [65,86]. Research has indicated that establishing relationships between research variables requires the application of regression models that provide accurate information about the connection linking single or multiple independent variables and a target variable [5,8,9,87,88]. Furthermore, research has demonstrated that applying truncated regression methodologies, such as the Tobit regression and DEA model, to investigate the influencing elements of enterprises’ performance yields accurate and precise findings [5]. Thus, advanced regression models help analyze multiple samples, achieve consistency in estimations, and identify the disparities between variables compared to conventional regression approaches [5]. The DEA-Tobit model was used to evaluate the pharmaceutical, sports, machinery and equipment, agriculture, food, and beverage industries in China, and it revealed dynamic results [89,90,91]. Moreover, factor analysis was employed to measure listed forestry companies’ profitability, debt servicing, operation, and development [92].

From the aforementioned literary works, there is evidence that some scholars have used many methodologies to evaluate the operating performance of listed forestry companies [62,89,93,94,95]. Thus, factor analysis and DEA are the most frequent performance evaluation methods applied to test listed forestry companies. In addition, most results were inconsistent. The factor analysis and traditional DEA methods have limitations to a certain extent. The factor analysis method uses financial data for evaluation based on economic indicators, primarily to compress most information from multiple indicators into fewer indicators to measure the enterprise’s comprehensive performance. The conventional DEA approach can select a representative sample of input-output indicators based on specific research purposes [28,87,96]. It can only estimate the potential of particular aspects of enterprises. Furthermore, the traditional DEA evaluation method can only be analyzed from a static perspective and cannot reflect the dynamic development trends of the entire industry. Thus, studies have shown that multiple decision-making units would be relatively effective simultaneously when the super-efficient DEA model is applied [26]. Research also suggests that the SFA model is adequate to obtain the robustness of the conventional and super-efficient DEA methods [29].

Scholars have applied extant methodologies to assess the performance of forestry enterprises. This study is unique in that it provides enhanced results based on the super-efficiency DEA evaluation method and the Malmquist index method. Therefore, this study combines the super-efficiency DEA and the Malmquist index methodologies to comprehensively evaluate the operating performance of listed forestry companies in China from a static and dynamic perspective lacking in most studies. Furthermore, the SFA approach is used for the robustness check, which helps add new knowledge to the existing literature.

3. Materials and Methods

3.1. Super Efficiency DEA Model

The new integrated super-efficiency DEA model is a widely recognized technique and mathematical programming approach for finding the most efficient DMU [81,97,98]. The super-efficiency DEA model can be divided into the CCR model with a CRS and the BCC model with a VRS. The CCR model is a recognized assumption for measuring the overall efficiency of a DMU [28,39,82,97]. Thus, changes in scale and technical efficiency marginally affect the actual output. Therefore, this study chose the CRS and super-efficiency DEA models to measure the comprehensive technical efficiency of each listed forestry company [99,100]. Studies have demonstrated that the DEA represents an analytical tool that supports identifying good practices in managing organizational resources [101,102]. Therefore, this research employed an advanced form of DEA, the super efficiency DEA, to produce extremely efficient findings and the desired effects by minimizing inefficiencies [101,102]. The principal benefit of the super-efficiency DEA is its capacity to accommodate multiple inputs and outputs and consider returns to scale when calculating efficiency. It also permits decreasing or increasing efficiency conditioned on output levels and size and does demand any assumption concerning productivity assessment [101,102]. In adopting the traditional DEA model to evaluate DMUs, multiple relatively effective conditions exist in the evaluation results [38]. This study employs the super-efficiency DEA model with each relatively effective DMU rank. The basic assumption is that when evaluating a certain DMU, it is appropriate to rank and exclude the DMU under evaluation. The rank of a given DMU is obtained by using other units. Thus, there may be an instance where $n$ listed companies may be evaluated based on their input and output factors, and each company $j(j=1,2,\dots ,n)$ uses $m$ kinds of inputs $X_{ij}(i=1,2,\dots ,m)$ to produce $s$ types of outputs $Y_{rj}(r=1,2,\dots ,s).$ Therefore, this study expresses the super-efficiency DEA model as shown in Equation (1):

$$\begin{gathered} \min \theta \\ s.t.{\sum }_{j=1}^n{\lambda }_j{X}_{ij}\le \theta X_{ik}+S_i^{-} \\ j=1 \\ \sum _{j=1}^n{\lambda }_j{Y}_{rj}\ge Y_{rk}-S_r^{+} \\ \sum _{j=1}^n{\lambda }_j=1 \\ \lambda \ge 0 \\ \theta ,{\lambda }_j,S_i^{-},S_r^{+}\ge 0,j=1,2,\dots ,n \end{gathered}$$

(1)

where $\mathrm{m}\mathrm{i}\mathrm{n}\theta$ is the super-efficiency value of the decision-making unit $\mathrm{D}\mathrm{M}\mathrm{U}$, and ${\lambda }_j$ is the combination ratio of reconstructing an effective $\mathrm{D}\mathrm{M}\mathrm{U}$, while, $S_i^{-}$ and $S_r^{+}$ are considered slack variables [75,82,103,104,105]. Additionally, $\theta$ denotes that $0$ ≤ $\theta$ ≤ 1. Thus, if $\theta$ = 1, the DMU is on the verge of the efficient frontier and demonstrates validity. Contrarily, if $\theta$ < 1, it demonstrates invalidity. In contrast, $1$ − $\theta$ indicates the excess proportion of inputs. Furthermore, $\lambda$ denotes the combined ratio and relative importance in measuring the scale of a DMU. $\sum \lambda <1,\sum \lambda =1$ and $\sum \lambda >1$ represent the increase, invariance, and reduction in scale benefits [106,107,108]. $S-\mathrm{a}\mathrm{n}\mathrm{d} S+$ represent the relaxation and residual variables, respectively, while $X_{ik}$ and $Y_{rk}$ denote the inputs and outputs of $\mathrm{D}\mathrm{M}\mathrm{U}.$ $\sum _{j=1,j\ne k}^n{\lambda }_j=1$ represents an additional constraint based on the super-efficiency model.

3.2. Malmquist Productivity Index

The Malmquist productivity index is widely used in the financial, industrial, medical, and other sectors to measure production efficiency [109,110] Famous economist Sten Malmquist proposed the Malmquist productivity index as a consumption index in 1953 [111,112]. Later, this methodology was applied to a productivity analysis [113]. Another study constructed the Malmquist index as a distance function model that performed linear operations through multiple directional outputs and input variables [114]. The Malmquist index methodology is essential for evaluating relative efficiency in the absence of production function. It can compare vertical data at different periods and is also critical for measuring dynamic efficiency relating to panel data [115]. The benefit of the Malmquist index approach in assessing productivity leading to its most extensive usage by researchers is that it demands a limited quantity of outputs and inputs for assessment without requiring modifications [116]. Furthermore, the evaluator does not need to minimize the inputs or maximize the outputs and is not dependent on fixed weights for the output and input indicators [116]. The Malmquist index is relevant when comparing varieties of industries and initiating policies [116]. It further suggests the core reason for this current study to employ the Malmquist index on the selected data to furnish in-depth knowledge of 32 companies in the Chinese economy.

This study is based on the productivity distance function model [114]. This research is premised on the Malmquist productivity index technology assumption of the CRS [116]. Therefore, it may be described as the product of the pure technical change index, scale efficiency (SE), and congestion change indexes [110,116,117]. Specifically, the CRS is essentially applied as a fundamental assumption to benchmark technology and further incorporate the VRS technological approach to quantify the unit of the SE change between two points [110,116,117]. The Malmquist index measures the relationship between technical efficiency changes, technological progress, and the TFP. The Malmquist model is combined with the DEA method to measure the dynamic changes in relative efficiency [116,117]. Therefore, this study employed these two methodologies to investigate the selected 32 forestry companies from 2014 to 2019 to determine their TFP efficiency and growth in the forestry enterprises [116]. To better analyze the dynamic changes of DMUs in this study, the TFP index model is chosen to conduct a longitudinal analysis of the indicator data [110,116,117].

Thus, under the technical conditions concerning years, the change in technical efficiency from distance $x$ to distance $t$ can be expressed in Equation (2) as:

$$M_s=\frac{D_0^s(x^t,y^t)}{D_0^s(x^s,y^s)}$$

(2)

where $M$ denotes the Malmquist total productivity, and the subscript $s$ represents the distance function at the time. $D$ represents the input distance function at time $t$, and the subscript $o$ denotes an input orientation. Hence, $s$ and $t$ are adjacent periods. $(x^t,y^t)$ and $(x^s,y^s)$ are production points concerning input $x^t$ and output $y^t$ within the technology period, respectively.

Additionally, under the technical conditions in year $t$, the change in technical efficiency from period $s$ to period $t$ can be expressed in Equation (3) as:

$$M_t=\frac{D_0^t(x^t,y^t)}{D_0^t(x^s,y^s)}$$

(3)

where $M$ denotes the Malmquist total productivity, and the subscript $t$ represents the time. $D$ indicates the input distance function at period s, and subscript $o$ indicates the input orientation. $(x^t,y^t)$ and $(x^s,y^s)$ are production points concerning input $x^t$ and output $y^t$ within the period of technology, respectively. The TFP from period $s$ to period $t$ can be expressed in Equation (4) as:

$$M_{st}(x^s,y^s,x^t,y^t)=\sqrt[2]{\left(M_s\times M_t\right)}=\sqrt[2]{\left[\frac{D_0^s(x^t,y^t)}{D_0^s(x^s,y^s)}\times \frac{D_0^t(x^t,y^t)}{D_0^t(x^s,y^s)}\right]}$$

(4)

Thus, the total productivity change index multiplies the SE and technological change index. These two variations constitute the total productivity index. Furthermore, in 1997, Ray and Desli proposed a decomposable efficiency change index (RD) model. Based on these two indexes, the decomposition form adopted for this study is expressed in Equation (5) as follows:

$$\begin{array}{ll}{M}_{RD}(x^s,y^s,x^t,y^t)& =\sqrt[2]{[\frac{D_0^s(x^t,y^t)}{D_0^s(x^s,y^s)}\times \frac{D_0^t(x^t,y^t)}{D_0^t(x^s,y^s)}]}\\ & =\frac{D_1^t(x^t,y^t)}{D_0^s(x^s,y^s)}\times \sqrt[2]{[\frac{D_0^s(x^t,y^t)}{D_0^s(x^s,y^s)}\times \frac{D_0^t(x^t,y^t)}{D_0^t(x^s,y^s)}]}=EC\times TP\end{array}$$

(5)

where EC is the technical efficiency change index, and TP is the technological progress change index. If $M_{st}>1$, the production efficiency has improved compared with the previous period. $M_{st}=1$ means that the production efficiency remains unchanged, and $M_{st}<1$ means that production efficiency is reduced. Furthermore, the production efficiency index can be decomposed into the technical efficiency change index and the technological progress change index. A technical efficiency change index or the technical progress change index > 1 promotes overall production efficiency. If it is $<$1, this factor is the root cause of the decrease in overall production efficiency.

4. Selection of Input-Output Indicators and Data Sources

4.1. Evaluation Indicators

The super-efficiency DEA and Malmquist index models have been widely used for multi-input and multi-output index measurements [84]. While applying the super-efficiency DEA evaluation model, it is crucial to establish an excellent input-output index system [75]. Thus, the number of evaluation indexes is at most 1/2 of the DMUs. There can be no strong correlation between similar indicators to avoid overshadowing the indicators; the output indicators continuously provide the input indicator requirements of the Malmquist index model to the decision-making unit within a certain period [26]. The evaluation results will be gravely affected if there is a strong correlation between the input and output indicators. Therefore, the number of input and output indicators should not be too much; simultaneously, a strong correlation between indicators should be avoided. In selecting the indicators of this study, the researchers emphasized the relevant analysis of the Cobb–Douglas production function Caves et al., 1982 [113] and the comprehensive analysis of this research [118,119,120]. The authors adopted the index selection principles such as systematicity and feasibility. They combined the characteristics of the forestry industry and raw materials processed into forestry end products to construct an input-output index system for the listed companies [119].

4.2. Input and Output Indicators

This study used the three main production factors, capital, labor, and land, to measure the variables by employing the super-efficiency DEA model [120]. Total assets (capital), number of employees (labor), and operating costs (investment in the acquisition of land) were selected as input indicators, and operating income and net profit were chosen as output indicators. All assets owned or controlled by the enterprise could bring economic benefits, reflecting the company’s strength [119,120]. The number of employees was considered the number of workers (labor) on the company’s register at the end of each year. This reflected the investment in human resources in an enterprise’s business activities. Operating costs were the cost of selling goods or providing labor services, which could reflect the direct input of the enterprise in production activities. Operating income was the direct income of the enterprise. Net profit reflected the profitability and operation level of the enterprise.

4.3. Data Sources

The forestry industry involves multiple sectors of the national economy’s first, second, and tertiary sectors, covering several raw materials and finished forestry products. This study analyzed 32 listed forestry companies on China’s Shanghai and Shenzhen A-share stock exchanges [119,120]. The enterprises whose main business activities were forestry and processing of forest by-products based on the Industry Classification Guidelines for Listed Companies revised in 2012 by the China Securities Regulatory Commission were employed as research objects. Data for the indicators were gathered from the financial statements published by the Shanghai and Shenzhen A-share stock exchanges from 2014 to 2019. The total assets (capital) data were obtained from the balance sheet, and information on operating costs, operating income, and net profit was obtained from the income statement.

The main business areas of the selected 32 listed companies differed based on the availability of various forestry products [119,120]. The core business divisions of the companies were categorized into forestry, wood processing, wood, bamboo, rattan, palm, grass products, furniture manufacturing, paper, and paper products. Among the forestry industries, there were two prominent listed companies in the forestry category; six listed companies in the wood processing and wood, bamboo, rattan, palm, and straw products category; four listed companies in the furniture manufacturing category; and 20 listed companies in the paper products category. From the regional distribution perspective, most listed forestry companies were located in East and South China, of which 19 were in East China and seven in South China. The other six were over the remaining regions. Supplementary Table S1 provides an overview of the specific enterprises and regions.

Integrating these products helped determine forestry products’ contributions to employment, the general growth of the economy, diversification purposes, and the shift to using more productive resources in forestry.

5. Empirical Results

The super efficiency DEA model was first applied to determine the comprehensive technical efficiency, pure technical efficiency (PTE), and SE of the 32 listed forestry companies from 2014 to 2019. Furthermore, an analysis of the annual static efficiency was performed. The TFP index model was applied to determine its TFP value, decomposed technological progress change index, and technological change efficiency index. Again, the authors analyzed the dynamic efficiency changes during performance evaluation using the Malmquist model. After selecting the DEAP version 2.1 software to measure the operating efficiency from 2014 to 2019, multiple DMUs with a total efficiency value of 1 were chosen as a measurement criterion [121,122]. The rankings of the various enterprises were compared, and the MaxDEA8 software was combined to calculate the super-efficiency DEA dynamic efficiency value.

5.1. Super-Efficiency DEA Static Efficiency Evaluation

The comprehensive technical efficiency value is obtained by multiplying the PTE and SE values. The PTE reflects the production efficiency of the input elements of the DMU at an optimal scale. Additionally, the PTE is affected by factors such as corporate management and technology. Total production and SE reflect the gap between the actual and optimal production scales. The overall efficiency value reflects the comprehensive combination of the DMUs to measure the resource utilization efficiency of the 32 listed forestry companies from 2014 to 2019, as shown in

Table 2. Comprehensive operating efficiency values and rankings of the 32 companies.

Company	Year						Mean Value	Rank
	2014	2015	2016	2017	2018	2019
Fujian Jinsen	2.804	1.863	2.232	4.273	4.002	1.326	2.750	1
Meike Home	1.390	1.467	1.430	1.274	1.265	1.235	1.343	2
Sophia	1.514	1.451	1.124	1.352	1.210	1.393	1.341	3
Baby rabbit	1.215	1.210	1.386	1.343	1.263	1.271	1.281	4
Sun Paper	0.998	1.194	1.203	1.044	1.066	1.438	1.157	5
Da Ya Shengxiang	1.070	1.227	1.109	1.135	1.138	1.015	1.116	6
Shun Hao	2.410	0.864	0.905	0.823	0.831	0.827	1.110	7
Del Future	1.601	1.351	1.247	0.849	0.807	0.785	1.107	8
Jingxing Paper	0.779	0.830	0.931	1.231	1.182	1.222	1.029	9
Hexing Packaging	0.967	0.850	0.781	0.931	1.203	1.383	1.019	10
Chenming Paper	0.807	1.035	1.053	1.060	1.025	1.064	1.007	11
Qifeng New Material	1.121	0.968	0.879	0.887	0.892	0.929	0.946	12
Zhongshun Jiezuo	0.846	0.874	0.919	0.904	1.006	1.116	0.944	13
Mountain Eagle Paper	0.750	0.823	0.862	0.926	1.287	0.936	0.931	14
Huatai shares	0.796	0.822	0.860	0.933	0.967	0.980	0.893	15
MeiYingSen	1.008	0.824	0.729	0.822	0.865	1.099	0.891	16
Fenglin Group	0.868	0.849	0.899	0.871	0.871	0.953	0.885	17
Bohui Paper	0.802	0.887	0.895	0.941	0.845	0.936	0.884	18
Yueyang Forest & Paper	0.757	0.810	0.837	0.873	0.913	0.901	0.848	19
Annie shares	0.763	0.853	0.789	0.812	0.889	0.881	0.831	20
Guanhao High-tech	0.732	0.779	0.845	0.795	0.870	0.964	0.831	21
Kane	0.865	0.812	0.795	0.816	0.818	0.853	0.827	22
Xilinmen	0.859	0.861	0.792	0.808	0.790	0.838	0.825	23
Yibin Paper	0.955	1.218	0.494	0.717	0.761	0.770	0.819	24
Hengfeng Paper	0.836	0.810	0.783	0.780	0.769	0.805	0.797	25
Minfeng Special Paper	0.788	0.767	0.790	0.769	0.748	0.778	0.773	26
Weihua	0.821	0.767	0.732	0.733	0.792	0.762	0.768	27
Merryun	0.549	1.177	0.602	0.685	0.712	0.723	0.741	28
Zhejiang Yongqiang	0.715	0.765	0.729	0.720	0.715	0.794	0.740	29
Qingshan Paper	0.689	0.711	0.778	0.733	0.745	0.716	0.729	30
Jilin Senkou	0.591	0.570	0.564	0.738	0.845	0.964	0.712	31
Pingtan Development	0.749	0.730	0.689	0.662	0.714	0.706	0.708	32
Average	1.013	0.969	0.927	1.008	1.025	0.980	0.987

. It shows that the top 11 companies in the average rank have production and operation efficiencies greater than 1, underlining that the super-efficiency DEA model is the most appropriate when measuring and comparing enterprises’ production and operation efficiency with values greater than 1 [96,122]. The enterprises’ aggregate average comprehensive technical super-efficiency value of the 32 listed forestry companies is 0.987 (less than 1), indicating the possibility of improvement. This exciting finding demonstrates that the comprehensive technical efficiency result of 0.987 shows a gain of approximately 1.3%, while the non-operational efficiency is less than 1.3%. Therefore, policymakers must implement effective strategic measures to enhance the performance of enterprises. Nevertheless, the average fluctuation range of the production and operation efficiency of the listed forestry companies from 2014 to 2019 was relatively stable.

Simultaneously, the dynamic results of the enterprises’ average efficiency values revealed that 11 out of the 32 companies, including Fujian Jinsen 2.750, Meike Home Furnishing 1.343, Sophia 1.341, Tubao 1.281, Sun Paper 1.157, Da Ya Dekor 1.116, Shun Hao shares 1.110, Del Future 1.107, Jingxing Paper 1.029, Hexing Packaging 1.019, and Chenming Paper 1.007 performed better. However, the minimal fluctuation range of these dynamic trends indicated a gradual decline in performance. Therefore, the integrated super-efficiency DEA model shows an enhanced average production and operation efficiency from 2014 to 2019. Thus, the scores obtained for the companies reached an effective state; this signified that the resource allocation capability and resource utilization efficiency of the 11 listed forestry companies during the performance evaluation period achieved the desired level of technology [82]. Based on the results obtained for the enterprises using the super-efficiency DEA model, Fujian Jinsen ranked first in average overall efficiency; its comprehensive efficiency values for each year were more promising. The high performance of Fujian Jinsen can be attributed to its importance in agricultural industrialization in Fujian Province, which has abundant and diverse forest resources, a wide range of operations, and a complete forestry industry chain. It comprises a dual-effect management model that improves efficiency and benefits, and its forest coverage rate is 82.1%. It has low logging and investment costs, and its forest area is easy to cultivate and harvest. In addition, reasonable allocation and use of resources are highly sustainable.

However, the overall efficiency value of the remaining 21 listed forestry companies could not attain the super-efficient DEA effective value of 1 [82]. The enterprises’ inability to reach the efficiency threshold of the super efficiency DEA value may be linked to the impact of the PTE, the SE values not reaching the optimal standard, or the combined influence of the technical and SE. This result suggests that enterprises whose comprehensive efficiency values have not yet reached the effective production state must adjust their input structure appropriately [123].

Furthermore, Table 2 underscores that the comprehensive efficiency value of Fujian Jinsen has a more significant impact on the average comprehensive efficiency within the analysis period. For instance, the result revealed that the comprehensive efficiency value of 2019 was lower than 2018. However, a critical observation of all 32 forestry companies’ total efficiency values pointed to improvement. Furthermore, the overall efficiency value of the 21 listed forestry companies with the most negligible value demonstrated an upward trend compared with the records of 2018. Notwithstanding, the remaining ten forestry companies with high-efficiency performance also experienced a downward trend, with a maximum difference of 0.123. These findings demonstrate that the decline in the overall efficiency value of Fujian Jinsen company significantly impacts the average efficiency value of the other forestry enterprises. When Fujian Jinsen is excluded, the overall efficiency average of the listed forestry companies needs to be recalculated, as shown in Figure 1. Thus, Figure 1 demonstrates the evaluation results’ trends after excluding the Fujian Jinsen company. It further illustrated that excluding Fujian Jinsen led to disparities in performance: it displayed a downward trend from 2014 to 2016 and an upward trend from 2016 to 2019. The dynamics in movement exhibit inconsistency during the period Fujian Jinsen is not excluded when their average overall efficiency is observed. The discrepancies highlight the impact of the overall efficiency value of Fujian Jinsen, which declined in 2019, causing a total decline in the overall average efficiency of the selected forestry companies.

Table 3. Performance evaluation results of the companies in 2019.

Serial No.	Company	crste	vrste	Scale	rts
1	Pingtan Development	0.706	0.720	0.980	irs
2	Zhejiang Yongqiang	0.794	0.814	0.975	irs
3	Xilinmen	0.838	0.854	0.981	irs
4	Weihua	0.762	0.782	0.973	irs
5	Bunny	1.271	1.348	0.943	irs
6	Sophia	1.393	1.645	0.847	drs
7	Qifeng New Material	0.929	0.963	0.964	irs
8	Meike Home	1.235	1.271	0.972	drs
9	Jilin Senkou	0.964	0.978	0.986	drs
10	Fujian Jinsen	1.326	6.400	0.207	irs
11	Fenglin Group	0.953	0.988	0.964	irs
12	Del Future	0.785	0.843	0.930	irs
13	Da Ya Shengxiang	1.015	1.016	0.999	irs
14	Sun Paper	1.438	3.033	0.474	drs
15	Shun Hao	0.827	0.859	0.964	irs
16	Shan Ying Paper	0.936	0.942	0.993	drs
17	Qingshan Paper	0.716	0.717	0.998	drs
18	Chenming Paper	1.064	1.000	1.064	drs
19	Yibin Paper	0.770	0.829	0.929	irs
20	Zhongshun Jiezuo	1.116	1.127	0.991	irs
21	Yueyang Forest Paper	0.901	0.908	0.993	irs
22	Minfeng Special Paper	0.778	0.909	0.857	irs
23	Meiying Sen	1.099	1.116	0.985	irs
24	Merryun	0.723	0.754	0.959	irs
25	Kane	0.853	1.155	0.739	irs
26	Jingxing Paper	1.222	1.276	0.958	irs
27	Huatai	0.980	1.095	0.895	drs
28	Hengfeng Paper	0.805	0.860	0.936	irs
29	Hexing Packaging	1.383	1.685	0.821	drs
30	Guanhao High-tech	0.964	1.011	0.954	irs
31	Bohui Paper	0.936	0.937	0.999	irs
32	Annie shares	0.881	0.928	0.949	irs
	Average	0.980	1.243	0.912

Note: crste refers to the comprehensive scale technical efficiency value under the fixed return to scale mode, and vrste refers to the pure technical efficiency value under the variable return to scale technical efficiency method. Scale refers to the scale efficiency value under the variable rts return to scale mode. Again, irs and drs denote increasing return to scale and decreasing return to scale.

shows the performance efficiency values of forestry listed companies obtained in 2019 based on the super-efficiency DEA model. Stemming from the impact of Fujian Jinsen company’s overall comprehensive efficiency average value in 2019, this study re-ranked the super-efficiency DEA for the listed forestry companies in 2019 to determine the DMUs with a total efficiency value above the criterion value 1 [122,124,125]. The results indicated that most high-performance companies attained an efficiency value above criterion value one after re-ranking [122]. The comprehensive efficiency value of Sun Paper reached 1.438 in 2019, Sophia 1.393, Hexing Packaging 1.383, Fujian Jinsen 1.326, Bunny 1.271, Meike Home Furnishing 1.235, Jingxing Paper 1.222, Zhongshun Jiezou 1.116, Meiying Sen 1.099, Chenming Paper 1.064, Da Ya Dekor 1.015. The super-efficiency DEA results of these 11 listed forestry companies are all effective. Interestingly, only Chenming Paper achieved comprehensive technical efficiency and an efficient scale. The remaining 10 companies with effective DEA also achieved pure technical efficiency in the production frontier. However, the SE level could not reach the production frontier threshold, illustrating the extreme imbalance between China’s forestry technology and the SE.

Furthermore, the average comprehensive technical efficiency of the listed forestry companies reached 0.980, demonstrating that there was still 2% room for input. In addition, the average PTE reached 1.243, and the SE average was 0.912. Ineffective SE lowered the overall operating efficiency of China’s forestry listed companies. Again, the results demonstrated that among the 32 listed forestry companies, only Chenming Paper has an effective SE, highlighting that China’s forestry listed companies have not reached the optimal production scale that invariably affected the SE of the companies. Hence, the impact of the ineffective scale led to the overall decline in the SE of China’s forestry industry [126]. Moreover, Fujian Jinsen’s comprehensive technical efficiency ranked highly among the companies, but its PTE value (6.400) and SE value (0.207) were highly unbalanced. Interestingly, Fujian Jinsen’s comprehensive technical efficiency demonstrates its matured forest resources and species structure. Fujian Jinsen has a comparative advantage in forest management and technological advancement. However, its small-scale SE value restricts its technological benefits. This research emphasizes that Fujian Jinsen’s company should broaden its business scope, expand its SE, and improve its profitability to improve company performance.

In summary, 23 forestry listed companies had increasing returns to scale, and nine demonstrated diminishing returns to scale when using the super-efficiency DEA model [88,127]. Companies in a state of increasing returns to scale should increase investment, optimize input structure, expand production scale, and improve company operating efficiency [61,68,69,70,115]. Furthermore, companies with diminishing returns to scale have redundant investment, and ineffective management of production and operation costs, leading to a decline in operating efficiency. It implies enterprises should manage their investment accordingly to achieve the target result.

5.2. Malmquist Dynamic Efficiency Evaluation

The dynamics in TFP during the performance evaluation of the forestry listed companies in China are shown in

Table 4. The average TFP results of the 32 companies from 2014 to 2019.

	Total Factor Productivity	Technical Efficiency Change Index	Technological Progress Change Index	Pure Technical Efficiency Change Index	Scale Efficiency Change Index
Pingtan Development	0.986	0.989	0.996	0.982	1.008
Zhejiang Yongqiang	1.027	1.023	1.004	1.025	1.023
Xilinmen	1.014	0.996	1.018	0.984	1.014
Weihua shares	0.999	0.986	1.012	0.986	1.004
Bunny	1.071	1.000	1.071	1.000	1.000
Sophia	1.006	1.000	1.006	1.000	1.000
Qifeng New Material	0.984	0.986	0.998	0.994	0.993
Meike Home	0.988	1.000	0.988	1.000	1.000
Jilin Forest Industry	1.108	1.110	0.997	1.106	1.005
Fujian Jinsen	0.902	1.000	0.902	1.000	1.000
Fenglin Group	1.039	1.020	1.020	1.021	1.002
Del Future	0.887	0.954	0.932	0.968	0.990
Da Ya Shengxiang	1.051	1.000	1.051	1.000	1.000
Sun Paper	1.074	1.000	1.074	1.000	1.000
Shun Hao	0.917	0.965	0.943	0.973	0.993
Shan Ying Paper	1.071	1.047	1.018	1.017	1.029
Qingshan Paper	1.036	1.009	1.026	1.002	1.006
Chenming Paper	1.088	1.048	1.046	1.000	1.048
Yibin Paper	0.944	1.013	0.923	1.011	0.996
Zhongshun Jiezuo	1.056	1.035	1.021	1.034	1.001
Yueyang Forest Paper	1.030	1.036	0.994	1.008	1.030
Minfeng Special Paper	1.011	0.998	1.013	1.019	0.982
Meiying Sen	1.013	1.009	1.007	1.008	1.000
Merryun	1.150	1.123	1.009	1.060	1.036
Kane	1.011	0.998	1.013	1.013	0.991
Jingxing Paper	1.155	1.052	1.093	1.046	1.007
Huatai	1.070	1.043	1.025	1.026	1.018
Hengfeng Paper	1.010	0.993	1.018	1.001	0.993
Hop Hing Packaging	1.052	1.013	1.036	1.008	1.008
Guanhao High-tech	1.055	1.059	0.998	1.057	1.001
Bohui Paper	1.073	1.034	1.038	1.017	1.016
Annie shares	1.039	1.031	1.006	0.991	1.048
2014–2015	1.019	1.033	0.984	1.006	1.026
2015–2016	0.995	0.971	1.022	0.970	1.004
2016–2017	1.142	1.036	1.102	1.043	0.993
2017–2018	1.020	1.025	0.995	1.019	1.006
2018–2019	0.968	1.025	0.944	1.017	1.009
Mean value	1.029	1.018	1.009	1.011	1.008

. As per the analysis dynamics, the TFP of forestry listed companies in China from 2014 to 2019 showed an increasing trend. This finding agrees with this research [128]; however, the growth rate was relatively slow, with an annual growth rate of 2.9%. In addition, the companies recorded an average yearly change of 1.8% in their technical efficiency, strongly driving an increasing TFP. Furthermore, the increase in technical efficiency changes is attributed to the combined growth rate of PTE and SE changes. Thus, the development of China’s forestry listed companies is stable, and the fluctuation range is minimal. It also suggests better utilization of multiple inputs of companies in production and operation [129]. Based on the outcome dynamics and periods of the performance evaluation, the TFP of China’s forestry industry in 2015–2016 and 2018–2019 demonstrated variations, indicating that the sector’s resource allocation standard experienced a dynamic shift within the said period.

Table 4 shows that the TFP of 24 listed forestry companies witnessed growth, while the TFP of the other eight listed companies exhibited a downward trend. The trend performance of the companies revealed that Jingxing Paper’s growth rate is the highest, with a TFP index of 1.155, therefore ranking first. Its technical efficiency change index is 1.052, while its technological progress change index is 1.093, which implies that technical efficiency change influences the growth of Jingxing Paper’s TFP index. The simultaneous development of the technological progress index demonstrates that the company focuses on technological innovation capacity and uses the technology available during performance evaluation well. However, Del Future company had the most significant decline in the TFP index value of 0.887. Compared to the other seven forestry listed companies with a downward TFP trend ranging between 0.9 and 1, its TFP index indicated that forestry listed companies paid little attention to technological progress and technology utilization efficiency. Based on the technical efficiency change index, from 2014 to 2019, the average operating efficiency of the listed forestry companies reached 1.018, exhibiting an increasing trend.

Additionally, this research decomposed the technical efficiency change index into the SE and PTE indexes [130]. The corresponding outcome was more significant than 1, indicating that the PTE and SE change indices jointly determined the rising trend of the technical efficiency index. Furthermore, Table 4 highlights the companies with a purely technical change index of less than 1. These companies include Annie, Xilinmen, Weihua, Pingtan Development, Qifeng New Materials, Shunhao, and Del Future. Their results show that these companies have not fully utilized their existing technologies to improve operating performance. Therefore, it is prudent to enhance their management and operational decision-making capabilities. Listed forestry companies with an SE change index of less than 1, including Kane, Minfeng Special Paper, Hengfeng Paper, Qifeng New Materials, Yibin Paper, Shunhao Stock, and Del Future, increased their input resource elements. However, the operating performance did not reach the anticipated target. The average value of the technological progress change index during the performance evaluation period reached 1.009, demonstrating that technical progress factors positively promoted the performance of listed forestry companies. For instance, companies such as Jingxing Paper, Xilinmen, Chenming Paper, Sun Paper, and other listed forestry companies obtained a technological progress change index greater than 1, indicating that they could transform their management concepts, develop new technologies, and improve their operational efficiency.

In summary, between 2014 and 2019, the selected forestry companies’ operating efficiency improved when they used the Malmquist model, notwithstanding fluctuations that stakeholders should address. These fluctuations may invariably influence domestic and foreign market policies, resulting in the instability of profit margins that may often affect the operating efficiency of enterprises. Stakeholders and companies should collaboratively determine external environment factors that drive performance and strive to improve their innovation capabilities and technical acumen. The innovation and critical strategies implemented will enhance the company’s core competitiveness and minimize the negative impact of unfavorable market conditions as well as policy environments on business performance.

5.3. Robustness Test Result Based on Stochastic Frontier Analysis

The SFA model was employed to recognize if all DMUs were robust, which was eventually proven based on statistical tests against data while making provisions for a statistical error [131]. While testing, the SFA model furnishes a standard where the production interrelationship is quantified as a conditional average; thus, outputs-inputs and other elements in support of the production function [131,132,133]. However, the aggregate deviation from the regression curve is decomposed into statistical noise and inefficiency. Thus, the SFA allows estimating samples representing an industry or firm for every DMU [96]. Again, this research adopted SFA because of its ability to consider stochastic noise in datasets and support the statistical evaluation of hypotheses regarding the degree of inefficiency and production structure [134]. The SFA has the propensity to accommodate random variations to measure consistency and accuracy in working conditions [134].

Therefore, based on the present study’s results obtained by applying the super-efficiency DEA and the Malmquist models, the authors used SFA to confirm the results’ robustness [29].

Table 5. SFA measurement results of enterprise efficiency (with operating income as the output variable).

Stock Code	Year						Average	Rank
	2014	2015	2016	2017	2018	2019
000910	0.994	0.994	0.994	0.995	0.995	0.995	0.995	1
600337	0.989	0.990	0.991	0.991	0.992	0.992	0.991	2
002078	0.964	0.966	0.968	0.971	0.973	0.974	0.969	3
002511	0.964	0.966	0.968	0.970	0.972	0.974	0.969	4
002572	0.945	0.948	0.952	0.955	0.958	0.961	0.953	5
600567	0.893	0.900	0.907	0.913	0.919	0.924	0.909	6
000488	0.883	0.891	0.898	0.905	0.911	0.917	0.901	7
600308	0.854	0.864	0.873	0.881	0.889	0.897	0.876	8
002228	0.848	0.858	0.868	0.877	0.885	0.892	0.871	9
002043	0.844	0.854	0.864	0.873	0.881	0.889	0.868	10
603008	0.834	0.845	0.856	0.865	0.874	0.882	0.859	11
600966	0.831	0.842	0.852	0.862	0.871	0.880	0.856	12
002067	0.817	0.829	0.840	0.851	0.861	0.870	0.845	13
002521	0.803	0.816	0.828	0.839	0.850	0.860	0.833	14
002565	0.790	0.804	0.817	0.829	0.841	0.851	0.822	15
600963	0.789	0.803	0.816	0.828	0.840	0.850	0.821	16
002303	0.781	0.795	0.809	0.821	0.833	0.844	0.814	17
002631	0.730	0.747	0.764	0.780	0.794	0.808	0.771	18
600433	0.727	0.745	0.762	0.778	0.792	0.806	0.768	19
600356	0.709	0.728	0.746	0.763	0.779	0.794	0.753	20
601996	0.705	0.724	0.743	0.760	0.776	0.790	0.750	21
002489	0.704	0.724	0.742	0.759	0.775	0.790	0.749	22
600103	0.692	0.713	0.732	0.749	0.766	0.781	0.739	23
002240	0.678	0.700	0.719	0.738	0.755	0.771	0.727	24
600235	0.669	0.691	0.711	0.731	0.748	0.765	0.719	25
002012	0.656	0.679	0.700	0.720	0.739	0.756	0.708	26
600189	0.557	0.586	0.613	0.639	0.663	0.685	0.624	27
000592	0.503	0.536	0.567	0.595	0.622	0.647	0.578	28
000815	0.469	0.504	0.537	0.568	0.596	0.623	0.550	29
002235	0.463	0.499	0.532	0.563	0.592	0.619	0.545	30
600793	0.437	0.474	0.509	0.541	0.572	0.600	0.522	31
002679	0.206	0.258	0.308	0.353	0.396	0.436	0.326	32
Mean Eff.	0.742	0.759	0.775	0.790	0.803	0.816	0.781

and

Table 6. SFA measurement results of enterprise efficiency (with net profit as the output variable).

Stock Code	Year						Average	Rank
	2014	2015	2016	2017	2018	2019
002078	0.918	0.924	0.931	0.937	0.942	0.947	0.933	1
002572	0.880	0.890	0.899	0.908	0.915	0.923	0.902	2
000910	0.746	0.767	0.787	0.805	0.821	0.836	0.794	3
600567	0.745	0.767	0.786	0.804	0.821	0.836	0.793	4
600337	0.676	0.703	0.728	0.751	0.772	0.791	0.737	5
002303	0.654	0.683	0.709	0.733	0.756	0.776	0.718	6
000488	0.641	0.671	0.699	0.724	0.747	0.768	0.709	7
002511	0.552	0.590	0.624	0.656	0.685	0.711	0.636	8
002043	0.536	0.575	0.610	0.643	0.673	0.700	0.623	9
002067	0.471	0.516	0.556	0.593	0.627	0.659	0.570	10
002521	0.422	0.470	0.515	0.555	0.593	0.627	0.530	11
002489	0.402	0.452	0.498	0.540	0.578	0.613	0.514	12
603008	0.328	0.384	0.435	0.483	0.526	0.566	0.454	13
002631	0.298	0.357	0.411	0.460	0.505	0.547	0.430	14
600308	0.271	0.332	0.388	0.439	0.486	0.529	0.408	15
002228	0.260	0.322	0.379	0.431	0.478	0.522	0.399	16
600966	0.259	0.321	0.378	0.430	0.478	0.521	0.398	17
601996	0.247	0.310	0.368	0.421	0.469	0.514	0.388	18
600433	0.244	0.307	0.365	0.418	0.467	0.511	0.385	19
600356	0.191	0.258	0.320	0.377	0.429	0.477	0.342	20
600103	0.124	0.198	0.265	0.326	0.383	0.434	0.288	21
002565	0.072	0.150	0.221	0.286	0.346	0.401	0.246	22
600963	0.064	0.142	0.214	0.280	0.340	0.396	0.239	23
600793	0.051	0.130	0.203	0.270	0.331	0.387	0.229	24
000592	0.007	0.078	0.155	0.226	0.290	0.350	0.182	25
002012	0.015	0.070	0.148	0.219	0.284	0.344	0.175	26
002240	0.105	0.012	0.072	0.150	0.221	0.286	0.102	27
600235	0.189	0.089	0.002	0.085	0.162	0.232	0.034	28
002679	0.274	0.167	0.069	0.020	0.102	0.177	0.035	29
000815	0.552	0.422	0.303	0.194	0.094	0.002	0.261	30
002235	0.568	0.437	0.317	0.206	0.105	0.013	0.274	31
600189	0.764	0.617	0.481	0.357	0.244	0.140	0.434	32
Mean Eff.	0.237	0.301	0.359	0.413	0.462	0.507	0.380

elaborate on the SFA model efficiency measurement results, with operating income as the output variable for Table 5 and net profit as the output variable in Table 6. Based on the outcome of SFA in quantifying operating income as the output variable in Table 5, 17 enterprises were efficient and revealed an average higher than 0.781. Additionally, the net profit measured as the output variable indicated that the efficiency of 19 enterprises revealed an average higher than 0.380. This demonstrates that the efficiency of the forestry enterprises constantly improved through internal management and external support, which were lower than the 0.987 of the super-efficient DEA analysis demonstrated in Table 2. Notwithstanding, the super-efficient DEA analysis showed that the comprehensive management efficiency of 13 forestry enterprises was higher than the average, rendering the two results homogeneous. Thus, the net profit and operating income efficiency results obtained based on SFA measurement were consistent with the measurement based on the super-efficient DEA, indicating that the operating efficiency of Chinese forestry companies was oriented toward profit efficiency improvement.

With the operating income as the output variable, the efficiency of enterprises with stock codes 000910, 600337, 002078, 002511, 002572, and 600567 ranked higher among the top six. In contrast, with net profit as the output variable, the efficiency of the enterprises with stock codes 002078, 002572, 000910, 600567, 600337, and 002303 were ranked higher, and only a unit difference between these two results was observed. Supplementary Table S2 presents the enterprise’s stock codes. Simultaneously, the super-efficient DEA model findings are similar to those of SFA for the enterprises with stock codes 002679, 600337, 002572, 002043, 002078, and 000910. Their rankings in the top six enterprises revealed the significance of the SFA analysis, indicating that the analyzed results were stable and robust.

Based on the study results presented in Table 5 and Table 6, Figure 2 demonstrates that the efficiency of the enterprises with operating income as the output variable performed better than those with net profit as the output variable. However, both groups experienced an increasing trend. Thus, from 2014 to 2019, the efficiency of enterprises with operating income as the output variable increased from 0.742 to 0.816. In comparison, the efficiency of enterprises with net profit as an output variable increased from 0.237 to 0.507. On average, forestry enterprises’ efficiency steadily increased. However, some meager differences existed between the comprehensive technical efficiency results and the super-efficient DEA model, as indicated in Figure 1 and the SFA trend of enterprise efficiency in Figure 2. This might be because a comprehensive error was considered in the SFA model, which could better reflect the essence and robustness of the results.

This study concludes that the comprehensive efficiency value is the product of the PTE and SE values. Furthermore, most enterprises’ resource allocation capacity and efficiency in China are reaching the optimum at their current technological level.

6. Discussion

The current research examined the efficiency measurement and the impact factors of sustainable forestry development in China by adopting the super-efficiency data envelopment analysis, Malmquist index methods for analysis, and applied the Stochastic Frontier approach for the robustness check. This study also examined the relevance of the forest products premised on static, dynamic, and scale efficiency. Based on the empirical results of this study, the authors find that the forestry industry in China is performing well and requires governmental and institutional support to achieve its core targets. Realizing the dynamic changes in the forestry industry will influence stakeholders to adapt responses that will enhance enterprises’ performance to cope with the current dynamics in the sector. The identified trend of performance and growth demonstrates that sustaining the industries will help alleviate unstandardized initiatives that decrease the enterprises’ performance. Analyzing the enterprises’ performance trajectory will underline the limitations and allow the adoption of appropriate mechanisms for future expansion. Identifying the limitations in the trajectory of performance will help future investigations. Thus, the super-efficiency DEA, Malmquist, and SFA models allowed researchers to measure the respective enterprises’ performance and quantify the variability of the enterprise performance and their geographical characteristics.

This study found that 11 companies were in the average rank of production and operation efficiencies more significant than 1, indicating that the super-efficiency DEA model was outstanding when used to measure and compare enterprises’ operation and production efficiency with efficiency values greater than 1 [96,122]. This demonstrates that enterprises with desired efficiency scores perform relatively better than those that do not reach the threshold. Therefore, enterprises that cannot attain the threshold must improve their performance. Nonetheless, the enterprises’ aggregate average comprehensive technical efficiency value reaches 0.987 (less than 1), demonstrating a likelihood of improvement. Additionally, interventional government policies should be tailored to the demands of the enterprises to help them cope with the current trajectory of development. The enterprises need to reduce ineffective asset inputs, strengthen the management of input resources, and increase input and output efficiency by reducing operating and management costs. Management should aim at transforming low-cost and inefficient into high-efficiency management methods. Efficient collaboration through economies of scale increases the operating efficiency of companies.

Furthermore, this research revealed that the forestry companies’ performance and efficiency scores in 2019 varied based on the super-efficiency DEA model findings. Based on the influence of Fujian Jinsen enterprise’s aggregate efficiency value in 2019, we re-ranked the enterprises’ scores using the super-efficiency DEA model to determine effective DMUs in 2019. Thus, scores with a total efficiency value above criterion 1 are recommended [122,124,125]. The outcome demonstrated that most high-performance companies achieved an efficiency value above criterion value one after the re-ranking [122]. Nevertheless, the SE level could not reach the threshold of the production frontier, demonstrating the high imbalance between China’s forestry technology and the SE. Therefore, this study suggests that companies adopt prudent approaches to enhance their performance.

In addition, dynamic analysis was performed by adopting the Malmquist model to determine the TFP trajectory in evaluating the industry performance. The findings from the dynamic investigation revealed that the TFP of the forestry enterprises between 2014 and 2019 showed an improved trend [128]. However, its development rate was relatively tardy, with an annual growth rate of 2.9%. The aggregate yearly average value of 1.8% reveals the dynamics of technical efficiency, serving as a strong driving force for increasing the TFP. Additionally, technological efficiency changes are linked to the combined growth rate of PTE and SE changes. This suggests that the development of China’s forestry listed companies tends to be stable, and the variation is observed to be low. The findings also demonstrate that domestic and foreign market conditions may have influenced these discrepancies. Capital investment factors must be adequately transformed into profits, flowing into business income and corporate performance.

This study employed SFA for a robustness check. This was performed to recognize whether all the DMUs were robust and aligned with the statistical findings of the super-efficiency DEA and Malmquist models [131]. The authors based their analysis on the measurement results with operating income as the output variable, as shown in Table 5, and net profit as the output variable, as shown in Table 6. The SFA findings based on operating income showed that 17 enterprises were efficient and demonstrated an average higher than 0.781. Furthermore, the SFA indicated that net profit as the output variable portrayed the efficiency of 19 enterprises, with an average higher than 0.380. This result illustrates that the efficiency of forestry enterprises constantly improves through internal management and external support, which confirms the super-efficient DEA analysis presented in Table 2. Based on the empirical findings of this study, we provide three crucial findings to assist stakeholders in decision making and address fundamental issues limiting forestry enterprises’ performance and management.

Firstly, this research found that most enterprises were affected by SE. Additionally, some enterprises could not reach effective DMUs, which may be due to domestic and foreign market conditions. Thus, the investment of capital elements did not translate into profits to flow into the enterprise’s operating income. Moreover, resources were successfully converted into the input of relevant factors, thus affecting enterprises. The production and operation of SE did not expand, and the target was not achieved. Hence, the performance trend suggests that stakeholders must implement policies that boost industries by limiting the influence of domestic and foreign market conditions.

Secondly, this study also demonstrates that not much focus has been placed on the technological progress of the forestry industry and that enterprises with low technical levels have not established technological awareness. Hence, understanding independent and technological innovation, which affects the further improvement of the efficiency of technological progress, remains insufficient. Thus, stakeholders must implement specialized policies to promote the performance of forestry enterprises.

Finally, the forestry industry has a long production cycle and investment return time. It is also affected by domestic and foreign market conditions and policies, which result in unstable profit returns that affect the flow of capital elements. Thus, these trends significantly affect enterprises’ operating efficiency. The challenges posed by foreign and domestic markets to enterprises demonstrate that stakeholders must implement effective and efficient policies to limit adverse conditions and promote efficiency.

7. Conclusions

This study investigated the performance of 32 forestry companies in China using data from the Shanghai and Shenzhen A-share stock exchanges. This research used forestry companies’ financial data from 2014 to 2019 to measure their operating efficiency and TFP. This empirical research focused on static and dynamic efficiency evaluation by adopting the super-efficiency DEA, Malmquist index, and SFA models for robustness checks. This study sought to aid the understanding of the operating performance of China’s listed forestry companies, ensuring their high-quality development, maintaining ecological balance, and promoting national economic growth. It intended to contribute to the existing knowledge and body of literature by combining the super-efficiency DEA and the Malmquist index methods to evaluate the operating performance of 32 listed forestry companies from a static and dynamic perspective in China. The main findings of this study are presented as follows:

First, the super-efficiency DEA model was used to conduct a static efficiency evaluation to determine the operating performance of forestry listed companies. Next, the Malmquist production index model was used to determine the dynamic efficiency of forestry listed companies. The static and dynamic evaluation results showed that the overall input-output efficiency of the listed forestry companies in China was relatively high. Thus, during the performance evaluation period, the average comprehensive technical efficiency of the 32 DMUs for the 6 years was 0.987. Regarding operating performance, 1.3% of the input factors of listed forestry companies did not contribute to the companies’ operating performance. Thus, the input production factors did not reach effective utilization standards without changing the technical level. The PTE for most companies is effective. However, the impact of SE efficiency affected further operating performance and improvement. From the dynamic efficiency, the average TFP index of the forestry listed companies reached 1.029, indicating an improvement in operating efficiency from 2014 to 2019. Notably, the PTE and TFP dynamics had equivalent outcomes. Both experienced similar patterns of fluctuation. The fluctuation was associated with the long-term forestry production cycle, unstable market factors, and susceptibility to policy influence.

Second, Fujian Jinsen, a leading company with the highest comprehensive efficiency value, significantly impacted the overall average efficiency of all forestry listed companies. The 2018–2019 average comprehensive efficiency results, excluding Fujian Jinsen, indicated an upward trend for the other companies. The technological standards of Fujian Jinsen concerning input resources were effective. However, the SE was insufficient. Integrated with the TFP index, its technological progress change index was less than one, indicating that new technologies are needed to increase the investment factors to focus on corporate strategy and leverage economies of scale better.

Among the 11 listed forestry companies that achieved comprehensive technical efficiency and effectiveness, the results indicated that the overall economies of scale of the industries were low.

7.1. Countermeasures and Suggestions

This study proposes corresponding countermeasures and suggestions at the macro and micro levels for stakeholders in the forestry industry:

(1) From a macro perspective, the state should improve the forestry industry development plan and policy implementation rules and give focus on forward-looking and guiding principles of industrial and technological innovation policies.

Concerning funding, China’s forestry economic development stakeholders should budget for an adequate funding source to avoid inadequacies. The funding support channels should be unidirectional and highly dependent on national budget funds. Therefore, financing channels should be expanded to promote financial support for forestry and economic development in the tertiary industry. The state should improve the loan acquisition system, simplify procedures, provide support through policy banks, provide financial subsidies, and appropriately expand the scope of funding subsidies.

Regarding taxation policies, the state should strengthen relevant forestry finance and taxation support policies to implement preferential tax bases, tax reductions, and exemption policies for all forestry listed companies. Thus, loan policies made for significant investments in forestry should have a prolonged payback time, and the collection system should be flexible.

Again, the state should strengthen scientific and technological innovation policies. The government must improve and innovate the governance policy system to guide listed forestry companies to carry out high-industrialization operations actively. The government should also encourage and support independent innovation by promoting high-tech imports and formulating cooperation between domestic and regional value chains. Furthermore, innovative power and international trade governance policies that align with the global value chain should provide a level playing field for the forestry market. In addition, China’s forestry industry remains immature; it is crucial to implement national policies and fund support projects to strengthen technological innovation. Systematic development plans and policies can help improve the quality of the forestry industry to regulate and support its high-quality development.

(2) At the micro level, this study proposes that companies need to deepen their use of emerging technologies, improve management systems, strengthen technological innovation capabilities, learn from global companies, and emulate the vital role of leading companies in the global market. Companies with high SE (Chenming Paper, Da ya Dekor, Qingshan Paper, etc.) can achieve complementary advantages through shareholding reforms and equity participation. They can jointly develop large emerging businesses with the parent company at the core of a stable community of interests.

Companies with low SEs, such as Fujian Jinsen, Sun Paper, and Kane Co., Ltd., should continue introducing advanced technology and equipment to enhance their companies. In addition, increasing scientific and technological research and development investments will strengthen the ability for independent innovation. Enterprises should optimize the input structure, reduce production and operation costs, pay maximum attention to existing resources, and maximize output by maintaining the best input.

Key core technologies and scale advantages are the cornerstones of companies’ economic and social progress in the global value chain. Taking Fujian Jinsen as an example, its complete development strategy and management level supported the growth of domestic forestry industries and served as an industrial supply chain leader. Enterprises that choose varied approaches for their operations should use existing resources to their full potential, control production costs, improve investment efficiency, and minimize inputs to obtain the maximum output. In addition, achieving the corresponding management status will invariably create effective resource utilization efficiency and capacity. Within the company’s complete-scale expansion and technological innovation, the improvement in the company’s operating efficiency will be minimal. Therefore, companies must analyze their conditions and the external environment and balance their interests to make scientific decisions.

7.2. Limitations of the Study and Future Directions

This study examined the performance of 32 forestry industries in China using data from the Shanghai and Shenzhen stock markets between 2014 and 2019. Owing to data unavailability for the forestry industry in China, short-term data characterized this research; therefore, the authors will examine this in the future using available data for analysis. This study only focused on China’s forestry industry; future research will compare China and other industries to ascertain dynamic findings.