Urban agglomeration-based green logistics efficiency evaluation analysis model construction method

Abstract

The invention provides an urban agglomeration-based green logistics efficiency evaluation analysis model construction method. The method comprises the steps of 1, constructing an urban agglomeration green logistics efficiency evaluation index framework; constructing an urban agglomeration green logistics environment variable index system, wherein the environment variable index system comprises a per capita total production value urbanization rate and a proportion of logistics fixed asset investment to regional fixed asset investment; and constructing a three-stage analysis model of a super-efficiency SBM-SFA-Malmquist index model according to the urban agglomeration green logistics efficiency evaluation index framework and the urban agglomeration green logistics environment variable index system. According to the method, the index system for systematically analyzing the urban agglomeration green logistics efficiency is constructed, the model method applied to urban agglomeration green logistics efficiency evaluation is constructed, and then the rationality of the evaluation index system and the model method is verified through urban agglomeration green logistics efficiency evaluation cases of the Yangtze River midstream. And factors restricting green logistics development are concluded.

Description

Technical Field

[0001] The present invention belongs to the technical field of green logistics, and in particular to a method for constructing a green logistics efficiency evaluation and analysis model based on an urban agglomeration. Background Art

[0002] Urban agglomerations are the symbol and product of urbanization that has reached a certain level and stage. Urban agglomeration-based areas, represented by urban agglomeration organizations, have become a huge space for gathering domestic and international economic and social elements.

[0003] The high-quality development of green logistics is an important driving force for the economic transformation of urban agglomerations. Green logistics promotes the spatial distribution of production factors and creates time benefits, thereby promoting the division of labor and cooperation among cities in urban agglomerations, making up for the lack of resource endowments of urban agglomerations, broadening the radiation range of central cities, and driving the economic development of surrounding cities; the development of green logistics can form a new industrial form, optimize the economic structure of urban agglomerations, promote the formation of innovative industries, enhance the competitiveness of pillar industries, reduce the economic operation costs of urban agglomerations, and improve economic efficiency; green logistics has a very large correlation effect on upstream and downstream related industries in the supply chain. Through the role of correlation effects, it can vigorously promote the development of other industries between regions; the core idea of ​​green logistics is environmental protection and sustainable development. While the development of green logistics produces huge economic benefits, it will also bring social benefits that cannot be ignored. Existing research on green logistics focuses on empirical interpretation at the theoretical level, mainly qualitative analysis, less quantitative analysis, and more subjective factors in evaluation. Some quantitative analysis is combined with subjective methods, which has the defect of strong uncertainty. The evaluation method in quantitative analysis is mainly based on traditional DEA and its improvement. Its basic idea is to make a homogeneous assumption on the environment in which the green logistics industry is located. The efficiency measurement method does not consider the influence of slack variables, environmental factors, management factors, and random interference factors. The development of green logistics is closely related to external factors such as the external economic environment and government execution management, but it is difficult for the traditional DEA method and its improvement to eliminate the above influences and reflect the actual level of green logistics development under heterogeneous conditions. At the same time, the accuracy of the traditional DEA method model measurement is difficult to solve the setting of slack variables and the distortion of the efficiency frontier, and has the defect of no feasible solution under variable scale returns. At present, the actual level of green logistics cannot be measured and evaluated objectively and truly based on qualitative analysis and homogenized efficiency evaluation methods.

[0004] After literature research and summary, in recent years, the literature with "urban agglomeration" and "logistics" as keywords tends to focus on the design and optimization of urban agglomeration logistics networks, the analysis of the combination of logistics industry and other industries, and focuses on the cross-influence between industries. There are few studies on the measurement of logistics efficiency. Papers studying the logistics efficiency of urban agglomerations pay more attention to the external space operation and connection between urban agglomerations, and rarely conduct detailed quantitative measurement and objective evaluation of the internal operation efficiency of urban agglomerations.

[0005] The current theory of the internal operation of green logistics in urban agglomerations is mainly based on the guidance of central or local policies. Generally, provincial capital cities are used as central cities to unite with surrounding cities to build an integrated economy as the guiding ideology for urban integrated development, and the role and strength of other cities are not fully utilized. As a result, some urban agglomerations in China do not have the basic characteristics of urban agglomerations. According to the traditional urban agglomeration classification standards, it is impossible to objectively and accurately position the development of the green logistics industry, and it is impossible to fully explain the current status of the internal operation of green logistics in urban agglomerations. Summary of the invention

[0006] The purpose of this invention is to propose a method for constructing a green logistics efficiency evaluation and analysis model based on urban agglomerations. A three-stage analysis method combining the super-efficiency SBM method, the SFA regression method and the Malmquist index method is used to conduct a systematic analysis of green logistics efficiency. The Cobb-Douglas production function model is applied when constructing the index system. Finally, the Person test and the LR test are performed before the model is used to determine the compatibility of the method theory with the green logistics case data of urban agglomerations.

[0007] In order to achieve the above object, the present invention provides a method for constructing a green logistics efficiency evaluation and analysis model based on an urban agglomeration, the method comprising:

[0008] Step 1: construct an urban agglomeration green logistics efficiency evaluation index framework; the urban agglomeration green logistics efficiency evaluation index framework includes input indicators and output indicators, and the output indicators include expected output indicators and non-expected output indicators;

[0009] The input indicators include the fixed asset investment in the logistics industry, the total mileage of highways, and the number of people employed in the logistics industry;

[0010] The expected output indicators include total energy consumption, total logistics industry output value and freight volume;

[0011] The undesirable output indicators include carbon dioxide emissions and carbon emission intensity of the logistics industry;

[0012] Step 2: Construct a green logistics environmental variable indicator system for urban agglomerations, including the urbanization rate of per capita regional GDP and the proportion of logistics fixed asset investment in regional fixed asset investment;

[0013] Step 3: Construct a three-stage analysis model of the super-efficiency SBM-SFA-Malmquist index model based on the urban agglomeration green logistics efficiency evaluation index framework and the urban agglomeration green logistics environmental variable index system;

[0014] The three-stage analysis model of the super efficiency SBM-SFA-Malmquist index model is constructed as follows:

[0015] In the first stage, a super-efficiency SBM model based on non-expected output is designed under the scenario of variable returns to scale. The input-output data of green logistics efficiency of each city in the urban agglomeration are evaluated and calculated to obtain the green logistics efficiency value and slack variable value. The green logistics efficiency value is the static efficiency value, and the slack variable value is the redundant term. At the same time, a Malmquist index model is constructed to calculate the original input and output, and the dynamic efficiency value based on the original input and output data is calculated;

[0016] In the second stage, the slack values ​​of the input variables of the super-efficiency SBM model results were combined with environmental variables to conduct SFA regression analysis to eliminate environmental factors;

[0017] In the third stage, the green logistics input data of urban agglomerations that have undergone SFA regression analysis are used to re-analyze the static and dynamic operational efficiency of green logistics in urban agglomerations using the super-efficiency SBM model with re-unexpected output and the Malmquist index model.

[0018] Furthermore, the environmental variables include environmental effects, management inefficiency and random interference terms.

[0019] Furthermore, the steps for constructing the green logistics efficiency evaluation index framework of the urban agglomeration are as follows:

[0020] The framework of the input-output evaluation index system is constructed based on the Cobb-Douglas production function, which is expressed as follows:

[0021] Y=A t k α L β μ

[0022] Among them, Y is the output, At is the comprehensive technical level, K is the invested capital, L is the invested labor, μ is the random interference; α and β represent the output elasticity of L and K respectively;

[0023] Combined with the selected non-desirable output indicators, the outputs with environmental impacts are incorporated into the output indicators.

[0024] Furthermore, the fixed asset investment amount of the logistics industry represents the capital input of the logistics industry; the total mileage of roads represents the infrastructure input of the logistics industry; the number of employees in the logistics industry represents the labor input of the logistics industry; the total energy consumption represents the material input of the logistics industry; the gross domestic product of the logistics industry represents the economic contribution level of the logistics industry; the freight volume represents the service output of the logistics industry; the carbon dioxide emissions of the logistics industry represent the carbon emission status of the logistics industry; the carbon emission intensity of the logistics industry represents the carbon dioxide emissions generated per unit GDP of the logistics industry, wherein carbon emission intensity = total carbon dioxide emissions of the logistics industry / gross domestic product of the logistics industry.

[0025] Furthermore, the environmental influencing factors include management inefficiency terms and random error terms.

[0026] Furthermore, the super-efficiency SBM model based on undesired output is expressed as follows:

[0027]

[0028] Among them, ρ * is the target efficiency value, ρ * ≥1 indicates that the efficiency is effective, and the maximum representative decision-making unit is the most efficient. Otherwise, ρ * <1 indicates that there is still room for improvement in efficiency; r is the rth expected output; j is the jth decision-making unit; n is the number of decision-making units, q is the number of decision-making unit inputs; u1 is the expected output; u2 is the undesired output; x i , are the elements in the corresponding input matrix, expected output matrix, and unexpected output matrix respectively; i is the i-th weight vector; are the variable reference points for eliminating the input matrix, expected output matrix, and unexpected output matrix of the kth decision-making unit; i is the i-th input of the decision-making unit; x ik is the input matrix of the kth decision-making unit; is the expected output matrix of the kth decision-making unit; is the undesired output matrix of the kth decision making unit.

[0029] Furthermore, the Malmquist productivity index model is used to measure the changes in global total factor productivity (TFP) before and after, establish a decision-making unit to evaluate dynamic efficiency, and construct a dynamic input-output distance function as follows:

[0030] TFP=TPR×TEC=TPR×PTE×SE

[0031] Among them, TPR represents the rate of technological progress, TEC represents the technical efficiency change index, PTE represents pure technical efficiency, and SE represents scale efficiency. Since technical efficiency is equal to pure technical efficiency multiplied by scale efficiency, the total factor productivity TFP is decomposed into:

[0032]

[0033] Among them, P c represents total factor productivity, t and t+1 represent the immediate and forward periods, N represents the scale index, M represents the pure technology index, a and b represent the input and output data respectively, They represent the dynamic changes of technical efficiency in the immediate and long term, according to the dynamic changes across time periods; It represents the scale efficiency;

[0034] represents pure technical efficiency; Indicates the technical efficiency change index

[0035] According to the Malmquist productivity index model, when p>1, it means that the total factor productivity is gradually increasing, and vice versa.

[0036] Furthermore, the second stage specifically includes:

[0037] The slack value of the input variable of the first-stage super-efficiency SBM model result is taken as the explained variable Y, and the environmental variable is taken as the explanatory variable X to construct the stochastic frontier model, which is expressed as follows:

[0038] s ni,t =f n (z i,t ;y n )+v ni,t +ε ni,t

[0039] where s ni,t represents the total value of the slack variable invested by the ith decision-making unit in year n, z i,t represents the environmental variable of the ith decision-making unit in year t, y n represents the unknown parameter to be estimated, y n represents the feasible stochastic frontier of the nth input, measuring the impact of environmental factors on the slack variable, v ni,t is the management inefficiency term, ε ni,t is a random error term.

[0040] Calculate the alpha value:

[0041]

[0042] Among them, when When , the inefficiency term has the greatest impact on each slack variable;

[0043] Calculate v ni and ε n The value, excluding the influence of external environment and random interference and adjusting the input-output variables, is expressed as follows:

[0044]

[0045] Calculate the adjusted input variable value It is expressed as follows:

[0046]

[0047] Among them, xni,t and Respectively represent the input amount before and after adjustment, and They represent the optimal external environmental factors and the decision-making unit with the least random interference, and They respectively represent the elimination of the influence of external environment and random interference.

[0048] Furthermore, the SFA regression analysis needs to verify the significance of the data, so a T test is performed on the data to determine whether the factors affecting efficiency screened out have an impact and interference on the explained variables at least at the level of 10%, that is, the α value is less than 0.1, which is used to measure the relative importance of non-efficiency in the total error. When α is close to 1, it means that the non-efficiency items have a major influence on the total error, which can support the applicability of the SFA model in this scenario; the test value of the likelihood ratio test LR needs to be greater than the critical value of the chi-square distribution with a degree of freedom of 1 at a significance level of 1‰.

[0049] Furthermore, in the third stage, the processed data are re-input into the super-efficiency SBM model based on non-expected output to obtain the adjusted input variables, and at the same time re-input into the Malmquist productivity index model to recalculate the adjusted global total factor productivity.

[0050] The beneficial technical effects of the present invention are at least as follows:

[0051] (1) The present invention selects the super-efficiency SBM-SFA-Malmquist index model three-stage analysis method from the two perspectives of static and dynamic analysis, considers the influence of slack variables, uses the SFA method to eliminate external macro factors, management factors and random interference factors, avoids subjective weighting, objectively and comprehensively analyzes green logistics efficiency, evaluates its development level, comprehensively constructs its evaluation index system, and conducts in-depth research on the factors affecting the logistics efficiency of urban agglomerations. It can be seen that the present invention is more logical and comprehensive in its research on green logistics efficiency, enhances the explanatory power of the results, and makes a certain contribution to the expansion of the green logistics efficiency research method system.

[0052] (2) As the high-quality development of green logistics is an important driving force for social and economic transformation, and urban agglomerations are important platforms for building international competitiveness. It is necessary to measure and analyze the level of green logistics development in urban agglomerations. At present, domestic scholars have conducted little research on the evaluation and analysis of green logistics efficiency. The research on green logistics efficiency with urban agglomerations as the research object focuses more on the external industrial linkages of urban agglomerations and the comparison between urban agglomerations, but neglects to explore the internal operating efficiency of urban agglomerations. This invention takes urban agglomerations as spatial carriers, takes green logistics efficiency as the evaluation object, focuses on the study of the internal operating efficiency of urban agglomerations, and enriches the research content of green efficiency measurement of urban agglomerations.

[0053] (3) Currently, domestic and foreign scholars have conducted little research on green logistics in the middle reaches of the Yangtze River. As one of my country's important economic zones, an important part of the national market and a spatial hub, the optimization of its green logistics system structure and the improvement of logistics efficiency are the key to the high-quality development, optimization and upgrading of my country's logistics industry. Therefore, this paper uses the green logistics efficiency of the middle reaches of the Yangtze River city agglomeration as a case study, and also enriches the theoretical research on the green logistics system of the middle reaches of the Yangtze River city agglomeration. The research conclusions can provide a theoretical basis for the decision-making of the regional government of the middle reaches of the Yangtze River city agglomeration, and also point out the direction for scholars to explore and analyze the measurement and influencing factors of green efficiency in logistics or other industries. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] The present invention is further described using the accompanying drawings, but the embodiments in the accompanying drawings do not constitute any limitation to the present invention. A person skilled in the art can obtain other drawings based on the following drawings without creative work.

[0055] Figure 1 This is a flow chart of the method for constructing a green logistics efficiency evaluation and analysis model based on urban agglomerations in the present invention.

[0056] Figure 2 This is a distribution diagram of the mean range of comprehensive technical efficiency of the Yangtze River Middle Reaches urban agglomeration in the third stage of an embodiment of the present invention.

[0057] Figure 3 This is a distribution diagram of the mean range of pure technical efficiency of the middle reaches of the Yangtze River city cluster in the third stage of an embodiment of the present invention.

[0058] Figure 4 This is a distribution diagram of the mean range of scale efficiency of the Yangtze River Middle Reaches urban agglomeration in the third stage of an embodiment of the present invention. DETAILED DESCRIPTION

[0059] Embodiments of the present invention are described in detail below, examples of which are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present invention, and cannot be understood as limiting the present invention.

[0060] The present invention understands the green logistics efficiency of urban agglomerations as follows: considering the comprehensive logistics efficiency of energy consumption and environmental costs within the urban agglomeration, taking natural resource consumption as an input variable, and incorporating pollutant emissions generated with value output as non-desired output variables into the logistics efficiency. The efficiency thus obtained is the green logistics efficiency of the urban agglomeration. The improvement goal is to minimize resource input and pollutant emissions and maximize economic output. The present invention will adopt the three-stage analysis method of the super-efficiency SBM-SFA-Malmquist model, and the same group of data can be used for the input indicators of the first and third stages. However, the second-stage SFA regression model needs to consider the influence of external environmental factors, so it is necessary to select external environmental factors that affect the level of green logistics development in urban agglomerations.

[0061] Embodiment 1

[0062] like Figure 1 As shown, the method provided by the embodiment of the method for constructing a green logistics efficiency evaluation and analysis model based on an urban agglomeration includes:

[0063] Step 1: As shown in Table 1, construct an urban agglomeration green logistics efficiency evaluation index framework; the urban agglomeration green logistics efficiency evaluation index framework includes input indicators and output indicators, and the output indicators include expected output indicators and non-expected output indicators;

[0064] The input indicators include the fixed asset investment in the logistics industry, the total mileage of highways, and the number of people employed in the logistics industry;

[0065] The expected output indicators include total energy consumption, total logistics industry output value and freight volume;

[0066] The non-desirable output indicators include carbon dioxide emissions from the logistics industry and the carbon emission intensity of the logistics industry.

[0067] Table 1

[0068]

[0069] Constructing an index system is a common method for quantitative research, which is generally divided into two steps: the first step is to build an index system framework based on the connotation and characteristics of the evaluation object according to relevant theories, and select targeted, representative and comprehensive indicators for research. The second step is to analyze and compare each indicator through literature review, and select appropriate indicators based on the actual situation of the research object. The present invention also follows the above steps when constructing the input-output index system of green logistics in urban agglomerations.

[0070] The first step is to build an input-output evaluation index system framework based on the Cobb-Douglas production function. The structural framework is as follows: Figure 1As shown. The Cobb-Douglas production function is a production function created by American mathematician CW Cobb and economist Paul H. Douglas when they jointly explored the relationship between input and output. This production function can better reflect the quantitative relationship between input and output under given production technology conditions. Therefore, the present invention uses the theoretical framework of the production function to build an urban agglomeration green logistics input-output evaluation index system framework.

[0071] Y=A t K α L β μ

[0072] Y is output, At is the comprehensive technology level, K is the capital input, L is the labor input, μ is the random interference; α and β represent the output elasticity of L and K respectively;

[0073] In the second step, based on production theory and combined with the characteristics of the selected non-expected output super-efficiency SBM model, the outputs that have an impact on the environment, such as carbon emissions, are incorporated into the output indicators.

[0074] Step 2: Construct a green logistics environmental variable indicator system for urban agglomerations, which includes the urbanization rate of per capita regional GDP and the proportion of logistics fixed asset investment in regional fixed asset investment.

[0075] The analysis of the green logistics efficiency of urban agglomerations is not only affected by internal factors, but also by external environmental factors. At the same time, since the above has confirmed that the SFA model is used in the second stage to eliminate influencing factors and perform regression analysis, it is necessary to consider the impact of external environmental factors when evaluating the green logistics efficiency of urban agglomerations. Considering that the green development of the logistics industry is affected by factors such as the local economic development level, government support, and population urbanization, combined with the availability of data, this paper finally selects the following three environmental variables:

[0076] ① Economic development level: The local economic level affects the development of the local logistics industry to a certain extent, and the economic strength of the population also affects the implementation of green and high-quality development of the logistics industry. Therefore, the present invention selects the per capita regional GDP to represent the economic development level of the urban agglomeration, and considers the impact of economic factors on the green development efficiency of the logistics industry.

[0077] ② Government support: The government's support for the green and high-quality development of the logistics industry is directly related to whether the logistics industry has more resources and development opportunities. The proportion of fixed asset investment in the logistics industry to regional fixed asset investment can, to a certain extent, represent the government's emphasis on the green and high-quality development of the logistics industry.

[0078] ③ Urbanization level: The logistics industry is labor-intensive, and its green development is inseparable from the comprehensive quality and living habits of urban residents. The higher the urbanization, the higher the production and consumption activity, the greater the energy consumption of the logistics industry, and the greater the carbon emissions, which may restrict the development quality of green logistics. This paper selects the urbanization rate as an external environmental variable that affects the efficiency of green logistics in urban agglomerations.

[0079] In summary, the construction of the green logistics efficiency environmental variable index system for urban agglomerations is shown in Table 2:

[0080] Table 2

[0081]

[0082] Step 3: Construct a three-stage analysis model of the super-efficiency SBM-SFA-Malmquist index model based on the urban agglomeration green logistics efficiency evaluation index framework and the urban agglomeration green logistics environmental variable index system;

[0083] Based on the requirements of the above-mentioned evaluation indicators and the characteristics that should be noted when evaluating the green logistics efficiency of urban agglomerations, the present invention proposes a three-stage analysis method that combines the super-efficiency SBM model of non-expected output, the stochastic frontier efficiency analysis method SFA and the Malmquist index model to measure the green logistics efficiency of urban agglomerations under the condition of variable returns to scale (VRS). The above model was selected based on the following reasons: (1) The decision-making unit of the urban agglomeration is the individual city, each of which is an independent geographical area and node. There is a high probability of connection between nodes. After the urban agglomeration is selected, the number of nodes is known, which is very consistent with the application scenario of the SBM model efficiency measurement; (2) After considering the influence of slack variables, the super-efficiency SBM model can further distinguish the decision-making units that are also effective in DEA, which is conducive to the ranking problem of high-efficiency cities; (3) The super-efficiency SBM model based on non-expected output fully considers the inclusion of non-expected output and is an industrial efficiency evaluation model that incorporates the concept of green and sustainable development; (4) Combined with the SFA model, it can eliminate the influence of environmental factors and random interference factors from the dimension of slack variables; (5) The application of the Malmquist index model makes up for the lack of dynamic analysis in the SBM model. This three-stage analysis method does not require any artificial weighting when measuring efficiency, making the quantitative analysis more objective. At the same time, it considers the influence of slack variables and the interference of external macro factors, management factors and random factors, and combines static analysis and dynamic analysis to comprehensively, objectively, scientifically and systematically evaluate the efficiency of green logistics.

[0084] Specifically, the present invention uses a three-stage analysis method that combines the super-efficiency SBM model of undesired output, the stochastic frontier efficiency analysis method SFA and the Malmquist index model to measure the green logistics efficiency of urban agglomerations. The specific steps of the three-stage analysis method are as follows: Figure 1 shown.

[0085] The first stage of the analysis method is divided into two steps. First, under the premise of selecting variable return to scale (VRS), the super-efficiency SBM model of non-expected output is used to evaluate and calculate the input-output data of green logistics efficiency of each city in the urban agglomeration, and the green logistics efficiency value and slack variable value are obtained; second, the Malmquist index model is used to calculate the original input and output to obtain a dynamic index value.

[0086] Phase 1: Super efficiency SBM-Malmquist index model to evaluate initial efficiency:

[0087] (1) Construction of super-efficiency SBM model based on unexpected output

[0088] When evaluating the super-efficiency SBM model, the DMU (decision-making unit) must first be evaluated by SBM, and then the super-efficiency SBM can be used for evaluation if the DMU is effective. In the previous discussion, it has been demonstrated that the decision-making unit of the urban agglomeration is the characteristic of individual cities, which is very consistent with the application scenario of the SBM model efficiency measurement. Under the premise of variable returns to scale (VRS), the super-efficiency SBM model of non-expected output is selected as follows:

[0089]

[0090] Among them, ρ * is the target efficiency value, ρ * ≥1 indicates that the efficiency is effective, and the maximum representative decision-making unit is the most efficient. Otherwise, ρ * <1 indicates that there is still room for improvement in efficiency; r is the rth expected output; j is the jth decision-making unit; n is the number of decision-making units, q is the number of decision-making unit inputs; u1 is the expected output; u2 is the undesired output; x i , are the elements in the corresponding input matrix, expected output matrix, and unexpected output matrix respectively; i is the i-th weight vector; are the variable reference points for eliminating the input matrix, expected output matrix, and unexpected output matrix of the kth decision-making unit; i is the i-th input of the decision-making unit; x ik is the input matrix of the kth decision-making unit; is the expected output matrix of the kth decision-making unit; is the undesired output matrix of the kth decision making unit.

[0091] (2) Construction of Malmquist productivity index model

[0092] In order to determine the dynamic efficiency value based on the original input and output data, in the first stage, the decision-making unit was established to evaluate the dynamic efficiency by calculating the changes of global total factor productivity (TFP) based on the Malmquist productivity index. The dynamic changes of the decision-making unit over time were further analyzed, and the distance function of dynamic input and output was constructed as follows:

[0093] TFP=TPR×TEC=TPR×PTE×SE

[0094] Among them, TPR represents the rate of technological progress, TEC represents the technical efficiency change index, PTE represents pure technical efficiency, and SE represents scale efficiency. Since technical efficiency is equal to pure technical efficiency multiplied by scale efficiency, total factor productivity (TFP) can be decomposed into:

[0095]

[0096] Where P c It refers to total factor productivity, t and t+1 represent the current and forward periods, N represents the scale index, M represents the pure technology index, a and b represent the input and output data respectively, that is, c The mathematical formula can be decomposed into the product of scale efficiency, pure technology index and technology efficiency change index, where It represents the scale efficiency; represents pure technical efficiency; Represents the technical efficiency change index.

[0097] According to the Malmquist productivity index model, when p>1, it means that the total factor productivity is gradually increasing, and vice versa.

[0098] Phase 2: SFA method eliminates environmental impact:

[0099] Under the super-efficiency SBM model conditions mentioned in the first stage of the present invention, the slack variable values ​​of various efficiency values ​​and various inputs may be affected by the random interference of environmental factors, thereby causing errors in the calculated green logistics efficiency values. The second stage is to use the slack values ​​of the input variables of the super-efficiency SBM model results of the first stage as the explained variable Y, and the environmental variables as the explanatory variables X, and perform stochastic frontier analysis (StochasticFrontierAnalysis), that is, SFA regression analysis. After SFA regression processing, the input data is adjusted for the purpose of eliminating influencing factors such as environmental effects, management inefficiency, and random interference.

[0100] The specific steps of the second-stage stochastic frontier model SFA method are divided into the following five steps:

[0101] (1) The first step is to build a stochastic frontier model:

[0102] s ni，t =f n (z i，t ; γ n )+v ni，t +ε ni，t

[0103] where s ni,t represents the total value of the slack variable invested by the ith decision-making unit in year n, z i,t represents the environmental variable of the ith decision-making unit in year t, yn represents the unknown parameter to be estimated, fn represents the feasible stochastic frontier of the nth input, and measures the impact of environmental factors on the slack variable, v ni,t is the management inefficiency term, ε ni,t is the random error term;

[0104] SFA needs to verify the significance of the data, so a T test is performed on the data to determine whether the factors affecting efficiency screened out have an impact and interference on the explained variables at least at the level of 10%, that is, the α value is less than 0.1.

[0105] (2) Therefore, it is necessary to calculate the α value:

[0106]

[0107] Among them, when When the inefficiency term v ni,t It has the greatest impact on each slack variable;

[0108] (3) Gamma and LR value test of SFA result data: In the SFA result data, the variable rate gamma is a key parameter used to measure the relative importance of inefficiency in the total error. When gamma is close to 1, it means that the inefficiency item has a major influence on the total error, which can support the applicability of the SFA model in this scenario; the test value of the likelihood ratio test LR must be greater than the critical value of the chi-square distribution with a degree of freedom of 1 at a significance level of 1‰. After the second and third steps of the T test, gamma verification, and LR test, the result data of the SFA regression analysis are meaningful.

[0109] (4) Calculate v ni and ε n The value, excluding the influence of external environment and random interference and adjusting the input-output variables, is expressed as follows:

[0110]

[0111] (5) Calculate the adjusted input variable value It is expressed as follows:

[0112]

[0113] Among them, x ni,t and Respectively represent the input amount before and after adjustment, and They represent the optimal external environmental factors and the decision-making unit with the least random interference, and They respectively represent the elimination of the influence of external environment and random interference.

[0114] Through the above calculation process, the main purpose of the second stage is to adjust the original input variables in order to eliminate external environment and random interference factors and obtain the adjusted input variable values.

[0115] Phase III: Efficiency evaluation after input variable adjustment:

[0116] The third stage is to use the adjusted green logistics input data of the urban agglomeration, and re-analyze the static and dynamic operational efficiency of the green logistics of the urban agglomeration using the super-efficiency SBM model of non-expected output and the Malmquist index model. The conclusions obtained in the third stage are the results obtained after excluding the influence of environmental variables and random errors. The analysis is more objective and can better reflect the actual situation of the development level of green logistics in the urban agglomeration.

[0117] Specifically, the third stage model input parameters are expressed as follows:

[0118]

[0119] Where ρ is the target efficiency value, ρ≥1 indicates that the efficiency is effective, and the maximum represents the most efficient decision-making unit. On the contrary, ρ<1 indicates that there is still room for improvement in efficiency; r is the rth expected output; j is the jth decision-making unit; n is the number of decision-making units, q is the number of decision-making unit inputs; u1 is the expected output; u2 is the undesired output; Z i , are the elements in the corresponding adjusted input matrix, expected output matrix, and non-expected output matrix respectively; i is the i-th weight vector; Z, are the variable reference points for eliminating the input matrix, expected output matrix, and non-expected output matrix of the kth decision-making unit; i is the i-th input of the decision-making unit; Z ik is the input matrix of the kth decision-making unit; is the expected output matrix of the kth decision-making unit; is the undesired output matrix of the kth decision making unit.

[0120] After investment:

[0121] The input value a becomes Z:

[0122]

[0123]

[0124] Where P c It refers to total factor productivity, t and t+1 represent the immediate and forward periods, N represents the scale index, M represents the pure technology index, Z and b represent the adjusted input data and the original output data, that is, Z t , Z t+1 are the current and forward input data, respectively, t 、b t+1 are the output data for the immediate period and the future period respectively. c The mathematical formula can be decomposed into the product of scale efficiency, pure technology index and technology efficiency change index, where represents the current scale efficiency index, represents the forward scale efficiency index, represents the spot pure technical index, Represents a forward pure technical index. An expression representing the adjusted scale efficiency index; The expression representing the adjusted pure technical efficiency index; Represents the adjusted technical efficiency change index.

[0125] The original input variable is represented by X, and the adjusted input variable is represented by Z.

[0126] Embodiment 2

[0127] The invention embodiment uses Dearun3.0 software to calculate and analyze the Malmquist total factor productivity model under the premise of variable returns to scale (VRS). The invention mainly analyzes the dynamic changes of the original data in the first stage after being processed by the Malmquist index model from the overall mean of the middle reaches of the Yangtze River city cluster to the individual cities.

[0128] From the perspective of time stage, the results of the Malmquist index model are shown in Table 1 for the overall urban agglomeration in the middle reaches of the Yangtze River.

[0129] Table 1 Malmquist productivity index and its decomposition mean for the Yangtze River Middle Reaches urban agglomeration during the first period 2012-2021

[0130]

[0131]

[0132] Note: The results are obtained by the author from the first stage Malmquist index model result data.

[0133] The average total factor productivity of the cities in the Yangtze River Middle Reaches Urban Agglomeration from 2012 to 2021 was 1.106, indicating that overall production efficiency has improved over the past decade and the overall situation is in a steady and positive state. The average value from 2020 to 2021 was 1.390, the highest among them, indicating that the productivity growth in 2021 was the highest in the past decade. Among them, the change in total factor productivity depends on the change in technical efficiency and the change in the efficiency of technological progress.

[0134] Technical efficiency declined in the two time periods of 2012-2013 and 2020-2021, and increased in the rest of the years, indicating that the technical management level was high during the past decade, and the efficiency of using resources in the production process has steadily improved. The main reason for the low technical efficiency in 2020-2021 is that the pure technical efficiency has declined by 5.3%, and the scale efficiency has declined slightly. From 2012 to 2013, both the pure technical efficiency and the scale efficiency have declined to varying degrees, among which the pure technical efficiency has declined by 7.0% and the scale efficiency has declined by 7.5%. This further shows that the overall pure technical efficiency and scale efficiency of the urban agglomeration in the middle reaches of the Yangtze River need to be improved, which is also one of the key factors to drive the improvement of total factor productivity. Therefore, the urban agglomeration in the middle reaches of the Yangtze River should continue to maintain the scale of the logistics industry, maintain the steady improvement of technical efficiency and technological progress index, enable the green development and transformation of the logistics industry, and increase the investment in scientific and technological resources that are conducive to energy conservation and emission reduction and the development of low-carbon green logistics to enhance its technological innovation momentum.

[0135] During this decade, the technology progress index was less than 1 only in the two time periods of 2018-2019 and 2019-2020. In particular, in 2019-2020, the technology progress index was only 0.64.

[0136] Furthermore, as shown in Table 2, from the perspective of the cities in the middle reaches of the Yangtze River urban agglomeration after adjusting the input, the total factor productivity change index of eight cities, including Yingtan, Xiantao, Huangshi, Pingxiang, Jiujiang, Xiangyang, Hengyang and Loudi, ranked at the bottom during the period 2012-2021, and their production efficiency was less than 1, indicating that there were extremely low years in the ten-year period or that the production efficiency was declining; the ten-year average values ​​of other cities basically exceeded 1, indicating that production efficiency was increasing.

[0137] Table 2 Malmquist productivity index and its decomposition of cities in the Yangtze River Middle Reaches Urban Agglomeration in the first phase

[0138]

[0139]

[0140] Note: The results are obtained by the author from the first stage Malmquist index model result data.

[0141] From the perspective of technical efficiency, the technical efficiency change index of Huangshi, Xiantao, Xiangyang and Pingxiang is lower than 1. The technical efficiency of these cities is declining, and they need to improve their technical level by effectively allocating various resources. The main reason why Huangshi's comprehensive technical efficiency change index is lower than 1 is that both pure technical efficiency and scale efficiency have declined by 3%; Xiantao's pure technical efficiency has declined by 1.5%, and scale efficiency has declined by 4.2%; Xiangyang's scale efficiency has declined by 2.1%, and Pingxiang's technical efficiency has declined by 4.4%. The technical efficiency change index of other cities is higher than 1, indicating that the technical efficiency of these cities has been increasing during this decade, and the resource allocation is reasonable, so the technical level has increased.

[0142] From the perspective of the technology progress rate index, the index of all cities is higher than 1. This shows that in the process of developing green logistics, the cities in the middle reaches of the Yangtze River urban agglomeration have a certain awareness of technological innovation, and each city also attaches great importance to it. The overall logistics industry is developing and transforming towards green development.

[0143] Furthermore, the Malmquist productivity index of each city in the Yangtze River Middle Reaches Urban Agglomeration from 2012 to 2021 before adjustment of input is shown in Table 3, which shows the changes in total factor productivity of each city year by year.

[0144] Table 3 Malmquist productivity index of cities in the Yangtze River Middle Reaches Urban Agglomeration in the first phase 2012-2021

[0145]

[0146]

[0147] Note: The results are obtained by the author from the first stage Malmquist index model result data.

[0148] Since 2012 is the base year, Table 4-14 starts from 2013. This table shows the ranking of total factor productivity for each year from 2013 to 2021, as well as the ranking of 31 cities. Before adjusting the input, the total factor productivity index of the Yangtze River Middle Reaches Urban Agglomeration in 2021 was the highest, reaching 1.39, indicating that the development and progress of the whole year were more prominent than those in previous years; 2017 and 2018 developed better than the previous two years, stagnated in 2019 and 2020, but the green logistics production efficiency in 2021 developed rapidly. In 2021, only Xiangtan, Jingdezhen and Yingtan saw a decline in production efficiency, while the rest of the cities saw an increase.

[0149] From the perspective of individual cities, there are differences in the development of cities in different years. Xiangtan has the highest average total factor productivity level, thanks to the rapid growth of productivity in 2017, because in 2017, Xiangtan's logistics industry labor input was reduced, but the expected output remained constant, the unexpected output decreased, and the overall efficiency increased.

[0150] Furthermore, the second stage model verification and result analysis:

[0151] The second stage SFA regression analysis was performed using the software frontier4.1 to obtain the result data, as shown in Table 4 below.

[0152] Table 4 Regression results of the second-stage SFA model

[0153]

[0154] First, the T test of the parameters can determine that the factors that affect efficiency are at least 10% significant at the level of influencing and interfering with the explained variables, indicating that the selection logic of environmental factors is feasible; second, the gamma values ​​are all higher than 0.75, indicating that the model has a high degree of fit; third, the test values ​​of the likelihood ratio test LR are all greater than the critical value of the chi-square distribution with a degree of freedom of 1 at a significant level of 1‰, indicating that there are management and technical efficiency losses, that is, the setting of the random frontier model is effective. The comprehensive detection of the three indicators shows that the SFA model analysis is necessary, and also shows that the model selection of the urban agglomeration logistics efficiency in Chapter 3 is in line with the actual law.

[0155] As can be seen from Table 4, the impact of external environmental factors on the slack value of input variables mainly lies in the value and positive or negative of their regression coefficients. If the regression coefficient is positive, it means that the increase in environmental variables will cause an increase in redundancy, thereby affecting the efficiency of green logistics management; if the regression coefficient is negative, it means that the increase in environmental variables will reduce redundancy, which will improve the efficiency of green logistics management. Table 5 describes the regression results of the three environmental variables on each input slack variable from 2012 to 2021.

[0156] Among them, per capita regional GDP is negatively correlated with fixed asset investment and total energy consumption in the logistics industry, and is positively correlated with relaxation of investment in logistics industry employees and relaxation of investment in transportation route network mileage. This shows that an increase in per capita regional GDP will reduce investment in fixed asset investment and total energy consumption in the logistics industry, but will increase redundant investment in logistics industry employees and transportation route network mileage, resulting in waste.

[0157] The increase in urbanization rate has a significant negative correlation with the slack variables of logistics personnel input, road mileage input, and energy consumption, and is only weakly positively correlated with the investment in fixed assets in the logistics industry. Therefore, the increase in urbanization rate and the improvement of people's quality will significantly improve the redundancy of input resources, thereby improving the efficiency of green logistics.

[0158] The proportion of logistics fixed asset investment in regional fixed asset investment is significantly negatively correlated with the fixed investment slack of logistics industry, highway mileage investment, and total energy consumption, and is only weakly positively correlated with the investment of employees. This shows that the government's increase in the proportion of logistics fixed asset investment in regional fixed asset investment during overall planning is helpful to reduce the redundant waste of green logistics efficiency.

[0159] The results of the above regression model are shown in the table above. The next step is to use Excel to calculate the input data based on the above results and separate the management inefficiency items.

[0160] Furthermore, the third stage of static analysis:

[0161] Based on the adjusted input data after SFA regression and the original output data, this paper again selects the super-efficiency SBM model, and uses the Dearun3.0 software to calculate the adjusted input-output panel data of 31 cities in the middle reaches of the Yangtze River urban agglomeration from 2012 to 2021, and obtains the adjusted green logistics efficiency values ​​and scale returns of each city in the middle reaches of the Yangtze River urban agglomeration from 2012 to 2021.

[0162] As shown in Table 5, in the third stage after adjustment, the average comprehensive technical efficiency of the Yangtze River Middle Reaches Urban Agglomeration during 2012-2021 was 0.236, which was a significant drop from 0.425 in the first stage. Changsha City had the best efficiency value of 0.822, while Tianmen City ranked last with 0.030, showing a huge gap between cities. Figure 2 As shown, there is a gap in the green logistics efficiency values ​​of the 31 cities in the third stage of the middle reaches of the Yangtze River urban agglomeration from 2012 to 2021. The ten-year average values ​​of comprehensive technical efficiency between 31 cities in the third stage have a disparity. Only five cities have an average value exceeding 0.5, and the rest of the cities are concentrated in the range of 0 to 0.3.

[0163] Table 5 Average comprehensive technical efficiency of cities in the Yangtze River Middle Reaches Urban Agglomeration during the third phase from 2012 to 2021

[0164]

[0165]

[0166] Combined with the above analysis, we can see that: first, judging from the ten-year average efficiency, the overall comprehensive technical efficiency of green logistics in the urban agglomeration in the middle reaches of the Yangtze River is low; second, the development of green logistics among cities in the urban agglomeration in the middle reaches of the Yangtze River is extremely different, there is a gap in the development between cities, there are few leading cities, the strength of the waist is almost zero, and the green logistics efficiency of each city is concentrated in the tail. This shows that the process of green logistics in the urban agglomeration in the middle reaches of the Yangtze River is still in the initial development stage. There are leading cities, but the development process of most cities is relatively slow, and the gap between cities is extremely large.

[0167] Table 6 Comprehensive technical efficiency values ​​of cities in the Yangtze River Middle Reaches Urban Agglomeration during the third phase from 2012 to 2021

[0168]

[0169]

[0170] Combining Table 5 and Table 6, it is found that the cities with the top four averages have achieved the best green logistics efficiency values ​​in individual years. For example, in the past ten years, Changsha City achieved the best SBM efficiency in 2018, 2019, and 2021, Wuhan City in 2019 and 2021, Yichang City in 2021, and Nanchang City in 2018. In 2021, the three cities with the highest averages all achieved the best, which shows that in recent years, there have been new planning configurations in green logistics management, technological development, or resources, and these configurations are effective.

[0171] As shown in Table 7, in the data results of the third phase, during the period 2012-2021, the average pure technical efficiency of the urban agglomeration in the middle reaches of the Yangtze River was 0.854, which is a small gap from reaching the input-output SBM effective value of 1. Overall, excluding the impact of environmental factors, the overall level of pure technical efficiency of the urban agglomeration in the middle reaches of the Yangtze River is much higher than the comprehensive technical efficiency.

[0172] Table 7 Average pure technical efficiency of cities in the Yangtze River Middle Reaches Urban Agglomeration during the third phase from 2012 to 2021

[0173] area Pure technical efficiency value Sorting area Pure technical efficiency value Sorting Overall mean 0.854 - Shangrao 0.864 16 Xinyu 1.011 1 Fuzhou 0.853 17 Changsha 1.008 2 Xianning 0.830 18 Tianmen 1.007 3 Hengyang 0.821 19 Yiyang 1.001 4 Yellowstone 0.815 20 Qianjiang 0.999 5 Wuhan 0.813 21 Jingdezhen 0.991 6 Ezhou 0.786 22 Xiantao 0.987 7 Xiangtan 0.780 23 Yingtan 0.958 8 Yichun 0.749 24 Pingxiang 0.954 9 Jiujiang 0.742 25 Changde 0.924 10 Yueyang 0.736 26 Zhuzhou 0.920 11 Huanggang 0.732 27 Nanchang 0.914 12 Jingzhou 0.729 28 Xiangyang 0.913 13 Loudi 0.717 29 Yichang 0.907 14 Xiaogan 0.591 30 Ji'an 0.871 15 Jingmen 0.560 31

[0174] Combined with Table 7 and the following Figure 3It can be seen that 93.5% of the cities in the middle reaches of the Yangtze River have a pure technical efficiency value of more than 0.7, and only Xiaogan and Tianmen have a lower pure technical efficiency, in the range of 0.5-0.7. Among them, the pure technical efficiency of Xinyu, Changsha, Tianmen, and Yiyang exceeds 1, indicating that these four cities have performed well in the management and technical development of green logistics efficiency. The pure technical efficiency is basically at the head and waist, indicating that after eliminating the influence of environmental factors, the cities have a good configuration for the management and technical development of green logistics efficiency.

[0175] From Table 8 below, we can see that the pure technical efficiency of Xinyu City and Tianmen City has achieved effective input-output every year during the ten years. Cities with high average rankings can also achieve effective input-output in half of the years. In 2021, one-third of the cities have achieved effective input-output. The lowest efficiency is Xiaogan City, which is only 0.579.

[0176] Table 8 Pure technical efficiency values ​​of cities in the Yangtze River Middle Reaches Urban Agglomeration during the third phase from 2012 to 2021

[0177]

[0178]

[0179] Note: *The results are expressed to achieve the optimal efficiency of the SBM model. The results are obtained by the author from sorting out the results data of the third stage super-efficiency SBM model.

[0180] Based on the above analysis of pure technical efficiency, we can see that: First, the overall efficiency of the urban agglomeration in the middle reaches of the Yangtze River has not reached the optimal level, but the gap from the optimal level is not large; second, the head city effect is obvious, the mid-level cities have great potential, the tail cities are not far behind, and the overall development structure is positive.

[0181] As shown in Table 9 below, in the data results of the third phase, during the period 2012-2021, the average scale efficiency of the urban agglomeration in the middle reaches of the Yangtze River was only 0.277. The overall distribution was consistent with the distribution of the adjusted comprehensive technical efficiency, and the range was very large. The highest average value in Wuhan was close to 0.9, while the lowest scale efficiency value in Tianmen City was only 0.03.

[0182] Table 9 Average scale efficiency of cities in the Yangtze River Middle Reaches Urban Agglomeration during the third phase from 2012 to 2021

[0183]

[0184] From above Figure 4As shown in Table 10, the overall scale efficiency of the urban agglomeration in the middle reaches of the Yangtze River is poor, and the leading cities are still more than 10% away from the optimal growth. Other cities are concentrated in the tail and have a large gap with the leading cities. Combined with Table 10, it can be seen that only Wuhan and Changsha performed well in 2021, both achieving effective input-output of scale efficiency, while the remaining cities did not achieve effective results in other years.

[0185] Table 10 Scale efficiency values ​​of cities in the Yangtze River Middle Reaches Urban Agglomeration during the third phase from 2012 to 2021

[0186]

[0187] Note: *The results are expressed to achieve the optimal efficiency of the SBM model. The results are obtained by the author from sorting out the results data of the third stage super-efficiency SBM model.

[0188] Based on the above analysis of scale efficiency, we can see that: first, the overall efficiency of the urban agglomeration in the middle reaches of the Yangtze River is relatively poor, and there is a large gap from the effective input-output ratio; second, there are fewer head and middle cities and more tail cities; third, there is a large gap between the head and the tail.

[0189] Furthermore, the results of the changes in returns to scale in the third phase of the cities in the Yangtze River Middle Reaches Urban Agglomeration from 2012 to 2021 calculated by Dearun3.0 software show that only Wuhan alone has unchanged returns to scale in 2021, while the rest of the regions are in a state of increasing returns to scale at all time points. The unchanged returns to scale in Wuhan in 2021 indicate that the superiority of Wuhan's scale production has been fully utilized as of 2021, the scale diseconomies are decreasing, and the overall configuration has reached a relatively good state. However, after the third phase of data adjustment for all other cities, the returns to scale have been decreasing in all years, which shows that the overall green logistics efficiency of the Yangtze River Middle Reaches Urban Agglomeration is in a stage of rapid development, and increasing investment will help the green and high-quality development of the logistics industry in the urban agglomeration and the transformation of the industrial structure. Therefore, to promote the development of green logistics in the Yangtze River Middle Reaches Urban Agglomeration, it is necessary to concentrate resources and efforts to expand the scale of the green logistics market.

[0190] Furthermore, a third-stage Malmquist index model dynamic analysis is performed. After eliminating external factors and management inefficiency in the second stage, the present invention still uses the Malmquist index model to perform data analysis by inputting adjusted input-output data under the condition of variable returns to scale.

[0191] Specifically, Table 11 shows the total factor productivity index and its decomposition of the cities in the middle reaches of the Yangtze River urban agglomeration in nine time periods from 2012 to 2021 after the SFA regression analysis in the second stage.

[0192] Table 11 Malmquist productivity index and its decomposition mean for the Yangtze River Middle Reaches Urban Agglomeration during the third period 2012-2021

[0193]

[0194] Note: The results are obtained by the author from the third stage Malmquist index model result data.

[0195] From a time perspective, the average total factor productivity of each city from 2012 to 2021 was 1.080, which performed well, indicating that overall production efficiency has improved over the past decade and is in a steady and positive state. The main driving force is still the high technological progress rate index, which has slightly decreased compared to the first stage, but the change is not obvious. The average value of 2020-2021 is 1.171, the highest among them, indicating that the productivity growth in 2021 is the highest in the past decade. Among them, the overall production efficiency in 2018-2019 and 2019-2020 has declined.

[0196] After the adjustment, the number of years when the comprehensive technical efficiency change index did not meet the standard changed from two to five, indicating that the changes in the overall external environment have a greater impact on the pure technical efficiency and scale efficiency of the urban agglomeration in the middle reaches of the Yangtze River. After the adjustment, the total factor productivity and its decomposition have declined to varying degrees, but the most obvious decline is the pure technical efficiency change index, which shows that the management level and environmental conditions have a greater impact on the pure technical efficiency of the urban agglomeration in the middle reaches of the Yangtze River. It is necessary to increase the investment of scientific and technological resources in the urban agglomeration in the middle reaches of the Yangtze River that are conducive to energy conservation and emission reduction, and the development of low-carbon green logistics in order to enhance its technological innovation momentum.

[0197] During this decade, the Technology Progress Index was less than 1 only in the two time periods of 2013-2014 and 2019-2020.

[0198] As shown in Table 12, the individual analysis of cities will still be conducted from two entry points: individual mean and time dimension.

[0199] Table 12 Malmquist productivity index and its decomposition of cities in the Yangtze River Middle Reaches Urban Agglomeration in the third stage

[0200]

[0201] From the perspective of the total factor productivity index, the average total factor productivity of the Yangtze River Middle Reaches Urban Agglomeration is lower than before the adjustment, indicating that the development of green logistics in the Yangtze River Middle Reaches Urban Agglomeration is promoted by external environmental factors. After eliminating external environmental factors, the total factor productivity has declined. After the adjustment, except for Pingxiang City, the average values ​​of other cities are all over 1, indicating that the overall production efficiency of the Yangtze River Middle Reaches Urban Agglomeration is increasing.

[0202] From the perspective of the comprehensive technical efficiency change index, between 2012 and 2021, the average values ​​of the fourteen cities of Huangshi, Ezhou, Xiantao, Xiangyang, Zhuzhou, Hengyang, Yueyang, Changde, Jingdezhen, Pingxiang, Xinyu, Yingtan, Ji'an, and Fuzhou were below 1, while the average values ​​of the remaining cities were all above 1. In the first stage, only the comprehensive technical efficiency change index of Huangshi, Xiantao, Xiangyang, and Pingxiang was below 1, indicating that the green logistics technical efficiency of the remaining ten cities, such as Ezhou and Zhuzhou, was overestimated. At the same time, from a decomposition point of view, the pure technical efficiency of these cities that did not meet the comprehensive technical efficiency standards was very close to 1, and the main reason for their poor performance was still the scale efficiency index, indicating that the logistics market scale of these cities was not large enough and the layout of green logistics was not in place.

[0203] From the perspective of the technology progress index, after adjustments, the technology progress index of each city has not decreased significantly, and the average value is above 1. This shows that in the past ten years, the urban agglomeration in the middle reaches of the Yangtze River has attached great importance to the transformation of the industrial structure and green development of the logistics industry as a whole, and has strengthened its emphasis on technological innovation. On the whole, the development of the logistics industry is relatively matched with the development of the entire region.

[0204] After adjusting the input, the Malmquist productivity index of each city in the Yangtze River Middle Reaches Urban Agglomeration from 2012 to 2021 is shown in Table 13, which shows the changes in total factor productivity of each city year by year.

[0205] Table 13 Malmquist productivity index of cities in the Yangtze River Middle Reaches Urban Agglomeration during the third period 2012-2021

[0206]

[0207]

[0208] Note: The results are obtained by the author from the first stage Malmquist index model result data.

[0209] The Malmquist index of each city in the Yangtze River Middle Reaches Urban Agglomeration from 2012 to 2021 after the adjustment of input is shown in Table 13. This table can show in more detail the changes and rankings of the total factor productivity of each city in each of the ten years after the adjustment of input. It can be seen from the table that after the adjustment of input, the total factor productivity level in 2016 was the highest, which was different from before the adjustment; the productivity in 2019 and 2020 ranked at the bottom. Yiyang City has the highest average total factor productivity level, while Pingxiang and Yingtan have the lowest. From the perspective of years, compared with the first stage, the total factor productivity level has increased and decreased, but most years have decreased; from the perspective of cities, the total factor productivity level has increased at the overall level, and only Pingxiang City is lower than expected.

[0210] Finally, from the perspective of system science and system engineering, the fundamental way to improve the green logistics efficiency of the urban agglomeration in the middle reaches of the Yangtze River is to grasp the following four directions: concentrate resources, eliminate redundancy, and strive for production benefits through efficiency. Combined with the above analysis, this section summarizes the four main data characteristics of the development of green logistics efficiency in the urban agglomeration in the middle reaches of the Yangtze River, and combined with the economic and geographical characteristics of the urban agglomeration in the middle reaches of the Yangtze River, puts forward five suggestions for improving the urban agglomeration in the middle reaches of the Yangtze River.

[0211] Results 1: The overall green logistics efficiency level of the urban agglomeration in the middle reaches of the Yangtze River is low. From the perspective of time, the overall development trend is positive. All cities are in the stage of increasing returns to scale. Industry management and technological development need to be continuously optimized, and the problem of unreasonable market scale allocation is serious. Overall, the green logistics efficiency of the urban agglomeration in the middle reaches of the Yangtze River is low, and it has not reached effectiveness in all years during the statistical period. Among them, the average technical efficiency reached 0.854, and the scale efficiency was only 0.277. This shows that the comprehensive technical efficiency in its development is more affected by scale efficiency, there is improper scale investment, excessive resource investment causing waste, and the scale effect has not yet been formed, which is a key factor restricting development. The main reasons are: low internal management efficiency, lagging technological development, and serious unreasonable market scale allocation. Combined with the analysis of changes in scale returns, the scale returns of green logistics development in all cities are increasing, indicating that increasing investment will have more benefits.

[0212] Regarding feature one, it can be seen that the optimization of green logistics efficiency in the middle reaches of the Yangtze River urban agglomeration must start with improving scale efficiency and expanding the logistics market, and implementing appropriate management strategies and methods.

[0213] In the market economy system, insufficient market resources will seriously affect the optimization of logistics resource allocation. The focus of expanding the logistics market is to improve people's livelihood economy, enhance urban economic vitality and purchasing power, stimulate the occurrence of business flow, and help the generation of logistics activities. For example, the phenomenon of empty vehicles is a very obvious waste of logistics resources due to insufficient market resources and planning. Reasonable planning of logistics resources in advance can also avoid the waste of logistics resources.

[0214] The key to improving economies of scale is to support advantageous logistics industries and achieve lean production. The implementation of lean management helps reduce waste in the production process of the logistics industry, thereby improving its scale economy potential

[74] . Taking Ezhou as an example, by supporting advantageous logistics industries, the logistics park with Huahu Airport as the core can play the role of an airport economic circle through reasonable planning and investment attraction, while radiating the distribution of logistics goods in neighboring cities such as Wuhan and Huanggang.

[0215] At the same time, based on the increasing returns to scale of green logistics in the middle reaches of the Yangtze River urban agglomeration, the government's increase in the planned fixed asset investment in the logistics industry, road resources, the number of people employed in the logistics industry, and energy resources will also have a positive effect on improving the efficiency of green logistics.

[0216] Result 2: The green logistics efficiency of cities in the middle reaches of the Yangtze River urban agglomeration is greatly affected by external environmental factors. Among them, the increase in urbanization rate and the proportion of fixed asset investment in the logistics industry are conducive to the improvement of overall green logistics efficiency. Combined with the scale efficiency SFA regression analysis, after eliminating environmental factors and interference factors, the comprehensive efficiency obtained by the super-efficiency SBM model in the third stage is significantly different from that in the first stage: the scale efficiency has decreased overall, the pure technical efficiency has increased, and the comprehensive efficiency has still decreased. This result also once again proves the necessity of eliminating external factors in the second-stage SFA regression model. Among them, the increase in urbanization rate will significantly improve the redundancy of input resources, thereby improving green logistics efficiency; the increase in the proportion of fixed asset investment in the logistics industry during overall planning will help reduce the redundant waste of green logistics efficiency and improve overall efficiency.

[0217] Countermeasure and suggestion 2: Promote new urbanization in urban agglomerations and optimize the structure of fixed asset investment.

[0218] With the green logistics pilot demonstration area of ​​the main center and the sub-center as the core, a green logistics information sharing platform will be established to improve the logistics operation network system. After the logistics information sharing platform is established, all cities in the Yangtze River Middle Reaches Urban Agglomeration need to participate in and maintain it together to ensure the publication and sharing of logistics information, achieve complementary advantages, and work together to promote the green transformation of the industrial structure of the logistics industry in the Yangtze River Middle Reaches Urban Agglomeration.

[0219] Result 4: The development of green logistics among cities in the Yangtze River Urban Agglomeration is highly differentiated, and different cities have different advantages and weaknesses. According to the data on urban efficiency values, they are mainly divided into four categories: the first category is cities with good performance in both pure technical efficiency and scale efficiency; the second category is cities with high pure technical efficiency but low scale efficiency; the third category is cities with high scale efficiency but low pure technical efficiency; the fourth category is cities with low scale efficiency and pure technical efficiency.

[0220] Countermeasures and suggestions 4: Implement urban green logistics policies and promote multi-level differentiated development

[0221] In the process of green optimization of the industrial structure of the logistics industry in the middle reaches of the Yangtze River urban agglomeration, the green logistics efficiency of different cities is inconsistent, and the problems they face are also significantly different. Therefore, it is necessary to implement a multi-level and differentiated green logistics policy based on the specific conditions of the green logistics efficiency level and investment redundancy of each city.

[0222] First, implement a green logistics policy with differentiated development. Combined with the green physical efficiency characteristics of each city. For cities with low green logistics efficiency and high carbon emissions, they should strengthen investment in logistics infrastructure and equipment, and guide urban enterprises to use new energy and low-carbon logistics facilities and equipment and transportation tools; for cities with a large proportion of redundant logistics practitioners, they should focus on improving the relevant skills and comprehensive quality of practitioners, and should also pay attention to the rational use of logistics robots and AGVs to improve the overall logistics efficiency; for cities with a large proportion of redundant logistics fixed asset investment and infrastructure investment, they should absorb the country's advanced logistics technology and management concepts, benchmark the green logistics development pilot areas, and improve the efficiency of resource allocation.

[0223] Second, implement a green logistics policy with multi-level development. Combined with the ranking of green logistics efficiency of each city and the city positioning, cities with low green logistics efficiency but good logistics scale and economic development should attach importance to and make in-depth layout of green logistics development, and highlight the importance of green development of industrial structure; for cities with poor logistics scale development and insufficient economic vitality, the focus should be on developing their logistics scale, rationally optimizing resource allocation, using the logistics industry to drive grassroots social employment, and realizing green logistics efficiency optimization with less investment and more returns; cities with low pure technical efficiency and scale efficiency still need to improve in technology innovation, industrial structure and scale resource allocation, learn from the experience of leading cities, and steadily achieve green logistics efficiency growth.

[0224] Geographical characteristics: The industrial structure of the Yangtze River Middle Reaches urban agglomeration is undergoing transformation and upgrading, and the exchange of materials within and outside the urban agglomeration is becoming more frequent. Based on the above-mentioned characteristics of green logistics efficiency of the Yangtze River Middle Reaches urban agglomeration, the Yangtze River Middle Reaches urban agglomeration has a large number of ports and inland rivers.

[0225] Countermeasure and Suggestion 5: Improve the logistics and transportation network and optimize the layout of green transportation infrastructure.

[0226] From the above data analysis, it can be seen that the poor efficiency of green logistics in most cities is mostly reflected in their poor scale efficiency. The development of urban transportation networks can alleviate the negative impact on regional ecosystem functions. Combined with the optimization of logistics and transportation networks, it will help to increase the logistics market of the urban agglomeration in the middle reaches of the Yangtze River, expand the scale of logistics in the urban agglomeration in the middle reaches of the Yangtze River, make it have scale effects, reduce marginal costs, and improve green logistics efficiency. The focus can be started from the following three aspects: first, promote the green and high-quality development of ports and inland rivers; second, promote the cross-regional integrated development of multimodal transport; third, optimize the urban distribution system.

[0227] Promote the green and high-quality development of ports and inland rivers. The urban agglomeration in the middle reaches of the Yangtze River has excellent ports and inland rivers. We should make every effort to promote the construction and high-quality development of green ports and inland rivers, and improve the shortcomings and weaknesses of the logistics and transportation system, which is of great significance to promoting the adjustment of transportation structure and implementing green logistics development.

[0228] Promote the integrated development of multimodal transport across regions. We should combine the characteristics of each city, take advantage of the geographical features of lakes, inland rivers, and branches of the Yangtze River, and combine road transport, rail transport, and civil aviation transport to achieve integrated development of multimodal transport across regions and improve the development market of the modern logistics industry.

[0229] Optimize the urban distribution system. This not only involves the planning of highway routes, but also involves the planning of storage facilities such as forward warehouses and central warehouses. Optimizing the network of warehouses, trunk lines and branch lines will help to optimize the urban distribution system, solve the last mile distribution problem, improve the logistics and transportation network from a finer granularity, and optimize the green logistics and transportation infrastructure.

[0230] In summary, the model method selected by this invention for the green logistics characteristics of urban agglomerations and the systematic analysis of urban agglomerations constructed

[0231] The framework system of green logistics efficiency of urban agglomerations is reasonable and feasible after data verification and validation. The green logistics efficiency evaluation of urban agglomerations is suitable for the analysis of super-efficiency SBM model, SFA model and Malmquist index combination of non-expected output under variable scale returns, and the constructed index system also meets the preconditions for the use of model methods.

[0232] Although embodiments of the present invention have been shown and described, those skilled in the art will appreciate that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the claims and their equivalents.

Claims

1. A method for constructing a green logistics efficiency evaluation and analysis model based on urban agglomerations, characterized in that: The method comprises: Step 1: construct an urban agglomeration green logistics efficiency evaluation index framework; the urban agglomeration green logistics efficiency evaluation index framework includes input indicators and output indicators, and the output indicators include expected output indicators and non-expected output indicators; The input indicators include the fixed asset investment in the logistics industry, the total mileage of highways, and the number of people employed in the logistics industry; The expected output indicators include total energy consumption, total logistics industry output value and freight volume; The undesirable output indicators include carbon dioxide emissions and carbon emission intensity of the logistics industry; Step 2: Construct a green logistics environmental variable indicator system for urban agglomerations, including the urbanization rate of per capita regional GDP and the proportion of logistics fixed asset investment in regional fixed asset investment; Step 3: Construct a three-stage analysis model of the super-efficiency SBM-SFA-Malmquist index model based on the urban agglomeration green logistics efficiency evaluation index framework and the urban agglomeration green logistics environmental variable index system; The three-stage analysis model of the super efficiency SBM-SFA-Malmquist index model is constructed as follows: In the first stage, a super-efficiency SBM model based on non-expected output is designed under the scenario of variable returns to scale. The input-output data of green logistics efficiency of each city in the urban agglomeration are evaluated and calculated to obtain the green logistics efficiency value and slack variable value. The green logistics efficiency value is the static efficiency value, and the slack variable value is the redundant term. At the same time, a Malmquist index model is constructed to calculate the original input and output, and the dynamic efficiency value based on the original input and output data is calculated; In the second stage, the slack values ​​of the input variables of the super-efficiency SBM model results were combined with environmental variables to conduct SFA regression analysis to eliminate environmental factors; In the third stage, the green logistics input data of urban agglomerations that have undergone SFA regression analysis are used to re-analyze the static and dynamic operational efficiency of green logistics in urban agglomerations using the super-efficiency SBM model with re-unexpected output and the Malmquist index model.

2. The method for constructing a green logistics efficiency evaluation and analysis model based on urban agglomerations according to claim 1 is characterized in that: The environmental variables include environmental effects, management inefficiency and random interference.

3. The method for constructing a green logistics efficiency evaluation and analysis model based on urban agglomerations according to claim 1 is characterized in that: The steps for constructing the green logistics efficiency evaluation index framework of the urban agglomeration are as follows: The framework of the input-output evaluation index system is constructed based on the Cobb-Douglas production function, which is expressed as follows: Y=A t k α L β m Among them, Y is the output, At is the comprehensive technical level, K is the invested capital, L is the invested labor, μ is the random interference; α and β represent the output elasticity of L and K respectively; Combined with the selected non-desirable output indicators, the outputs with environmental impacts are incorporated into the output indicators.

4. The method for constructing a green logistics efficiency evaluation and analysis model based on urban agglomerations according to claim 1 is characterized in that: The fixed asset investment amount of the logistics industry represents the capital input of the logistics industry; the total mileage of roads represents the infrastructure input of the logistics industry; the number of employees in the logistics industry represents the labor input of the logistics industry; the total energy consumption represents the material input of the logistics industry; the gross domestic product of the logistics industry represents the economic contribution level of the logistics industry; the freight volume represents the service output of the logistics industry; the carbon dioxide emissions of the logistics industry represent the carbon emission status of the logistics industry; the carbon emission intensity of the logistics industry represents the carbon dioxide emissions generated per unit GDP of the logistics industry, where carbon emission intensity = total carbon dioxide emissions of the logistics industry / gross domestic product of the logistics industry.

5. The method for constructing a green logistics efficiency evaluation and analysis model based on urban agglomerations according to claim 1 is characterized in that: The environmental influencing factors include management inefficiency items and random error items.

6. The method for constructing a green logistics efficiency evaluation and analysis model based on urban agglomerations according to claim 1 is characterized in that: The super-efficiency SBM model based on non-expected output is expressed as follows: Among them, ρ * is the target efficiency value, ρ * ≥1 indicates that the efficiency is effective, and the maximum representative decision-making unit is the most efficient. Otherwise, ρ * <1 indicates that there is still room for improvement in efficiency; r is the rth expected output; j is the jth decision-making unit; n is the number of decision-making units, q is the number of decision-making unit inputs; u1 is the expected output; u2 is the undesired output; x i , are the elements in the corresponding input matrix, expected output matrix, and unexpected output matrix respectively; i is the i-th weight vector; are the variable reference points for eliminating the input matrix, expected output matrix, and unexpected output matrix of the kth decision-making unit; i is the i-th input of the decision-making unit; x ik is the input matrix of the kth decision-making unit; is the expected output matrix of the kth decision-making unit; is the undesired output matrix of the kth decision making unit.

7. The method for constructing a green logistics efficiency evaluation and analysis model based on urban agglomerations according to claim 6 is characterized in that: According to the Malmquist productivity index model, the changes of global total factor productivity TFP before and after are calculated to establish a decision-making unit for dynamic efficiency evaluation, and the dynamic input-output distance function is constructed as follows: TFP=TPR×TEC=TPR×PTE×SE Among them, TPR represents the rate of technological progress, TEC represents the technical efficiency change index, PTE represents pure technical efficiency, and SE represents scale efficiency. Since technical efficiency is equal to pure technical efficiency multiplied by scale efficiency, the total factor productivity TFP is decomposed into: Among them, P c represents total factor productivity, t and t+1 represent the immediate and forward periods, N represents the scale index, M represents the pure technology index, a and b represent the input and output data respectively, They represent the dynamic changes of technical efficiency in the immediate and long term, according to the dynamic changes across time periods; It represents the scale efficiency; represents pure technical efficiency; Indicates the technical efficiency change index According to the Malmquist productivity index model, when p>1, it means that the total factor productivity is gradually increasing, and vice versa.

8. The method for constructing a green logistics efficiency evaluation and analysis model based on urban agglomerations according to claim 1 is characterized in that: The second stage specifically includes: The slack value of the input variable of the first-stage super-efficiency SBM model result is taken as the explained variable Y, and the environmental variable is taken as the explanatory variable X to construct the stochastic frontier model, which is expressed as follows: s ni,t =f n (z i,t ;y n )+v ni,t +ε ni,t where s ni,t represents the total value of the slack variable invested by the ith decision-making unit in year n, z i,t represents the environmental variable of the ith decision-making unit in year t, y n represents the unknown parameter to be estimated, y n represents the feasible stochastic frontier of the nth input, measuring the impact of environmental factors on the slack variable, v ni,t To manage inefficiencies, ε ni,t is a random error term. Calculate the alpha value: Among them, when When , the inefficiency term has the greatest impact on each slack variable; Calculate v ni and ε n The value, excluding the influence of external environment and random interference and adjusting the input-output variables, is expressed as follows: Calculate the adjusted input variable value It is expressed as follows: Among them, xni,t and Respectively represent the input amount before and after adjustment, and They represent the optimal external environmental factors and the decision-making unit with the least random interference, and They respectively represent the elimination of the influence of external environment and random interference.

9. The method for constructing a green logistics efficiency evaluation and analysis model based on urban agglomerations according to claim 8 is characterized in that: The SFA regression analysis needs to verify the significance of the data, so the data is subjected to a T test to determine whether the factors affecting efficiency screened out have an impact and interference on the explained variables at least at the level of 10%, that is, the α value is less than 0.1, which is used to measure the relative importance of inefficiency in the total error. When α is close to 1, it means that the inefficiency items have a major influence on the total error, which can support the applicability of the SFA model in this scenario; the test value of the likelihood ratio test LR must be greater than the critical value of the chi-square distribution with a degree of freedom of 1 at a significance level of 1‰. .

10. The method for constructing a green logistics efficiency evaluation and analysis model based on urban agglomerations according to claim 7, characterized in that: In the third stage, the processed data are re-entered into the super-efficiency SBM model based on non-expected output to obtain the adjusted input variables, and at the same time re-entered into the Malmquist productivity index model to recalculate the adjusted global total factor productivity.

Images