Theoretical framework
The traditional high-carbonization economic development mode pursues the maximum economic output with the minimum input of factors. Although this mode greatly accelerates the process of industrialization and urbanization, it also causes severe environmental problems. The key to solve the problem lies in the transformation from the high-carbonization economic development mode to the low-carbonization economic development mode. Low-carbonization economic development mode pursues the maximum economic output and less carbon emissions with the minimum factor input compared with the high-carbonization economic development mode [27]. Therefore, Based on the theory of environmental externality and sustainable development, the paper firstly constructs the evaluation index system of economic development performance under carbon emission constraint according to the connotation of low-carbon economy. Secondly, the DEA model and Malmquist productivity index are used to calculate technical efficiency, pure technical efficiency, scale efficiency, total factor productivity, technological change, scale efficiency change index and pure technical efficiency change, and analyze the temporal and spatial evolution characteristics of static and dynamic performance of economic development. Finally, the evaluation index system of influencing factors on economic development performance is constructed from endogenous and exogenous variables, and the Tobit model is used to find its key driving factors, and the optimization and promotion strategy of low-carbon economy is proposed (Figure 1).
On the basis of referring to relevant research, the paper selects input indicators from land, labor, capital and technology, and selects output indicators from economy and carbon emissions to construct the evaluation index system of economic development performance under carbon emissions constraint. Among them, urban built-up area, number of social employees at the end of the year, fixed asset investment and R&D (Research and Development) expenditure are selected as input indicators, and Gross Domestic Product and carbon emissions are selected as output indicators (Table 1). It should be noted that the output index of carbon emission is treated as an input index in the paper to ensure the objectivity of the evaluation results because the carbon emission is an index of non-expected output.
Table 1 Index system of economic development performance under carbon emission constraints
Index attribute
|
Index selection
|
Definition of Indicator
|
Input indicators
|
Urban built-up area /km2
|
land input
|
Number of employees at year-end/ten thousand
|
Labor input
|
Total investment in fixed assets / 100 million yuan
|
Fund input
|
R&D expenditure / 100 million yuan
|
Technology input
|
Total carbon emission /t
|
Carbon emission pressure
|
Output indicators
|
Gross domestic product / 100 million yuan
|
Economic output
|
Research methods
Data envelopment analysis method
Data envelopment analysis (DEA) is a method to evaluate the efficiency of decision units with multiple inputs and outputs. Assuming that there are L kinds of input and M kinds of output indicators, xjl represents the input amount of the L kinds of resources, and yjm represents the output amount of the M kinds of resources in the J province. For n (n= 1, 2..., K) province, its formula is [28]:
Where θ(0<θ≤1) is the technical efficiency index, λj(λj ≥0)is the weight variable, s-(s-≥0)is the relaxation variable, s+(s+≥0) is the residual variable, ε is non-Archimedean infinitesimal, e1T=(1,1,…,1)∈Em and e2T=(1,1,…,1)∈Ek are m and k-dimensional unit vector spaces respectively, the closer θ is to 1, the higher the score is for n province, Otherwise, the lower the score is. Equation (1) can be transformed into a DEA model with variable returns to Scale (VRS) by introducing the constraint condition =1. Using VRS model, the technical efficiency can be decomposed into pure technical efficiency (Vrste) and Scale efficiency (Scale).
Malmquist productivity index
The efficiency value obtained by DEA model can be used to measure the relative performance of each decision-making unit. When studying the change of efficiency value in different periods, it is necessary to introduce Manquist productivity index. Let (Xt,Yt) and (Xt+1,Yt+1) be the input-output relationship in t and t+1 period, the change of input-output relationship from (Xt,Yt) to (Xt+1,Yt+1) is the change of total factor productivity. Dct (xt,yt) and Dct+1(xt+1,yt+1) are distance functions. Malmquist productivity index based on t and t+1 reference technology is [27]:
- (2)
(3)
According to the idea of rational index, the geometric average value between the two is defined as the total factor productivity index, and its formula is:
(4)
Malmquist productivity can be decomposed into technical efficiency (Effch) and technological change (Techch) under VRS hypothesis. Besides, the change in technical efficiency can be decomposed into pure technical efficiency (Pech) and scale efficiency (Sech). Equation (4) can be decomposed into:
(5)
Where M(xt,yt,xt+1,yt+1) is the total factor productivity, Techch, Pech and Sech are the change indices of technological change, pure technical efficiency and scale efficiency. When the change index of total factor productivity, technological change, pure technical efficiency and scale efficiency is greater than 1, it indicates a positive trend, otherwise, it tends to deteriorate.
Tobit regression model
The paper chooses Tobit regression model to analyze the influencing factors of economic development performance under carbon emission constraint because the technical efficiency score value obtained by DEA model is truncated, and its formula [29] is as follows:
Where is the dependent variable, Yi is the technical efficiency index, Xi is the independent variable, a represents the coefficient, and b is the error term.
Data source
The paper takes 30 provincial units in mainland China (except Hong Kong, Macao, Taiwan, and Tibet) as the basic regional object, measures the score of economic development performance under carbon emissions constraint, and analyzes its influencing factors. Related data mainly come from China Statistical Yearbook of 2009, 2013, 2017 and 2021, China Energy Statistical Yearbook and provincial statistical yearbook. It should be noted that the carbon emissions data are calculated using the methods provided by the IPCC (Intergovernmental Panel on Climate Change), that is, the total amount of various energy consumption is multiplied by their average low calorific value and CO2 emission coefficient [30]. In addition, since the data of energy consumption per unit GDP of each province in 2016 and 2020 cannot be directly obtained, the paper uses the total energy consumption of each province divided by its total Gross domestic product to obtain the conversion.
Evaluation of China's economic development performance
Static performance evaluation of economic development
Measure analysis of static performance level
The software Deap2.1 is used to measure the static performance level of economic development of 30 provincial units based on the collected input-output index data, and the technical efficiency of each province and its exponential decomposition results are shown in Table 2.
Table 2 Economic development performance score under carbon emission constraints from 2008 to 2020
Province
|
2008 year
|
2012 year
|
2016 year
|
2020 year
|
mean
|
crste
|
vrste
|
scale
|
crste
|
vrste
|
scale
|
crste
|
vrste
|
scale
|
crste
|
vrste
|
scale
|
crste
|
Beijing
|
1.00
|
1.00
|
1.00
|
1.00
|
1.00
|
1.00
|
1.00
|
1.00
|
1.00
|
1.00
|
1.00
|
1.00
|
1.00
|
Tianjin
|
0.76
|
0.82
|
0.93
|
0.78
|
0.78
|
1.00
|
0.93
|
1.00
|
0.93
|
1.00
|
1.00
|
1.00
|
0.87
|
Hebei
|
0.50
|
0.67
|
0.74
|
0.48
|
0.52
|
0.93
|
0.52
|
0.52
|
1.00
|
0.48
|
0.53
|
0.91
|
0.50
|
Shanxi
|
0.55
|
0.56
|
0.99
|
0.57
|
0.63
|
0.91
|
0.67
|
0.70
|
0.95
|
0.78
|
0.86
|
0.91
|
0.64
|
Inner Mongolia
|
0.78
|
0.84
|
0.93
|
0.75
|
0.87
|
0.86
|
0.70
|
0.72
|
0.97
|
0.74
|
0.76
|
0.98
|
0.74
|
Liaoning
|
0.84
|
1.00
|
0.84
|
0.87
|
1.00
|
0.87
|
1.00
|
1.00
|
1.00
|
1.00
|
1.00
|
1.00
|
0.93
|
Jinlin
|
1.00
|
1.00
|
1.00
|
1.00
|
1.00
|
1.00
|
1.00
|
1.00
|
1.00
|
1.00
|
1.00
|
1.00
|
1.00
|
Heilongjiang
|
1.00
|
1.00
|
1.00
|
1.00
|
1.00
|
1.00
|
1.00
|
1.00
|
1.00
|
1.00
|
1.00
|
1.00
|
1.00
|
Shanghai
|
0.63
|
0.66
|
0.95
|
0.88
|
0.89
|
0.99
|
0.61
|
0.64
|
0.95
|
0.70
|
0.71
|
0.98
|
0.71
|
Jiangsu
|
0.70
|
0.78
|
0.90
|
0.79
|
0.88
|
0.90
|
0.80
|
0.90
|
0.89
|
0.78
|
0.97
|
0.80
|
0.77
|
Zhejiang
|
0.68
|
0.72
|
0.95
|
0.76
|
0.76
|
1.00
|
0.74
|
0.76
|
0.98
|
0.90
|
0.77
|
0.90
|
0.77
|
Anhui
|
0.94
|
0.94
|
1.00
|
0.97
|
0.98
|
0.99
|
0.91
|
0.92
|
0.99
|
0.86
|
0.86
|
1.00
|
0.92
|
Fujian
|
0.59
|
0.59
|
1.00
|
0.67
|
0.68
|
0.98
|
0.64
|
0.65
|
0.99
|
0.61
|
0.62
|
0.99
|
0.63
|
Jiangxi
|
0.84
|
0.84
|
1.00
|
0.92
|
0.99
|
0.93
|
0.90
|
0.90
|
1.00
|
0.89
|
0.91
|
0.98
|
0.89
|
Shandong
|
0.65
|
0.88
|
0.74
|
0.70
|
0.85
|
0.82
|
0.75
|
0.90
|
0.83
|
0.80
|
1.00
|
0.80
|
0.73
|
Henan
|
0.60
|
1.00
|
0.60
|
0.66
|
0.95
|
0.69
|
0.64
|
0.74
|
0.87
|
0.70
|
0.80
|
0.88
|
0.65
|
Hubei
|
0.78
|
0.79
|
0.99
|
0.74
|
0.74
|
1.00
|
0.77
|
0.79
|
0.98
|
1.00
|
1.00
|
1.00
|
0.82
|
Hunan
|
0.61
|
0.61
|
1.00
|
0.60
|
0.60
|
1.00
|
0.59
|
0.59
|
1.00
|
0.64
|
0.64
|
1.00
|
0.61
|
Guangdong
|
1.00
|
1.00
|
1.00
|
1.00
|
1.00
|
1.00
|
1.00
|
1.00
|
1.00
|
1.00
|
1.00
|
1.00
|
1.00
|
Guangxi
|
0.77
|
0.89
|
0.87
|
0.80
|
0.84
|
0.95
|
0.81
|
0.86
|
0.94
|
0.80
|
0.81
|
0.99
|
0.80
|
Hainan
|
1.00
|
1.00
|
1.00
|
1.00
|
1.00
|
1.00
|
1.00
|
1.00
|
1.00
|
1.00
|
1.00
|
1.00
|
1.00
|
Chongqing
|
0.69
|
0.70
|
0.99
|
0.79
|
0.80
|
0.99
|
0.85
|
0.85
|
1.00
|
0.92
|
0.93
|
0.99
|
0.81
|
Sichuan
|
0.61
|
0.66
|
0.93
|
0.67
|
1.00
|
0.67
|
0.79
|
0.96
|
0.82
|
0.87
|
1.00
|
0.87
|
0.74
|
Guizhou
|
0.58
|
0.59
|
0.98
|
0.62
|
0.64
|
0.97
|
0.64
|
0.65
|
0.99
|
0.61
|
0.62
|
0.99
|
0.61
|
Yunnan
|
0.63
|
0.82
|
0.77
|
0.74
|
0.81
|
0.91
|
0.71
|
0.72
|
0.99
|
0.55
|
0.56
|
0.99
|
0.66
|
Shaanxi
|
0.55
|
0.55
|
1.00
|
0.56
|
0.56
|
1.00
|
0.58
|
0.59
|
0.98
|
0.57
|
0.58
|
0.99
|
0.57
|
Gansu
|
0.89
|
0.89
|
1.00
|
0.88
|
0.89
|
0.99
|
0.97
|
0.98
|
0.99
|
0.89
|
1.00
|
0.89
|
0.91
|
Qinghai
|
0.57
|
1.00
|
0.57
|
0.46
|
1.00
|
0.46
|
0.77
|
1.00
|
0.77
|
0.69
|
1.00
|
0.69
|
0.62
|
Ningxia
|
1.00
|
1.00
|
1.00
|
1.00
|
1.00
|
1.00
|
1.00
|
1.00
|
1.00
|
1.00
|
1.00
|
1.00
|
1.00
|
Xinjiang
|
1.00
|
1.00
|
1.00
|
1.00
|
1.00
|
1.00
|
1.00
|
1.00
|
1.00
|
1.00
|
1.00
|
1.00
|
1.00
|
mean
|
0.76
|
0.83
|
0.92
|
0.79
|
0.86
|
0.93
|
0.81
|
0.84
|
0.96
|
0.83
|
0.86
|
0.95
|
,0.80
|
According to the Table 2, it can be found that China's economic development performance under carbon emission constraint and its decomposition index have the following characteristics: First, the static performance level is generally low, and only a few provinces have reached the optimal level of economic development performance. The technical efficiency index of 2008, 2012, 2016 and 2020 were 0.76, 0.79, 0.81 and 0.83 respectively, indicating that the static performance level of China's economic development under carbon emissions constraint only reached 76%, 79%, 81% and 83% of the optimal level. The average value of the technical efficiency index of China's economic development in the four years is 0.80, indicating that the static performance level in the research period is only 80% of the optimal level, and there is still great potential for improvement. In the future, the static performance level can be improved by optimizing the allocation of all input factors. In terms of provinces and regions, the number of provincial units with technical efficiency index 1 in 2008, 2012, 2016 and 2020 is 7, 7, 8 and 10, respectively, accounting for 23%, 23%, 27% and 33% of the total number of research units. Among them, the technical efficiency index of Beijing, Jilin, Heilongjiang, Guangdong, Hainan, Ningxia and Xinjiang is always 1, indicating that the static performance level of economic development under carbon emissions constraint of these provinces has reached the optimal level, and there is no input redundancy and output shortage. In addition, the technical efficiency index of the other provincial units fluctuates and does not reach the optimum at the above time nodes, which also indicates that the static performance of most provincial units still has a large room for improvement. Second, the pure technical efficiency index is higher than the technical efficiency index and slightly lower than the scale efficiency index. The pure technical efficiency index is the key to the improvement of economic development performance. In 2008, 2012, 2016 and 2020, the pure technical efficiency index was 0.83, 0.86, 0.84 and 0.86 respectively, indicating that the pure technical efficiency only played 83%, 86%, 84% and 86% of the optimal level. The average value of the pure technical efficiency index of China's economic development in the four years is 0.85, indicating that the pure technical efficiency only reaches 85% of the optimal level during the study period, which has great potential for mining. Specifically, the number of provincial units with optimal pure technical efficiency in 2008, 2012, 2016 and 2020 is 10, 10, 10 and 14, accounting for 33%, 33%, 33% and 47% of the total number of units. The pure technical efficiency index of Beijing, Liaoning, Jilin, Heilongjiang, Guangdong, Hainan, Qinghai, Ningxia and Xinjiang is always 1, indicating that the pure technical efficiency of these provincial units has been fully utilized in the process of economic development. The pure technical efficiency index of other provincial units fluctuates to different degrees and fails to reach the optimum at the four time nodes at the same time, indicating that the pure technical efficiency of these provincial units has not been fully played, and the pure technical efficiency can be improved by adjusting the quantity of each input factor in the future. In addition, the pure technical efficiency index of China's provinces is slightly smaller than the scale efficiency index, it is not difficult to deduce that pure technical efficiency is the key to restrict the static performance improvement of China's economic development combined with the characteristics of technical efficiency and its decomposition index. Third, the number of provincial units with the best scale efficiency index is significantly higher than that with the best technical efficiency. Giving full play to scale efficiency is still an effective way to improve the performance of economic development. In 2008, 2012, 2016 and 2020, China's scale efficiency index is 0.92, 0.93, 0.96 and 0.95 respectively, indicating that scale efficiency is only 92%, 93%, 96% and 95% of the optimal level. The average value of scale efficiency in the four years is 0.94, indicating that the scale efficiency of economic development in the research period has been fully played, reaching 94% of the optimal level. Besides, the number of provincial units with optimal scale efficiency in 2008, 2012, 2016 and 2020 was 13, 12, 12 and 12, which were higher than the technical efficiency index and pure technical efficiency index in most years. The scale efficiency indexes of Beijing, Heilongjiang, Jilin, Hunan, Guangdong, Hainan, Ningxia and Xinjiang are all 1 in different years, indicating that there is no factor input redundancy and output insufficiency from the perspective of scale efficiency and the scale efficiency indexes for other provincial units show the fluctuation characteristics in the aspect of time series. It is suggested that these provincial units should adjust the allocation ratio of various factors in time to avoid the phenomenon of excessive or insufficient resource input. Remarkably, the scale efficiency index is relatively close to the optimal level, so the static performance can be boosted to a certain extent by expanding production scale.
Spatial differentiation characteristics of static performance
According to the measurement results of economic development performance in 2008, 2012, 2016 and 2020, the paper classifies them into four types by using Jerks classification method, and analyzes the spatial distribution pattern of static performance of economic development (Figure 2).
As can be seen from Figure 2, the spatial distribution characteristics of static performance of China's economic development under carbon emissions constraint from 2008 to 2020 are as follows: First, the static performance of economic development is significantly different and shows a trend of rising first and then falling. The maximum value of static performance of China's economic development in 2008, 2012, 2016 and 2020 is 1. Meanwhile, the provincial units with low static performance are Hebei, Shaanxi, Guizhou, Hunan and Qinghai respectively. The average value of technical efficiency index is 0.50, 0.57, 0.61, 0.61 and 0.62. It is only 50%, 57%, 61%, 61% and 62% of optimal level, far lower than the national average level of 80% in the same period, indicating that the static performance of China's economic development under carbon emissions constraint is very significant. In terms of the spatial distribution of static performance, there were 8, 9, 9 and 10 provincial units in the higher level area, 7, 6, 4 and 7 provincial units in the high level area, 6, 10, 9 and 7 provincial units in the medium level area, and 9, 5, 8 and 6 provincial units in the low level area in 2008, 2012, 2016 and 2020. According to the spatial variation of higher and high level areas, it is not difficult to find that the static performance of China's economic development generally showed an upward trend from 2008 to 2012, while it showed a downward trend from 2012 to 2020.
Second, the number of provincial units in different performance categories gradually stabilized over time, and showed obvious staggered layout characteristics in spatial distribution. From 2008 to 2012, the number of provincial units that changed in the static performance category was 10, namely Jiangxi, Chongqing, Jiangsu, Sichuan, Guizhou, Henan, Hubei, Zhejiang, Fujian and Inner Mongolia, the number of provincial units that changed in static performance category from 2012 to 2016 was 11, including Liaoning, Gansu, Anhui, Jiangxi, Qinghai, Guangxi, Jiangsu, Shanghai, Henan, Guizhou and Fujian, the number of provincial units with changes in static performance type from 2016 to 2020 was 9, including Hubei, Tianjin, Gansu, Sichuan, Zhejiang, Shanghai, Shanxi, Henan and Yunnan. It can be found that the static performance type areas of economic development in China have changed over time, mainly in Henan, Hubei, Yunnan, Guangxi, Jiangsu, Zhejiang and Shanghai, while the static performance in the vast majority of other provincial units remain stable by comparing the evolution characteristics of spatial pattern. In addition, the static performance of China's economic development shows that different type areas are mixed with each other in spatial distribution, showing a development trend of scattered layout. However, the reasons for the differences in static performance in these provincial units are different. Henan, Hubei and Sichuan have extensive land use patterns, limited capital input, low energy utilization efficiency, less advanced technology level and huge population base, which make them have great redundancy in land, capital, technology, labor force and environment input, and great output deficiency in economic development. Input and output elements are generally in a state of medium level "imbalances", which is an important reason limiting the static performance improvement for above provinces. Yunnan and Guangxi are located in less developed areas in China, with less carbon emissions from energy consumption and less input of land, labor, capital and technology. Therefore, these provinces are in a low level of "imbalance" in terms of factor input and output, which is the reason for their low static performance of economic development. Shanghai, Jiangsu and Zhejiang are located in developed coastal areas with solid economic development foundation, advanced technology level and high degree of economical and intensive use of land. However, in the process of development, the consumption of fossil energy is huge, resulting in high carbon emissions, and all factors are in a high level of "imbalance". Therefore, high carbon emissions are the reason for the low static performance in Shanghai, Jiangsu and Zhejiang.
Dynamic performance evaluation of economic development
In order to deeply analyze the dynamic change characteristics of China's economic development performance from 2008 to 2020, the paper also uses Deap2.1 software to process the input-output index data, and uses Malmquist productivity index to calculate the dynamic change value of economic development performance under carbon emissions constraint. The obtained variable value of economic development performance and its decomposition index are shown in Table 3.
Table 3 Dynamic changes of China's economic development performance from 2008 to 2020
Province
|
2008-2012 year
|
2012-2016 year
|
2016-2020 year
|
techch
|
pech
|
sech
|
tfpch
|
techch
|
pech
|
sech
|
tfpch
|
techch
|
pech
|
sech
|
tfpch
|
Beijing
|
1.43
|
1.00
|
0.50
|
0.72
|
0.75
|
1.00
|
1.55
|
1.16
|
1.06
|
1.00
|
1.08
|
1.14
|
Tianjin
|
1.11
|
0.87
|
0.97
|
0.94
|
1.20
|
1.15
|
0.96
|
1.32
|
1.15
|
1.00
|
1.07
|
1.23
|
Hebei
|
0.77
|
0.78
|
1.27
|
0.76
|
0.87
|
1.00
|
1.07
|
0.93
|
1.08
|
1.01
|
0.91
|
0.99
|
Shanxi
|
0.73
|
1.11
|
0.94
|
0.75
|
0.83
|
1.11
|
1.04
|
0.95
|
0.90
|
1.24
|
0.96
|
1.08
|
Inner Mongolia
|
0.97
|
1.00
|
0.93
|
0.90
|
0.99
|
0.81
|
1.15
|
0.92
|
1.11
|
1.05
|
1.01
|
1.18
|
Liaoning
|
1.04
|
1.00
|
0.99
|
1.03
|
1.32
|
1.00
|
1.09
|
1.43
|
0.89
|
1.00
|
1.00
|
0.89
|
Jinlin
|
0.94
|
1.00
|
1.00
|
0.94
|
0.89
|
1.00
|
1.00
|
0.89
|
1.10
|
1.00
|
1.00
|
1.10
|
Heilongjiang
|
0.75
|
1.00
|
1.00
|
0.75
|
0.85
|
1.00
|
1.00
|
0.85
|
1.13
|
1.00
|
1.00
|
1.13
|
Shanghai
|
1.04
|
1.06
|
1.00
|
1.10
|
1.08
|
0.72
|
0.99
|
0.77
|
1.03
|
1.16
|
1.00
|
1.20
|
Jiangsu
|
1.07
|
1.11
|
0.99
|
1.17
|
0.92
|
1.02
|
1.02
|
0.96
|
1.10
|
1.08
|
0.89
|
1.05
|
Zhejiang
|
1.03
|
1.04
|
0.99
|
1.06
|
0.87
|
0.99
|
1.01
|
0.87
|
1.06
|
0.99
|
0.96
|
1.01
|
Anhui
|
1.08
|
1.02
|
1.01
|
1.11
|
0.81
|
0.93
|
1.00
|
0.75
|
1.02
|
0.93
|
1.01
|
0.96
|
Fujian
|
1.00
|
1.07
|
1.00
|
1.06
|
0.81
|
0.97
|
1.03
|
0.80
|
1.06
|
1.02
|
1.00
|
1.08
|
Jiangxi
|
1.10
|
1.10
|
0.99
|
1.20
|
0.77
|
0.91
|
1.08
|
0.75
|
1.03
|
1.00
|
0.99
|
1.02
|
Shandong
|
0.91
|
0.97
|
1.09
|
0.95
|
0.88
|
1.06
|
1.01
|
0.93
|
1.08
|
1.11
|
0.99
|
1.19
|
Henan
|
0.88
|
0.95
|
1.15
|
0.96
|
0.81
|
0.77
|
1.25
|
0.79
|
1.03
|
1.09
|
1.02
|
1.14
|
Hubei
|
0.82
|
0.94
|
1.01
|
0.78
|
0.83
|
1.06
|
0.98
|
0.86
|
1.05
|
1.27
|
1.02
|
1.37
|
Hunan
|
0.82
|
0.97
|
1.02
|
0.81
|
0.80
|
0.97
|
1.00
|
0.78
|
1.02
|
1.10
|
0.99
|
1.11
|
Guangdong
|
0.95
|
1.00
|
1.00
|
0.95
|
0.83
|
1.00
|
1.00
|
0.83
|
0.97
|
1.00
|
1.00
|
0.97
|
Guangxi
|
0.84
|
0.95
|
1.08
|
0.87
|
0.80
|
1.02
|
0.99
|
0.81
|
0.99
|
0.94
|
1.06
|
0.99
|
Hainan
|
0.39
|
1.00
|
1.00
|
0.39
|
0.91
|
1.00
|
1.00
|
0.91
|
0.95
|
1.00
|
1.00
|
0.95
|
Chongqing
|
0.96
|
1.14
|
1.00
|
1.10
|
0.88
|
1.06
|
1.00
|
0.93
|
1.03
|
1.13
|
0.99
|
1.15
|
Sichuan
|
0.83
|
1.52
|
0.72
|
0.91
|
0.80
|
0.96
|
1.22
|
0.94
|
1.02
|
1.04
|
1.12
|
1.18
|
Guizhou
|
0.66
|
1.09
|
0.99
|
0.71
|
0.82
|
1.02
|
1.02
|
0.86
|
1.05
|
0.97
|
0.99
|
1.01
|
Yunnan
|
0.61
|
0.98
|
1.18
|
0.70
|
0.85
|
0.89
|
1.09
|
0.82
|
1.04
|
0.80
|
0.99
|
0.83
|
Shaanxi
|
0.96
|
1.01
|
1.00
|
0.97
|
0.88
|
1.05
|
0.98
|
0.91
|
1.08
|
1.00
|
1.00
|
1.08
|
Gansu
|
0.70
|
0.98
|
1.00
|
0.69
|
0.84
|
1.11
|
1.00
|
0.93
|
1.03
|
1.02
|
0.90
|
0.94
|
Qinghai
|
0.65
|
1.00
|
0.80
|
0.52
|
0.90
|
1.00
|
1.70
|
1.53
|
1.12
|
1.00
|
0.89
|
0.99
|
Ningxia
|
0.78
|
1.00
|
1.00
|
0.78
|
0.87
|
1.00
|
1.00
|
0.87
|
1.06
|
1.00
|
1.00
|
1.06
|
Xinjiang
|
0.66
|
1.00
|
1.00
|
0.66
|
0.88
|
1.00
|
1.00
|
0.88
|
1.09
|
1.00
|
1.00
|
1.09
|
mean
|
0.86
|
1.02
|
0.98
|
0.85
|
0.88
|
0.98
|
1.06
|
0.92
|
1.04
|
1.03
|
0.99
|
1.06
|
Table 3 showed that the total factor productivity of China's economic development performance under carbon emissions constraint from 2008 to 2012 was only 0.85, indicating that the dynamic performance level of China's economic development declined by 15%. From the decomposition value of total factor productivity index, technological change index, pure technical efficiency index and scale efficiency index are 0.86, 1.02 and 0.98, respectively, indicating that technological progress decreased by 14%, pure technical efficiency increased by 2%, and scale efficiency decreased by 2% from 2008 to 2012. Therefore, it is not difficult to infer that technological degradation is the key factor causing the decline of dynamic performance of economic development in this period. The total factor productivity index from 2012 to 2016 was 0.92, with an overall decrease of 8%. Among them, the technological change index, pure technical efficiency index and scale efficiency index were 0.88, 0.98 and 1.06, respectively, indicating that China's technological progress and pure technical efficiency in 2016 decreased by 12% and 2% compared with 2012. It can also be seen that technological degradation is still the main reason for the improvement of dynamic performance of economic development. The difference from 2008 to 2012 is that the impact of technological progress on dynamic performance is constantly weakened, while the driving effect of scale efficiency on dynamic performance is gradually enhanced. The total factor productivity index from 2016 to 2020 is 1.06, indicating that the overall dynamic performance level of China's economic development increased by 6%. The technological change index, pure technical efficiency index and scale efficiency index are 1.04, 1.03 and 0.99 respectively. It shows that China's technological progress and pure technical efficiency increased by 4% and 3%, while scale efficiency decreased by 1% from 2016 to 2020, which also reflects that technological progress and pure technical efficiency are the fundamental reasons for the improvement of dynamic performance. It should be noted that the impact of scale efficiency on the dynamic performance has changed from pulling to constraining.
Specifically, there were 8 provincial units with total factor productivity index greater than 1 from 2008 to 2012, namely Liaoning, Shanghai, Jiangsu, Zhejiang, Anhui, Fujian, Jiangxi and Chongqing, indicating that the dynamic performance level of economic development of the above provinces showed an overall upward trend, with an upward range of 3%-20%. The total factor productivity index of other provinces is less than 1, indicating that the dynamic performance of these provinces showed a downward trend, with a decline range of 3% to 61%. In addition, it can be found that the number of provincial units with technological change index, pure technical efficiency index and scale efficiency index all less than 1 is 21, 9 and 11 by observing the decomposition value of total factor productivity index, which can also reflect from the side that technological progress is the primary factor affecting the improvement of dynamic performance. The second is scale efficiency, pure technical efficiency has little influence on dynamic performance of economic development. From 2012 to 2016, there were only 4 provinces with total factor productivity index greater than 1, namely Beijing, Tianjin, Liaoning and Qinghai, indicating that the dynamic performance of economic development in Beijing, Tianjin, Liaoning and Qinghai showed an upward trend, and the total factor productivity index in other provinces were all less than 1, indicating that the dynamic performance of the vast majority of provincial units in 2016 showed a downward trend compared with 2012. In addition, the number of provincial units with technological change index, pure technological efficiency index and scale efficiency index all less than 1 from 2012 to 2016 is 27, 10 and 5, respectively, indicating that technological progress and scale efficiency have a greater impact on the dynamic performance of economic development, while pure technological efficiency has a smaller impact. The number of provincial units with total factor productivity index greater than 1 from 2016 to 2020 is 21, namely Beijing, Tianjin, Shanxi, Inner Mongolia, Jilin, Heilongjiang, Shanghai, Jiangsu, Zhejiang, Fujian, Jiangxi, Shandong, Henan, Hubei, Hunan, Chongqing, Sichuan, Guizhou, Shaanxi, Ningxia and Xinjiang. It shows that the dynamic performance of the above provinces shows an increasing trend, and the growth momentum of most provincial units is rapid, with the increase of more than 8%. According to the decomposition value of total factor productivity index, the number of provincial units with technological change index, pure technical efficiency index and scale efficiency index less than 1 is 5, 5 and 11, indicating that scale efficiency restricts the improvement of dynamic performance.
Analysis on the influencing factors of economic development performance
Selection of influencing factor evaluation factors
The spatial-temporal variation characteristics of China's economic development performance have been comprehensively analyzed above, but the deep-seated reasons for the differences in economic development performance remain unknown. Therefore, the paper selects specific indicators from urbanization, economic development, environmental regulation and openness as independent variables, takes static performance score as dependent variable, and uses Tobit model to explore the influencing factors of economic development performance. Table 4 shows the index system of influencing factors.
Table 4 Evaluation index system of influencing factors of economic development performance
Variable (abbreviated)
|
Meaning of variables
|
Specific indicators
|
Industrialization(IN)
|
Reflect the overall level of industrial development
|
Proportion of output value of secondary industry /%
|
Urbanization(UR)
|
Reflect the overall level of urban development
|
Proportion of Urban Population in total Population /%
|
Economic development(EC)
|
Reflect the overall level of economic development
|
Per capita GDP /yuan
|
Energy efficiency(EN)
|
Reflect the utilization efficiency of energy
|
Energy consumption per Unit GDP /ton of standard coal/ten thousand yuan
|
Vegetation coverage(VE)
|
Reflect the vegetation coverage
|
Forest coverage rate /%
|
Environmental regulation(ENV)
|
Reflect comprehensive environmental protection efforts
|
Per capita SO2 emissions /t
|
Motorization(MO)
|
Reflect the circulation of economic activities
|
Number of civil motor vehicles owned by ten thousand people
|
Openness(OP)
|
Reflecting the region's opening up
|
Total actual foreign investment /US $100 million
|
Government intervention(GO)
|
Reflects the government's ability to intervene in the economy
|
Government general budget revenue /100 million yuan
|
Analysis of Tobit model results
Firstly, the original data of independent variables were processed by logarithm in order to eliminate the interference caused by heteroscedasticity of data to the regression model. On this basis, Tobit model is constructed to analyze the influencing factors of static performance of China's economic development in 2008, 2012, 2016 and 2020 respectively, and the model results are shown in Table 5.
Table 5 Tobit analysis results of influencing factors of economic development performance
Variable
|
2008 year
|
2012 year
|
2016 year
|
2020 year
|
Coefficient
|
Prob
|
Coefficient
|
Prob
|
Coefficient
|
Prob
|
Coefficient
|
Prob
|
Ln IN
|
-0.0631
|
0.7431
|
-0.2454
|
0.1793
|
-0.1303
|
0.8253
|
-0.2080
|
0.1900
|
Ln UR
|
0.7536
|
0.0150**
|
0.8783
|
0.0278**
|
1.0350**
|
0.0220**
|
1.2894
|
0.0013***
|
Ln EC
|
-0.1400
|
0.6168
|
-0.4653
|
0.1020
|
-0.5733
|
0.0183**
|
-0.5124
|
0.0137**
|
Ln EN
|
0.0392
|
0.8210
|
-0.4540
|
0.0177**
|
-0.4947
|
0.0216**
|
-0.3285
|
0.0501*
|
Ln VE
|
-0.0062
|
0.8888
|
-0.0284
|
0.5067
|
-0.0939
|
0.0957*
|
-0.0995
|
0.0815*
|
Ln ENV
|
-0.1290
|
0.1789
|
0.1740
|
0.1060
|
0.0697
|
0.4128
|
0.1086
|
0.1174
|
Ln MO
|
-0.0663
|
0.5303
|
-0.0040
|
0.9741
|
0.2180
|
0.1307
|
0.1216
|
0.5357
|
Ln OP
|
-0.0730
|
0.1563
|
-0.0612
|
0.4220
|
-0.0726
|
0.2636
|
0.0268
|
0.4164
|
Ln GO
|
0.0235
|
0.6861
|
0.0095
|
0.9170
|
-0.0554
|
0.4300
|
-0.0472
|
0.3981
|
C
|
-0.3825
|
0.8451
|
4.3111
|
0.0469**
|
2.7715
|
0.1397
|
1.9672
|
0.1232
|
Log likelihood
|
18.4181
|
—
|
19.2355
|
—
|
20.6364
|
—
|
19.6211
|
—
|
AIC
|
-0.4945
|
—
|
-0.5490
|
—
|
-0.6424
|
—
|
-0.5747
|
—
|
SC
|
0.0192
|
—
|
-0.0353
|
—
|
-0.1287
|
—
|
-0.0610
|
—
|
Note: "***" means significant at 1% level, "**" means significant at 5% level, "*" means significant at the 10% level, “—”indicates that the item does not exist.
As can be seen from Table 5, the absolute values of AIC and SC in 2008, 2012, 2016 and 2020 models are all less than 1, and the Log likelihood value is high, so it can be inferred that the Tobit model constructed in this study is significantly effective according to the model judgment rules. Specifically, among the 9 independent variables in the 2008 model, only the urbanization level passed the significance test of the 5% critical level, while the rest of the independent variables failed the significance test of the 10% critical level, indicating that the urbanization level was the main factor affecting the economic development performance score under carbon emissions constraint. The influence of other factors on economic development performance is not obvious. In addition, the estimated value of urbanization level is 0.7536, indicating that urbanization level has a positive effect on economic development performance, which is reflected in the increase of urbanization rate by 1%, the score of economic development performance will increase by 0.7536 units. In 2012, both urbanization level and energy use efficiency passed the significance test of 5% critical value level, indicating that urbanization level and energy use efficiency are the main factors influencing the economic development performance score, while the other 7 independent variables have no obvious effect. The estimated parameter values of urbanization level and energy use efficiency are 0.8783 and -0.4540 respectively, indicating that urbanization level has a positive effect on economic development performance, while energy use efficiency has a negative effect. The relationship of numerical variation is as follows: if the urbanization rate increases by 1%, the economic development performance will increase by 0.8783 units, if the energy consumption per unit GDP increases by 1%, the economic development performance will decrease by 0.4540 units. The number of independent variables included in Tobit model increased from 2 in 2012 to 4 in 2016, including urbanization level, economic development level, energy efficiency and vegetation cover, which were significant at the critical value level of 5%, 5%, 5% and 10%, respectively. It indicates that urbanization level, economic development level, energy use efficiency and vegetation coverage in 2016 are the main factors influencing economic development performance, while motorization level, environmental regulation and government intervention are not very obvious. In addition, the estimated values of the above independent variables are 1.0350, -0.5733, -0.4947 and -0.0939 respectively, indicating that the economic development performance score will increase by 1.0350 units for every 1% increase in the urbanization rate. If per capita GDP increases by 1%, the score of economic development performance will decrease by 0.5733 units. If energy consumption per unit GDP increases by 1%, the economic development performance score will decrease by 0.4947 units. For every 1% increase in forest coverage, the economic development performance score will decrease by 0.0939 units. The number of independent variables included in the model in 2020 remains the same as that in 2016, and their effect directions on economic development performance score remain highly consistent. It should be noted that they have slight changes in significance level and effect intensity. Specifically, urbanization level is significant at the critical level of 1%, economic development level is significant at the critical level of 5%, and energy efficiency and vegetation cover status are significant at the critical level of 10%. In terms of parameter estimates, the impact of urbanization rate and forest coverage rate on economic development performance score gradually increased, while the impact of per capita GDP and energy consumption per unit GDP on economic development performance score weakened compared with 2016.
The urbanization level is always the main factor affecting economic development performance under carbon emissions constraint by comparing the estimation results of model parameters in four years, and the influence of the factor on economic development performance increases from 0.7536 in 2008 to 1.2894 in 2020, with an average annual increase of 0.0447 units. It is found that the urbanization process promotes the rapid agglomeration of factors such as population, industry, capital and technology to urban areas, which makes it possible to attract a large amount of investment in fixed assets and the inflow of rural labor force, this will help improve the efficiency of economical and intensive use of land and energy, change people's way of production, life and consumption, and introduce and promote advanced production technologies and energy-saving and environmental protection technologies. In this way, carbon emissions will decrease significantly while the economy develops rapidly, and the low-carbon and green transformation development goals of the national economy will be realized. The influence of economic development level, energy efficiency and vegetation cover on economic development performance gradually increases over time. Specifically, the economic development level has the greatest negative impact on the performance of economic development because the extensive mode of economic development is difficult to maintain under the background of increasing economic downward pressure and incomplete transformation and upgrading of industrial structure and energy structure. Unswervingly taking the road of sustainable development is the inevitable choice to achieve win-win economic development and environmental protection. Energy use efficiency has a great impact on economic development performance because low energy use efficiency will cause people to consume more energy in the production process. However, China's current energy structure is still dominated by fossil energy, so low energy efficiency will directly lead to more carbon emissions, which impedes the improvement of economic development performance. Vegetation coverage can also affect economic development performance. The reason is that green vegetation, especially forest, can use photosynthesis to convert carbon dioxide into living matter and release oxygen, which can reduce carbon dioxide emissions in the process of human activities and help to alleviate the environmental load faced by human activities. Therefore, the increase of vegetation coverage can improve economic development performance.