Nexus Between Energy Policy and Environmental Performance in China: 1 the Moderating Role of Renewable Energy Patents. 2

13 The socioeconomic and environmental considerations of energy production have become 14 crucial due to the increasing complexity of the relationship between energy and the 15 environment. In this context, this study aims to develop possible mechanisms for perspectives 16 of energy policy on environmental by exploring the mediating role of renewable energy patents. 17 The study used a non-radial data envelopment analysis (DEA) model and panel data model for 18 30 Chinese provinces by taking the panel data from 2010 to 2017. The results show that the 19 overall environmental performance index (EPI) of Chinese areas is improved by 9.88% from 20 2010 to 2017. Further, The econometric model findings offer evidence that provincial 21 renewable energy policies and emission reduction policies positively impact the enhancement 22 of EPI. The results also show that the P values of the single-threshold model and the double- 23 threshold model both passed the 1% significance test, so it can be concluded that there is a 24 double-threshold effect. Finally, the research findings posed several policy implications based 25 on the research findings.


Introduction
In the current ages, various researches are dedicated to measuring, investigating, and 30 enhancing energy efficiency. At that time, the environmental problem is a severe issue in the 31 world. Mainly the question of the world's environmental problems due to the greenhouse 32 emissions in which carbon-dioxide (co2), which generally connected to the blazing of fossil 33 fuels. However, this process also becomes the source of wasteful usage of natural resources 34 and significant environmental issues. Today the country that consumes considerable energy 35 will also become the source of highly intense carbon dioxide (Ozturk and Acaravci, 2010), 36  and (Anser et al., 2020). The government must shape a new planned 37 objective that helps save natural resources, develop a healthy environment for enhancing 38 energy efficiency and defending territory, and attain environmental development. Some 39 countries are already declared that they reduce their co2 releases per unit of GDP by 40% to 40 45% (Hou et al., 2019) and (Usman et al., 2021). At present, all nations are already facing 41 enormous challenges of environmental issues. In this situation, these countries must enhance 42 their energy efficiency and carefully think about environmental restrictions to lower their 43 energy utilization and bring down environmental pollution , (Öztürk and 44 Altınok, 2021). There are three types of index in which thermodynamic index, physical-based 45 index, and currency-based indicators index. With index, we usually measure energy efficiency. 46 The combined manufacturing procedure usually used energy such as natural gas, oil, and coal, 47 etc. In this, labor and capital are also used to create unique productivity, e.g., GDP and 48 unfortunate productivity, including emissions of pollutants (CO2) and (SO2) (Iqbal et al., 49 2019b). The environmental efficiency must not be unnoticed to deliver the more Similar type 50 of Efficiency Mark. 51 Industries across the world consume more than one-third of the global energy. The 52 proportion of C02 emission is slightly higher in this regard. China's development model relies 53 heavily on the industry, so local infrastructure development, manufacturing of export-oriented 54 Since fossil energy is considered the main factor in global warming, emphasizing energy-70 efficient production and distribution processes may be the key to solving this hazardous issue 71 (Meng et al., 2020). Economic developments across the nation are correlated with energy 72 intensity, carbon emission, and global warming issues. Therefore, businesses and governments 73 should care about the human and wildlife, climate, and environmental aspects while setting up 74 their respective growth strategies. In this connection, the green movement initiation by 75 adopting green technology solutions for industrial production and distribution can be a useful 76 undertaking. A robust energy and carbon management and control, if possible, with a strict 77 regulatory framework and better energy policy can also improve the environmental quality to 78 a great extent. 79 the existing studies how different types of regional-level energy policies affect environmental 182 performance. Therefore, investigating the effect of other regional level energy policies a timely 183 attempt for ensuring an environmentally sustainable developed economy in China. 184  (Inman et al., 2006). Efficiency, in this case, has been defined by several 189 scholars. In this study, we consider Farrell's (1957) (Farrell, 1957) definition of efficiency, 190 which was drawn from the study of (Koopmans 195) (Koopmann, 1951) to describe the 191 measure of efficiency that constitutes multiple inputs. Farrell (1957) (Farrell, 1957) states that 192 two components form an organization's efficiency: technical efficiency and allocative 193 efficiency. In an input-oriented efficiency measurement, technical efficiency refers to the ratio 194 of optimal input to the actual information. For an output-oriented efficiency measurement, it 195 relates to the rate of real output to the optimal production. 196

Methodology
On the other hand, allocative efficiency manifests an organization's ability to utilize its 197 inputs optimally in respect to its prices and technology. Based on the objective of  Making Units (DMU), production frontier or cost frontier is used to determine the optimal 199 input and optimal output. Two different methods are recommended in the literature in this 200 regard: parametric and non-parametric approaches. A functional plan is assigned for the 201 frontier in the parametric approach, but no preceding specification is applied for the non-202 parametric approach's border. (Charnes et al., 1978) followed the non-parametric process to 203 develop the DEA model for measuring the single DMU's efficiency for the first time. They 204 designed an input-oriented DEA model assuming a constant return to scale. However, the 205 following studies of the DEA model considered different assumptions set. For example, 206 (Banker, 1984) proposed variable returns to scale (VRS) and used mathematical programming 207 in their DEA model to generate a linear best practice frontier that relies on experimental input-208 output data. This new DEA approach received wider acceptance for measuring the efficiency 209 of different DMUs across industries or countries. 210 Assume there are n DMUs and individually signifies an administration zone. Separately, 211 DMU's non-energy input and L energy input produce the predictable cost Output or low output 212 of K. DMUs helps make as much wanted production as possible and spent the resources in less 213 amount. Therefore, in the usual methods, the decreasing of pollutants is not allowable. This 214 trouble can be resolved using different methods like using the opposite of worse output, bad 215 behavior output as input, and statistically renovating the unwanted result into the desired 216 outcome. In the study of energy and environmental efficiency, low production is mostly 217 produced by fossil fuels through manufacture, which must be minimized if we use energy in 218 less amount. 219 b kj = θ k b b kj , k = 222 1, . . . , k , λ j, S i x− ,S l e− , S r y+ ≥ 0, for all j, i, l, r, 223 Consider that method decreases the unwanted output and possible and a certain level of 224 ideal output and non-energy input. For sections between 0 and 1, the energy and environmental 225 efficiency index is θ, the superior the index, well the region's performance in terms of reducing 226 pollutant discharges and energy saving. The corresponding part is measured to be energy and 227 environmentally efficient. It cannot decrease its pollutant releases and energy consumption If 228 EPI1 = 1 (θ = 1) is zero, but if the EPI1 < 1 (θ < 1) are not zero, then the corresponding region 229 is environmentally inefficient, and can decrease energy utilization and pollutant discharges. decision-makers can give. We will use this method to estimate the total-factor energy and 242 environmental efficiency of various areas because this model has a developed perceptive power 243 than the 1st model. 244 In this research, years between 2000 to 2008, we make a strategy to compute the energy 245 and environmental efficiency in various areas. This forceful assessment can give us data about 246 efficiency variations. However, It is more important and significant to discovering energy and 247 the environment by putting on DEA window analysis to increase efficiency. The DEA window 248 analysis method is used to develop time-varying data and cross-section data to compute 249 dynamic properties. This method works by affecting medians and creates efficiency measures 250 after this, treating every DMU as an individual unit at various times. Therefore, we find the 251 environmental efficiency of various areas of different ages through overlapping windows using 252 this technology. 253 During possible measurement of efficiency, it has been seen that the width of the window 254 has tended to yield 3or 4 periods of time. This paper considers (w=3) a window with three 255 widths for attaining consistent environment and energy efficiency results. For such purpose, 256 the first three years, 2010 to 2012, have been utilized for the first window. After that, we will 257 more for a further one-year window and release base year while adding the next one, and this 258 procedure will continue till the last window is installed. Thus, radial and non-radial 259 environmental energy efficiency (EPI1 &EPI2) of each underlined province can be attained by 260 Appling DEA window analysis. 261

262
In the above equation, we used three inputs labor, capital and energy use, one good output 263 provincial Gros regional product which used as proxy for GDP and one bad output CO2 264 emissions for environmental performance index over the period of 2010 to 2017.    The following equation is constructed to find the relationship between environmental 334 regulation and total factor energy efficiency: 335 EPI = α + βEPI , −1 + γenergy_policy + θ + + +

Empirical results and discussions
(3) 336 In this equation, represents the intercept and , and are coefficients to be estimated. 337 energy_policy is the independent variable, that is, the vector that represents the energy policy.    Note: Standard errors are in parentheses (). ***= 1% significant level; **= 5% significant 380 level and *=10% significant level 381 Table 4  The results show the same scenario as the baseline regression as it is seen that there is the 413 existence of statistically significant EmRP over EPI, and the control variables also consistent 414 with the estimated coefficients. However, for the case of RePo, the coefficients (0.0605) of 415 lagged variables are found to be positive and statistically significant at 1% level, proving that 416 environmental performance is affected by the lag terms (table 4). It refers to the fact that in 417 the case of a rigorous CCER practice, the influence of "innovation offset" is more powerful 418 than the result of the "compliance cost" effect in the long run. This proposition matches with 419 the findings of Guo and Yuan (2020). They argue that taking the lagged variables instead of 420 current variables might increase to chances to generate a positive and significant effect on 421 energy efficiency.  Note: ***/1% significant level; **/5 significant level and */10 significant level (4) 452 In this model above, C is the estimated threshold value, and I(·) is the symptomatic function, 453 which will be true if the corresponding condition is equal to 1 and false if the value is 0. The 454 test results might come up with the presence of multiple thresholds, which can further be 455 stretched to double and numerous threshold models from the base single threshold model. 456

Analysis of Threshold Regression Test 457
We first checked the number of thresholds to perform threshold regression analysis. In 458 this study, we used Hansen's threshold panel model, used bootstrap technology, and repeated 459 it 500 times to test the threshold. We found that the impact of pollution mitigation policies and 460 clean energy policies on the environmental efficiency index has a major dual-threshold effect, 461 in which energy policy is the threshold component. The results of the importance assessment 462 are summarized in Table 6 are the corresponding 95% confidence intervals (table 7). 471  Figure 3 shows the result of the likelihood ratio (LR) function. The likelihood ratio (LR) 475 function of the dual-threshold model is used to test the consistency of the threshold estimate to 476 better understand the authenticity and confidence interval of the threshold estimate. When the 477 LR value is 0, the regional environmental performance index is the threshold estimate. As 478 shown in Figure   period. Provinces with an EPI higher than 0.438 also showed a similar steady growth trend. 499 This is happened due to the level of regional economic and technological development 500 continued to increase , from 9.7% to 16.1%, and the regional innovation 501 system continued to shift to the best level. However, after 2013, provinces' environmental 502 performance with a regional economic development level higher than 0.438 has changed. 503 Between 2010 and 2017, the proportion of regions with a regional economic development level 504 of less than 0.438 declined from approximately 90.3-83.9%. The story of regional 505 technological innovation has different positive effects on these regions' regional sustainable 506 development capabilities. During the study period, the proportion of areas where regional 507 technological innovation contributed to promoting sustainable development dropped from 87.1% 508 to 61.3%. Also, during the study period, the proportion of regions where regional technological 509 innovation only had a weak impact on regional sustainable development increased from 3.2% 510 to 22.6%. Therefore, it can be concluded that, overall, the situation in China is optimistic. 511  Table 9 presents the results of regression for the threshold model. As, the values of 514 emission reduction policies and renewable energy policies exceed the levels of corresponding 515 thresholds, the positive impact of energy policy on environmental performance gradually 516 increases. It is observed that the coefficient estimates for the threshold effect model are 0.0571, 517 0.012, respectively, and there is an upbound of their corresponding level of significance from 518 5% to 1%. This proposes that when the energy policy's pull-out position improves by 1%, the 519 environmental performance increases by 3.32% to 8..05%. It proves that the "J-shape" has a 520 marginal Growth trend. These investigation results depict how different regulations affect the 521 causality between the surrounded position of environmental regulations, the EP, and the 522 threshold or turning point in this relationship. 523 Secondly, the impact of energy consumption on carbon dioxide emissions coefficient 543 estimates for the threshold effect model is 0.0571, 0.012, respectively. There is an upbound of 544 their corresponding level of significance from 5% to 1%. This proposes that when the pull-out 545 position of the environmental regulations improves by 1%, the high emitting industries' total 546 factor energy efficiency increases by 1.2% to 5.7%. The central region is higher in level and 547 scale than the eastern and western regions. Also, the mechanism has apparent heterogeneity in 548 the east of, west and central regions. Regardless of the environmental regulations or industrial 549 structure, the central area's transmission path is much more important and more extensive than 550 that in the eastern and western regions. This phenomenon may affect environmental regulations 551 and industrial structure. However, this insignificant impact hinders the transmission path in the 552 region of the west. 553

Discussion and Policy Implication
The study pointed out that revenue must match responsibilities so that local governments 554 have both financial resources and corresponding environmental management rights and 555 obligations to improve environmental quality. Also, making full use of transfer payments, tax 556 rebates, and other support systems to enhance the environmental governance and public service 557 capabilities of local governments is another thing that needs to be done to stimulate the 558 enthusiasm and efficiency of local governments in protecting the environment. 559 Next, policymakers should pay close attention to the distorting effects of energy 560 consumption on local government behavior and are optimistic that they will maintain close 561 supervision to improve environmental governance efficiency. Therefore, it is necessary to 562 reduce pollution through coexistence and flexible government expenditure in environmental 563 protection and monitoring environmental protection. 564 An increase in environmental protection expenditures will help formulate pollution 565 transmission in underdeveloped areas and financial self-sufficiency areas. In contrast, 566 developed areas with strong economic capacity need to improve environmental supervision 567 further. This will help prevent the government's production expenditures from squeezing out 568 fiscal expenses related to technological innovation. 569 This is also conducive to changing the phenomenon that the upgrading of the industrial 570 structure leads to an increase in pollution emissions, which weakens the second stage of the 571 transmission path (related to media effects) and reduces the negative impact on the 572 environment. 573 Ethical Approval and Consent to Participate: The authors declare that they have no known 574 competing financial interests or personal relationships that seem to affect the work reported in 575 this article. We declare that we have no human participants, human data or human tissues. 576