3.1.Measurement method of green innovation of regional enterprises
The entropy method measures the weight of each indicator layer in the composite system. If there are p provinces (cities), m indicators y years, Xɑβθ is the value of the βth indicator of the ɑth province in the θth year(ɑ = 1, 2, 3. . p; β = 1, 2, 3. . m, θ = 1, 2, 3. . y)。The formulae are as follows:
1. Indicator standardisation: Different indicators have different scales and units and therefore need to be standardised.
If the indicator is positive,
$$\begin{array}{c} {\text{Y}}_{\text{ɑ}{\beta }{\theta }}=\raisebox{1ex}{$\left({\text{X}}_{\text{ɑ}{\beta }{\theta }}-\text{min}\left({\text{X}}_{\text{ɑ}{\beta }{\theta }}\right)\right)$}\!\left/ \!\raisebox{-1ex}{$\left({\text{m}\text{a}\text{x}(\text{X}}_{\text{ɑ}{\beta }{\theta }})-\text{min}\left({\text{X}}_{\text{ɑ}{\beta }{\theta }}\right)\right)$}\right. \left(1\right) \end{array}$$
If the indicator is negative
$${\text{Y}}_{\text{ɑ}{\beta }{\theta }}=\raisebox{1ex}{$\left({\text{m}\text{a}\text{x}(\text{X}}_{\text{ɑ}{\beta }{\theta }})-{\text{X}}_{\text{ɑ}{\beta }{\theta }}\right)$}\!\left/ \!\raisebox{-1ex}{$\left({\text{m}\text{a}\text{x}(\text{X}}_{\text{ɑ}{\beta }{\theta }})-\text{m}\text{i}\text{n}({\text{X}}_{\text{ɑ}{\beta }{\theta }})\right)$}\right. \left(2\right)$$
where min represents the minimum value and max represents the maximum value.
2. Calculation of characteristic proportions or contributions Zɑβθ。
$${\text{Z}}_{\text{ɑ}{\beta }{\theta }=}\frac{{\text{Y}}_{\text{ɑ}{\beta }{\theta }}}{\sum _{1}^{\text{p}}\sum _{1}^{\text{y}}{\text{Y}}_{\text{ɑ}{\beta }{\theta }}},\text{ɑ}=\text{1,2},3\dots \dots \text{p},{\theta }=1, 2, 3 . . . \text{y} \left(3\right)$$
3. Calculation of entropy Eβ:
$$\begin{array}{c}{\text{E}}_{{\beta }=}K\sum _{\text{a}=1}^{\text{p}}\sum _{{\theta }=1}^{\text{y}}{\text{Z}}_{\text{ɑ}{\beta }{\theta }}\text{ln}\left({\text{Z}}_{\text{ɑ}{\beta }{\theta }}\right),K=-\frac{1}{\text{ln}\left(\text{y}\text{p}\right)},{0\le \text{E}}_{{\beta }}\le 1 (4)\end{array}$$
4. Calculate the information utility value of the βth indicator:
$$\begin{array}{c}{\text{G}}_{{\beta }}=1-{\text{E}}_{{\beta }} \left(5\right)\end{array}$$
Determining the weights of evaluation indicators Wβ:
$$\begin{array}{c}{\text{W}}_{{\beta }}=\frac{{\text{G}}_{{\beta }}}{\sum _{1}^{\text{m}}{\text{G}}_{{\beta }}} ,\beta = 1, 2, 3 . . . . . . m\#\left(6\right) \end{array}$$
3.2 Constructing a Comprehensive Green Innovation Indicator System for Regional Enterprises
Based on the panel data of each province in China, this paper refers to the research of Kang Yumei et al. and constructs the comprehensive green technology indicators from five dimensions, namely, the amount of patents granted, the share of technology market turnover, the government's investment in green technology and innovation, the urban unemployment rate, and the energy utilisation rate.(Yumei & Chengxing, 2023)。The number of patents granted reflects how active a region or organisation is in technology development and innovation. A higher number of granted patents may imply that the region or organisation has a high level of innovative capacity and dynamism in green technologies. This lays the foundation for further development of green technologies. The ratio of technology market transactions reflects the activity of the technology market and the efficiency of the transformation of scientific and technological achievements. A higher technology market turnover ratio means that more green technology achievements have been applied and promoted. It helps to popularise and deepen green technology. Government expenditure on science and technology as a proportion of finance: This indicator reflects the importance the government attaches to science and technology and green technology. Government investment in science and technology and green technology can promote relevant research and development, and promote the progress and innovation of green technology. Urban registered unemployment rate can reflect the economic situation and employment. A healthy economic environment can provide more employment opportunities, thus attracting more talents to invest in green technology R&D and innovation.Electricity consumption per unit of GDP is an important criterion for measuring the efficiency of energy use in a region or organisation. By reducing electricity consumption per unit of GDP, energy consumption and environmental pollution can be reduced, and green and low-carbon development can be promoted, which is one of the important goals of green technology innovation. As shown in Table 1.
Table 1
Comprehensive evaluation index system of green innovation index for regional enterprises
Objective level
|
Criteria level
|
Indicator layer
|
Weights
|
Properties
|
Comprehensive Green Technology Indicators for Regional Enterprises
|
Patent grants
|
Patent Applications and Authorisations(item)/ Year-end Resident Population (10,000)
|
0.3331
|
Forward
|
Percentage of Technology Market Transactions
|
Technology market turnover (billion yuan) / GDP (billion yuan)
|
0.3205
|
Forward
|
Government Investment in Science and Technology Innovation
|
Share of Science and Technology Expenditure in Fiscal Expenditure
|
0.2303
|
Forward
|
Urban Unemployment Rate
|
Urban registered unemployment rate
|
0.0716
|
Forward
|
Energy Utilisation Rate
|
Electicity consumption per unit of GDP
|
0.0443
|
Negative
|
Based on the entropy method and the comprehensive index of green technology innovation, this paper measures the development level of green technology innovation in each region of China from 2012 to 2021, as shown in Fig. 3. It can be found from 2012 to 2021. The green technology innovation score of the eastern region ranges from 0.19 to 0.35, that of the central region ranges from 0.09 to 0.26, while that of the western region ranges from 0.09 to 0.17. The eastern region, with its higher level of economic development and better infrastructure, has always scored at a higher level in terms of green technological innovation. As shown in Fig. 3, from the overall trend, China's green technology innovation index shows a gradual upward and steady development trend. The eastern region has been in the lead. However, it is worth noting that the inter-regional green STI gap between the central and western regions is gradually widening. This suggests that the western region may need more attention and investment in the development of green STI in China in order to narrow the regional gap.
3.3 Calibration
In this paper, based on Boolean algebra theory and previous studies, the data were calibrated precisely to ensure the consistency and coverage of the analysis. The direct calibration method is used, and the 95% quartile, 50% quartile and 5% quartile are set as calibration anchor points, and the specific results are shown in Table 2.
Table 2
|
Variable name
|
Fully affiliated
|
Intersections
|
Completely unaffiliated
|
Result Variables
|
Green innovation of local enterprises (Y)
|
0.469
|
0.139
|
0.068
|
Conditional variables
|
Per capita gross regional product(A)
|
116664
|
50242
|
28622
|
Percentage of government expenditure on social security(B)
|
45.215
|
40.4
|
33.88
|
Urban Green Coverage Rate(C)
|
0.194
|
0.134
|
0.084
|
Level of aging population(D)
|
0.159
|
0.108
|
0.072
|
Tourist demand for green innovation(E)
|
5.511
|
0.995
|
0.01
|
Number of foreign-invested enterprises(F)
|
79639.850
|
6698
|
737.8
|
Demand for innovation by regional service industries(G)
|
37660.94
|
9850.3
|
1542.95
|
3.4 Necessity Analysis of Individual Conditions
According to the set theory of Boolean algebra and QCA design principles and applications, it is known that the smaller the adjustment distance of QCA panel data, the higher the consistency accuracy. However, the adjustment distance is not clearly defined in statistics.QCA experimental analysis needs to consider the data size and data inclusion, so the median value of the adjustment distance used in this experiment is 0.3.As shown in Table 3, if per capita GDP (A), the proportion of government social security expenditures (B), the percentage of urban green coverage (C), the level of the aging population (D), the demand of tourists for green innovations (E ), number of foreign-invested enterprises (F), and regional enterprises' demand for innovation (G), the seven indicators have an adjusted distance to green innovation greater than 0.3, and coverage less than 0.5 requires researchers to further explore the necessity.
Table 3
Analysis of the necessary conditions
variant
|
High level of local business green innovation (Y).
|
Low level of local business green innovation (~ Y).
|
Aggregate Consistency
|
Aggregate coverage
|
Inter-group consistency
|
Intra-group consistency
|
Aggregate Consistency
|
Aggregate coverage
|
Inter-group consistency
|
Intra-group consistency
|
A
|
0.85
|
0.84
|
0.10
|
0.24
|
0.47
|
0.56
|
0.52
|
0.52
|
~A
|
0.56
|
0.46
|
0.18
|
0.45
|
0.87
|
0.88
|
0.11
|
0.21
|
B
|
0.75
|
0.69
|
0.06
|
0.38
|
0.58
|
0.65
|
0.28
|
0.47
|
~B
|
0.62
|
0.55
|
0.18
|
0.40
|
0.73
|
0.78
|
0.09
|
0.40
|
C
|
0.65
|
0.61
|
0.20
|
0.42
|
0.63
|
0.72
|
0.29
|
0.39
|
~C
|
0.71
|
0.61
|
0.25
|
0.35
|
0.86
|
0.70
|
0.15
|
0.39
|
D
|
0.77
|
0.71
|
0.20
|
0.30
|
0.54
|
0.61
|
0.36
|
0.49
|
~D
|
0.85
|
0.51
|
0.37
|
0.41
|
0.75
|
0.80
|
0.19
|
0.32
|
E
|
0.61
|
0.66
|
0.50
|
0.41
|
0.51
|
0.67
|
0.45
|
0.56
|
~E
|
0.69
|
0.54
|
0.29
|
0.32
|
0.74
|
0.70
|
0.19
|
0.34
|
F
|
0.77
|
0.79
|
0.06
|
0.37
|
0.88
|
0.61
|
0.15
|
0.60
|
~F
|
0.62
|
0.49
|
0.04
|
0.43
|
0.83
|
0.81
|
0.08
|
0.30
|
G
|
0.78
|
0.75
|
0.07
|
0.40
|
0.50
|
0.60
|
0.27
|
0.55
|
~G
|
0.85
|
0.49
|
0.20
|
0.47
|
0.79
|
0.81
|
0.07
|
0.35
|
average
|
0.7
|
0.6
|
0.2
|
0.4
|
0.7
|
0.7
|
0.2
|
0.4
|
By analysing the intergroup consistency and coverage of the corresponding variables (as shown in Tables 3 and 4). There are the following findings: firstly, in the process of analysing the necessary conditions, we did not find that any single factor can constitute the necessary conditions for green innovation of local enterprises alone. This means that green innovation of local enterprises is a complex process that requires multiple factors to work together to achieve it. Secondly, in the inter-group data with an adjusted distance greater than 0.3, we observe that in cases a, b, and c, the level of consistency across years does not reach 0.9. therefore the necessary relationship is not satisfied. Meanwhile, by plotting the scatterplot of coverage and consistency, we find that the coverage is mainly concentrated on the right y-axis, which passes the test of the non-essential condition. However, consistency does not pass the test of the non-essential condition. This further suggests that these factors may play a role in the green innovation process. However, they are not decisive and have their own research value.
Table 4
Data between groups with adjusted distances greater than 0.3
Situation.
|
Causal combination situations
|
|
2012
|
2013
|
2014
|
2015
|
2016
|
2017
|
2018
|
2019
|
2020
|
2021
|
a
|
~A and Y
|
Intergroup consistency
|
0.68
|
0.68
|
0.66
|
0.63
|
0.59
|
0.58
|
0.54
|
0.50
|
0.47
|
0.41
|
Intergroup coverage
|
0.26
|
0.32
|
0.34
|
0.37
|
0.39
|
0.49
|
0.63
|
0.70
|
0.75
|
0.91
|
b
|
~F and Y
|
Intergroup consistency
|
0.62
|
0.61
|
0.64
|
0.64
|
0.63
|
0.63
|
0.65
|
0.61
|
0.59
|
0.57
|
Intergroup coverage
|
0.31
|
0.37
|
0.38
|
0.41
|
0.44
|
0.50
|
0.59
|
0.62
|
0.67
|
0.74
|
c
|
~G and Y
|
Intergroup consistency
|
0.72
|
0.73
|
0.71
|
0.68
|
0.61
|
0.54
|
0.50
|
0.50
|
0.48
|
0.44
|
Intergroup coverage
|
0.31
|
0.38
|
0.40
|
0.43
|
0.46
|
0.53
|
0.62
|
0.64
|
0.66
|
0.74
|
3.5 Configuration analysis results
Table 5
Configuration truth table
Conditional variables
|
parameterisation1
|
Parameterisation2
|
Parameterisation3
|
Gross regional product per capita (A)
|
⊗
|
⊗
|
●
|
Percentage of government expenditure on social security(B)
|
●
|
●
|
●
|
Urban Green Coverage Rate(C)
|
⊗
|
⊗
|
●
|
Level of aging population(D)
|
●
|
|
●
|
Tourist demand for green innovation(E)
|
|
●
|
⊗
|
Number of foreign-invested enterprises(F)
|
●
|
●
|
●
|
Demand for innovation by regional service industries(G)
|
●
|
●
|
●
|
Consistency
|
0.838
|
0.827
|
0.845
|
Original Coverage
|
0.402
|
0.411
|
0.366
|
Unique Coverage
|
0.019
|
0.08
|
0.095
|
Inter-group consistency adjusted distance
|
0.011
|
0.012
|
0.012
|
Intra-group consistency-adjusted distance
|
0.026
|
0.029
|
0.024
|
Overall PRI
|
0.611
|
|
|
Overall Consistency
|
0.827
|
|
|
Overall Coverage
|
0.402
|
|
|
Note: ● and ⊗ indicate presence and absence of core; blank indicates that presence and absence are also possible. |
Table 5 shows that the consistency of grouping 1, grouping 2, and grouping 3 is 0.828, 0.827, and 0.845, respectively. The overall consistency is greater than 0.75. And the adjustment distance between intra-group and inter-group for individual grouping is less than 0.3. It shows that the aggregated consistency has a better explanatory strength. These three groupings can be regarded as sufficient conditions affecting the generation of sustainable green innovation in local enterprises. From the study of group state 1, we observe the influence of different factors on the green innovation of enterprises. Cohort 1 shows that economic drivers such as the number of foreign-invested firms and the demand for innovation in the regional service sector, as well as socio-environmental factors such as a high level of government social security expenditure and the level of ageing, are the main drivers of green innovation in firms, while market demand has a limited impact. Configuration 2 further emphasises the importance of high levels of foreign investment and service sector innovation demand. At the same time market demand for tourists' preference for green begins to emerge. By configuration 3, economic, social and market demand factors show a more balanced state. The economic drivers are GDP per capita, high level of foreign investment, and high level of regional demand for innovation in the service sector. The social environment factor is dominated by the high level of government social security expenditure share. Market demand is dominated by the urban greening coverage rate.
These findings suggest that, in the context of China's geographical resource differences, localities should combine their own characteristics to achieve factor linkages in order to promote local firms' green innovation. It is worth noting that the multidimensional linkage model demonstrated in Grouping 3, although the study shows that the green innovation of local enterprises is influenced by multiple factors such as economic drive, social environment and market demand. However, this multidimensional linkage model still needs further in-depth exploration. The key lies in how enterprises balance supply and demand to achieve multi-dimensional power. Only by comprehensively considering economic, social and market demands can enterprises formulate a more reasonable and effective green innovation strategy. This is not only the key to enhance the competitiveness of enterprises, but also the way to realise the sustainable development of green innovation. Therefore, future research should pay more attention to the balance and synergy of enterprises under the effect of multidimensional factors, so as to promote the in-depth development of green innovation.
3.5 Between and within group results
It was found that the adjusted distance of intergroup consistency for all 3 groupings was not greater than 0.3, indicating that there was no significant time effect. Further examination of its temporal changes revealed that the consistency levels of the 3 groupings showed a decline from 2012–2016. However, they collectively showed a period of rapid growth in 2016–2021. As shown in Fig. 4. Among them, the fastest growth rate of intergroup consistency for group state 3 grows from 0.84 to 1.00. The reason for this is that our government intervention plays a crucial role. Checking government websites and local service platforms found that 60% more documents were released in 2016 to promote green development compared to 2015. Meanwhile, the government sends out strong intervention signals In terms of policy intervention: the Chinese government has introduced a series of policies to encourage green innovation and sustainable development, such as providing financial subsidies, tax incentives, and loan facilitation, in order to incentivise enterprises to increase their research and development in green technologies. Regulatory constraints: The government has strengthened the formulation and enforcement of environmental protection regulations, imposing strict limits on pollution emissions and energy consumption. Government in green procurement: As one of the largest consumers, the government has given priority to green products and services to encourage enterprises to actively develop green products and enhance their green innovation capability.
Table 5
|
Eastern China
|
Central China
|
Western China.
|
Configuration 1
|
0.56
|
0.62
|
0.55
|
Configuration 2
|
0.38
|
0.46
|
0.38
|
Configuration 3
|
0.39
|
0.46
|
0.38
|
The range of intra- and inter-group consistency adjustment distances is almost the same, and the intra-group consistency adjustment distance is not greater than 0.3. The variability in the distribution of geographic coverage of the grouping models revealed in Table 5. The explained cases of group states 2 and 3 are mainly concentrated in the central region, which may stem from the unique resource conditions and policy environment in these regions. However, histogram 1 shows stronger explanatory power in East, Central and West China, with a coverage of more than 0.5, indicating its universality. This shows that firms in different regions may be affected by different factors when facing green innovation and sustainability challenges. Such geographical differences may stem from the diversity of regional levels of economic development, market demand, resource distribution and policy orientations. For policy makers and entrepreneurs, an in-depth understanding of the geographical characteristics of the model can help develop more targeted strategies and measures to promote green transformation and sustainable development of enterprises.
3.6 Consistent prediction of sustainable development of green innovation of local enterprises
Using MATLAB, the article constructed a consistency prediction model for the sustainable development of green innovation in local enterprises, as shown in Fig. 5. The root mean square error (RMSE) of the model is 0.9, a value that indicates that the predictive accuracy of the model can be applied to consistency prediction.RMSE is a commonly used metric for assessing the predictive ability of a model, which measures the magnitude of the model's error by calculating the mean of the squared difference between the predicted value and the actual value. A lower RMSE value means that the model has a higher prediction accuracy. From the trend of the model's prediction graph, it can be observed that the overall trend of the consistency index of the sustainable development of green innovation of local enterprises shows a gradual decrease followed by a rapid growth trend. According to the prediction trend, China's progress in green technological innovation is gradually accelerating, and the consistency index of enterprises' sustainable development of green innovation will be maintained around 0.90.