Multi-Period Double Difference Model

As this paper focuses on evaluating the advantages of implementing policies over multiple time periods and EDC is designated as a “quasi-natural experimental zone”, DID is chosen as the empirical model of this paper. DID can effectively eliminate possible individual fixed effects and time fixed effects and effectively control for potential other variables so as to derive more accurate causal effects. As far as the research content of this paper is concerned, the establishment of ecological civilization advance demonstration zones may have a positive promotion effect on EER. On the other hand, there may be a significant differential effect in the pilot cities in comparison with the cities that have not implemented this policy. Therefore, this paper presents DID as depicted in the subsequent equation:

Where i, t indicates city and year; EER_it is an explanatory variable indicating the degree of coordination between urban eco-efficiency and economic development coupling; Treat_i · Post_t is a dummy variable for EDC policy; β is the core concern coefficient, responding to the effect of EDC on the degree of coupling and coordination of EER_it effects; X_it is the control variable µ_i,γ_t sample city individual effects and time fixed effects, ε_it is a random perturbation term.

Coupled Evaluation Model

This paper constructs a coupling evaluation model of ecological efficiency and economic development. The EER is determined through a comprehensive assessment of various indicators. In this study, we employ AHP hierarchical analysis to establish the weighting for each characteristic indicator and assess the EER through a series of steps. As the sample areas are widely spread, the ecological basis and stock changes of the selected indicators have resource basis and governance differences. Using the AHP hierarchical analysis method, the problem can be qualitatively and quantitatively decomposed into multiple levels for processing, and the study starts from the national situation of China’s resource and environmental development, combines with the hotspot of future work on sustainable development, and is more practicable and scientifically significant.

(1) Integrated assessment modeling

An important prerequisite for conducting coupling measurement is to construct a comprehensive evaluation model, and the indicator values in this evaluation model need to be standardized to remove the influence of the scale on the indicators. Based on this, this paper adopts the method of extreme deviation to standardize the indicator data, as shown in Equations (2) and (3).

In the above equation, x_i' denotes the standardized value; max x_i', min x_i' denote the maximum and minimum values of indicator i, The comprehensive evaluation model is developed based on the indicator values and weights of each indicator characteristic, as illustrated in Equations (4) and (5).

In the above equation, F(x), F(y) represent the combined benefits of environmental efficiency and regional economic subsystems, respectively. a_i, b_i are the weights of the indicators in the environmental efficiency and economic development subsystems, respectively, which are derived from the AHP hierarchical analysis method. x_i', y_i' are the values of the indicators standardized using the extreme variance method.

(2) Measuring Coupling and Coupling Harmonization

EER can be used to illustrate the degree of mutual influence between the two subsystems of environment and economy [27]. Still, the results of its measurement may cause “pseudo” coupling results due to the low evaluation index of the subsystems and the similarity of the results. Therefore, this paper eliminates the misestimation due to the “pseudo” coupling results by introducing a comprehensive evaluation index and carries out the measurement of the level of coordination of the coupling of eco-efficiency and economic development so as to reflect the extent of coordinated development of the various subsystems. The process of measurement is shown in Equations (6), (7), and (8) [28].

In the above equation, K is the systematic coupling of environmental efficiency and economic development. K∈[0,1], K larger represents a higher degree of coupling and vice versa; T is the value of the combined evaluation index for environmental efficiency and economic development; and D is the coupled harmonization degree of environmental efficiency and regional development.

Explained Variable

EER as an explanatory variable. The selection of indicators for the system coordination ratio includes both the economic development subsystem [29], the environmental efficiency subsystem, in which the economic development system is measured in terms of the quality of economic development, the quality of the industrial structure, and the living standard of the people, and the environmental efficiency system is measured in terms of the utilization rate of waste and energy treatment and the level of ecological protection. This paper refers to the “National Ecological Civilization Construction Demonstration Zone Construction Program (Trial)” (2013) formulated by China’s government branch in conjunction with six ministries and commissions for the selection of indicators, it explains regional economy development quality is illustrated by the GDP per capital, the GDP growth rate of each prefecture-level city is indicative of the extent to which the secondary industry contributes to industrial pollution in the region, measures the people’s living standard in terms of the retail sales of consumer goods and the local financial expenditures; at the same time, solid waste recycling rate and sewage treatment rate are used to explain the utilization rate of waste energy treatment, and the sulfur dioxide emission, industrial soot emission and per capital green space rate are used to explain Urban eco-efficiency. The nature of each indicator is explained and weighted, as shown in Table 1.

Core Explanatory Variables (Civ)

The primary explanatory factor examined in this study is the application of EDC “Treat·Post”. If the city is EDC and the observation time is after realizing policy, the interaction term between Treat and Post takes the value of 1, and vice versa, it is zero.

Control Variable

(1) Refer to the literature on relevant EDC mechanisms of action, this paper controls the impact of the following variables on EER in each city:

(2) The total amount of urban GDP and economic development may result in a situation where “development is emphasized over ecology,” and therefore GDP development may have negative externalities on the ecosystem.

Science expenditure and the government’s increased investment in scientific research may promote the construction of the ecological civilization advance demonstration zone, which in turn will lead to green innovation and improve the efficiency of energy use.

(3) The actual amount of foreign investment may cause the transfer of highly polluting industries to China, resulting in environmental degradation, but also the transfer of high-tech industries to enhance the efficiency of China’s environmental utilization. The increase in foreign investment may also contribute to the enhancement of the comprehensive strength of China’s economic development and put more funds into the construction of ecological civilization.

(4) Domestic waste disposal rate: the rise in its value contributes to enhancing people’s quality of life, thus promoting the simultaneous advancement of the ecological economy.

(5) The park green space area is an important indicator of the ecological development of the city and the well-being of the people, and the expansion of the green space area represents the mutual promotion and coordinated development of the economy and the environment.

Parallel Trend Test

To further verify whether the implementation of the policy of EDC has had a significant impact on the explanatory variables, this paper carries out the test of the parallel trend for all the samples before and after the implementation of the policy. As shown in Fig. 1, the figure uses the policy implementation year (2014) as the policy point in time (d_0) and the year before the policy implementation as the base period (d_-1), and the event study method is used to test for parallel trends.

As can be seen from Fig. 1, none of its coefficient estimates were significant before the implementation of EDC. The study found that there was no significant difference between the cities in the control and experimental groups before the policy was implemented. In the year of policy implementation (d_0), the construction of ecological civilization’s first demonstration zone did not immediately play a policy effect; there is a lag. In the second year of policy implementation (d_1), the policy effect gradually appeared and has been there since then. There has been a clear upward trend between the policy cities and the control group of cities, appearing with obvious differences. The policy is confirmed to be effective and passes the parallel trend test.

Fig. 1.

Parallel trend test.

Placebo Test

As there may be potential influencing factors in the estimation results, the baseline regression may also be biased. To further validate the accuracy of the study findings, the paper conducted individual placebo tests on the research subject. That is, a random sample is drawn to observe the impact of EDC on EER and the extent to which policy impacts are coordinated. If the regression coefficient is centrally distributed around 0, i.e., the randomly selected cities will not be significantly affected by EDC, then it proves that the policy can be excluded from the potential interference influencing factors, and the regression results are credible. Otherwise, it is considered that there is an error. In this paper, 46 randomly selected city samples were selected as a „pseudo-experimental” group, randomly repeated 500 times, and the results shown in Fig. 2 were obtained. The figure indicates that the estimated coefficients of the pseudo-experimental group are centrally distributed around 0, and the results are not significant. In addition, the distribution of the baseline regression coefficients for the pseudo-experimental group is at the high point of the distribution of the spurious regression coefficients, and the true estimated coefficient (0.0519) is in the range of small probability events in the distribution plot. All the above results prove that the benchmark regression results of this paper are reliable, and the coupling and coordination degree of the ecological environment and economic development in EDC is not much affected by potential factors.

Fig. 2.

Placebo test.

Other Robustness Tests

(1) Replacement of explanatory variables [30–32]. Considering that the coupling and coordination degree of ecological environment and economic development integrates the two aspects of environment and economy, this paper replaces the explanatory variables with real GDP and non-green energy intensity, in which the non-green energy intensity is measured according to the city’s total tonnage of coal use and then taken as the logarithm of it. According to columns (1) and (2), it can be seen that after the replacement of the explanatory variables, the real GDP is positively significant at the level of 1%. It shows that after the implementation of the policy of ecological civilization advance demonstration zone, the pilot city insists on the synergistic development of economy and environment, abandons the situation of „wanting economy but not green, and strengthens investment in ecological management while the level of economic development is improving. The intensity of non-green energy use is negatively significant at the 10% level, indicating that the structural upgrading of the energy industry is being gradually implemented under the influence of the policy and that the dependence on non-green energy is gradually decreasing.

(2) Removing provincial capital cities. As the provincial capital city focuses on more high-quality resources in the province and is the leader of a province’s development, it may itself have a good ecological and economic foundation. This paper removes the provincial capital city from the sample after the analysis, and Column (3) of Table 4 shows that the multi-period double-difference coefficients of the pilot policy on the basis of the removal of the provincial capital city are positively significant at the level of 1 percent, which indicates that EDC does produce policy benefits for the pilot cities.

(3) Excluding other policy interference, Ecological civilization’s first demonstration zone implementation period, there are other environmental policies (such as a low-carbon pilot city) implementation of superposition [33]. In order to further exclude the impact of interference policies, this paper excludes the sample period implementation of low-carbon pilot policy cities and excludes other policy cities, The pilot policy of DID is still at the level of 1% positive and significant, indicating that the ecological civilization’s first demonstration zones have a robust effect.

(4) The counterfactual test. In order to further test whether EDC policy played a role during the policy implementation period and to exclude the progressive role brought about by the city’s development, the study assumes that the policy occurred in 2009. Pseudo-dummy variables are set, indicating that EDC plays a robust role in improving EER.

Endogeneity Test

There may be reverse causality in the bidirectional development between EER and EDC, and when the EER of a place continues to rise, it may stimulate investment efforts in ecological civilization construction. In order to circumvent the problem of endogeneity, the study introduces instrumental variables, and in order to satisfy the characteristics of exogenous and correlation of the instrumental variables, the study selects the instruments that have a strong correlation with the core explanatory variable EDC, but do not have a path effect on the interpreted variable EER variables.

To ensure the usability of the indicator, the study selects instrumental variables from a geographic perspective, and by measuring the degree of terrain undulation from Feng’s experience [34], the interaction term between the degree of terrain undulation and the time lag of one period is used as an instrumental variable for EDC, which can be used to more scientifically verify whether EDC works effectively in the enhancement of EER. On the one hand, geographic indicators are naturally formed and are the core indicators for assessing regional environmental suitability and resource and environmental carrying capacity, and areas with greater topographic relief are rich in ecological resources and have a better ecological foundation, which is closely related to the construction of ecological civilization. At the same time, the historical natural formation of terrain undulation is far away from the current stage of development and does not change with the passage of time, which will not have a direct impact on the explained variable EER in the experiment and can meet the basic conditions of instrumental variables. On the other hand, since there is a time-lag effect on the utility of the EDC generating mechanism and terrain undulation as fixed cross-section data that does not change, the study draws on Sun’s practice [35] of interacting the degree of terrain undulation with a time-lagged period in a 2SLS regression.

The results of the experiment are shown in the Table 5, where IV is the interaction term between the degree of relief and the time lag for one period. The experimental results show that in the first period, the coefficient of instrumental variable IV is significant and the LM statistic is significantly positive; the Cragg-Donald Wald F statistic of 167.308 is greater than the critical value at the 10% significance level, and the experimental results indicate that there is no weak instrumental variable problem. The second stage of the experiment shows that EDC has a significant promotion effect on EER, and the growth of EDC inputs is 0.1588 per unit of enhancement, which is a significant enhancement compared with the benchmark regression. All of the above analyses indicate that the instrumental variables selected in this paper are reasonable and effective, so the benchmark results have a certain degree of robustness.

Analysis of Urban Location Heterogeneity

China has a vast land area and suffers from unbalanced regional development, with a wide development gap between the eastern and western parts of the country. In terms of China’s current national conditions, the eastern part of the country has a superior geographic environment, a solid economic foundation, a good business environment, and is new and high-tech; the central part of the country is industrially developed, rich in human resources, and sufficiently motivated to enhance the ecological environment and economic development; while the western part of the country is relatively backward in development, with overexploitation of resources, and may be insufficient in its ability to enhance the degree of coordination of the coupling of the ecological environment and economic development. The results of the previous study divide all the samples into a whole, and the results may be affected by the radiation of the developed areas. In order to further refine the results, this paper divides the 213 sample cities into three regions of the center, east, and west based on the location of their geographical neighbors for the analysis of heterogeneity. This paper constructs location dummy variables (area), assigns values to the Central and Western Cities for regression, respectively (each regression of the research point city assigns the value of 1), and brings the location dummy variables into the model after interacting with the policy multi-period double-difference terms to get the results in columns (2)-(4) of Table 5, it shows that the estimated coefficients of the interaction terms of the policy of EDC for the cities in the central and western regions are significant at the levels of 1% and 5%, respectively, and they are significantly affected by the policy of ecological civilization construction. In contrast, the interaction term coefficient of EDC policy for eastern cities is significant at the 10% level, and the coefficient of the core explanatory variable is also smaller than that of the west and similar to that of the central part of the country. The explanation for this phenomenon may be due to the fact that the eastern region itself has a good economic and ecological foundation, with less room for progress, and is not significantly affected by the construction of ecological civilization.

Heterogeneity Analysis of City Types

(1) Distinguish according to resource abundance.

The degree of urban resource abundance will produce different responses to the ecological construction according to the degree of urban dependence on local resources for the division, the construction of the dummy variable resource, the strong degree of resource dependence of the city assigned a value of 1, and the non-resource-oriented cities assigned a value of 0. After interacting the dummy variables with the policy multi-period double-difference term and bringing them into the model, which show that the estimated coefficient of the interaction term of resource-dependent cities is significant at the level of 1%, which indicates that EDC has a significant effect on the promotion of the optimization and rational use of resources, and it has a positive externality on resource-dependent cities, which significantly enhances the degree of coordination of the coupling of the ecological environment and the economic development of the resource-dependent cities, whereas for the cities with weaker resource-dependence, the change of the level of resources doesn’t affect the development of the cities directly, and therefore, the effect of EDC on this is relatively minor.

(2) Differentiation according to the degree of technological endowment

Cities with a high degree of technological endowment are those with advanced science and technology industries and innovation capabilities. These cities usually have high-level scientific research institutions, a large number of high-tech enterprises, and human resources, and are able to attract and cultivate high-quality scientific and technological talents to promote scientific and technological innovation and industrial development. 213 cities are divided into two categories; cities with a high degree of technology endowment and cities with a low degree of technology endowment. They construct the dummy variable technology, assign a value of 1 to the cities with a high degree of technology endowment, and assign a value of 0 to the cities with a low degree of technology endowment. They interact the dummy variable with the policy multi-period double-difference term, bring it into the model, and get the regression results of the (5) columns in Table 5. The results show that cities with higher degrees of technological endowment are significantly affected by the policy of ecological civilization advance demonstration zone, which indicates that under the influence of the policy of ecological civilization advance demonstration zone, cities with a high degree of technological endowment can be effectively pushed to carry out the industrial transformation, technological innovation, and reduce the proportion of sewage-emitting industries, so as to stimulate the efficacy of promoting the degree of coordination of the coupling between the ecological environment and the economic development, whereas, for the cities with an insufficient technological endowment, the response to the influence of the policy is obviously lagging behind.