As is common knowledge, energy use and carbon emissions are intimately correlated with human activity. Especially, population mobility has a major negative impact on carbon emissions in regions where people live (34, 35). Mobility of the population will unavoidably result in frequent use of transportation, which uses a lot of fossil fuels and raises carbon emissions. In the transportation sector, intelligent mobility can significantly reduce carbon emissions (36). Economic manufacturing will raise carbon emissions in cities, which is another significant element affecting China's industrial development. Some academics contend that while economic progress aids in the rationalization of industrial structure and carbon emissions, the upgrading and optimization of industrial structure has a detrimental effect on carbon emissions (19, 37).

China has a significant agricultural sector, thus its influence on carbon emissions cannot be understated. Research on the steadily rising rural carbon emissions in China's provinces discovered that the provinces closest to the ministry created the most emissions (38). Numerous academics have studied how the use of science and technology in diverse areas influences carbon emissions in light of China's tremendous advancement. Some academics state that China's total carbon emissions are somewhat increasing due to low innovation efficacy of green technology (39, 40). Also, some scholars think that the advancement of green technology benefits the research and creation of renewable resources. Eventually, it has a major detrimental influence on carbon emissions, but has less of an effect in the short term (41). Science and technology advancements encourage the development of digital technology, and these advancements can decrease carbon emissions as a result of their positive knock-on effects (42). According to the study, China's embedded carbon emissions from 2002 to 2017 were significantly impacted negatively by factors relating to the production structure of the digital economy (43).

We can assume that when we talk about the development zone, we are talking about development-related information. Therefore, the focus of this research material is on the development and economic development zones, as well as on how these development zones impact affect the local economy. According to some academics, the creation of development zones can, in some cases, encourage industrial agglomeration, boost industrial productivity, and boost exports. Also, the upgrading strategy policy of development zones influences both imports and exports (44). Moreover, China's development zones have a variety of repercussions, the most notable of which is a considerable impact on nearby manufacturing businesses (45). The creation of the economic development zone serves as a foundation for improving the organization of the industrial land, maximizing its use, and providing a solid land guarantee for the industrial transfer of the manufacturing industry (46).

Development zones can boost investment effectiveness, increase financial openness, and enhance the region's overall economic growth in terms of fostering regional economy (47). Development initiatives strategies has considerably significant spillover and driving impacts and serves as a launchpad for communities to pursue innovation-driven development initiatives (46). Over time, the spillover impacts will differ in numerous ways, largely depending on the special zone's strategic development and policy objectives (48). By distributing resources fairly, the spillover effect can effectively encourage regional economic growth, and help close the economic development gap between regions (22). By maximizing the arrangement of public service facilities, the high-tech zone diversifies urban public service goals in order to better the city as a whole (49). At the macro level, it encourages regional economic growth, and at the micro level, it benefits business performance and the improvement of individual performance (50, 51).

Development zones influence our ecological environment to varied degrees in addition to having an effect on our economic development. Theoretically, the establishment of the development zone will attract a large number of industries to settle in these cities, since industrial manufacturing is an indispensable real economy in a city. Unfortunately, it contaminated the environment of the city in the process of creating the economy. According to Guo et al. (52), the industrial pollution index records an initial decline and a gradual rise with time. Furthermore, the research shows that industrial production will produce a large amount of polluting gases, the most important of which are sulfur dioxide and nitrogen dioxide (52). Simultaneously, the production of heavy industry produces a large amount of industrial wastewater. Palani et al. (53) explain that improper treatment of wastewater will cause pollution of water resources, and the long-term consequence is the decline in the quality of the ecological environment.

Arguably, developing digital industries and sustainable industries are critical skills that can be used to solve this problem. Some researchers hold that developing the digital economy will significantly reduce the industrial pollution (35). The argument is that the high-tech sector puts emphasis on progressing science and technology (54). Consequently, the upgrading of science and technology will improve the efficiency and quality of environmental pollution treatment (55). Because of the varying stages of development zones, the pollution situation after implementation is also very different. The formation of provincial-level development zones further aggravated China's contamination intensity, but rising to the national level, they no longer play the same role (56). Likewise, concerning the smog pollution in the nation, the provincial development zone policies significantly increased smog pollution. On the other hand, the national development zone policies have no significant effect (57). Contrasting the development zones, the establishment of high-tech zones can improve the environment, and the increase of green patents and economic agglomeration are the mechanisms by which high-tech zones can improve the environment (47).

From the perspective of technological innovation, on the one hand, the establishment and construction of the development zone will attract a large number of high-tech enterprises (22), and at the same time, it will encourage local enterprises to carry out technological innovation activities and increase investment in innovative scientific research (26). High and new technology will improve the efficiency of production and manufacturing, meanwhile, it will increase the utilization rate of raw materials, promoting production efficiency and reducing the consumption of production resources and power resources (58), to achieve the effect of reducing carbon dioxide emissions. On the other hand, the establishment of the development zone will promote the development of emerging green industries such as high-end equipment manufacturing, new energy industries, and lead the direction of urban industrial transformation (59). With the acceleration of the industrial structure upgrading process led by the establishment of the development zone, the living space of the traditional “three high” industries in the city has been further reduced, and the vacated development space will be more occupied by strategic emerging green industries, which makes pollution-free, Clean and green production factors have been widely gathered and applied (60), it can also reduce carbon dioxide emission intensity and improve urban air quality.

From the perspective of resource allocation, on the one hand, the lower land price inside the development zone and related policy grants enable enterprises in the park to enjoy low supply of factors and ensure the continuity of normal production activities of enterprises (13, 61), to promote the efficiency of enterprise factor utilization. In addition, the complete infrastructure of the national development zone has successfully attracted foreign high-quality innovation capital, creating conditions for the emergence and development of high-end productive services (61). On the other hand, the high-end production factors attracted by the development zone can fully replace traditional production factors, promote the development of resource-intensive industries, and gradually reduce environmental pollution (62). To sum up, the optimization of resource allocation caused by the establishment of national development zones will help reduce undesired outputs in the industrial production process (63), and accordingly reduce carbon dioxide emissions.

From the perspective of the city's own conditions, under the background of the establishment of national development zones, there are large differences between different cities in terms of economic scale, innovation factor agglomeration capacity, resource allocation efficiency, and industrial policy formulation and implementation capacity (46). Generally speaking, regional economic development lags behind that of central cities (64). Under the guidance of the construction of ecological civilization, compared with other cities, the central city with stronger agglomeration ability of innovation elements can give full play to its advantages in terms of policy pilot, economic development scale and innovation element agglomeration, so as to achieve the transformation of economic development model (65). And the first-mover advantage of industrial structure adjustment, promote the full release of the potential of urban environmental improvement (66). Due to the relatively poor elements to promote the upgrading of the industrial structure and the poor external environment for industrial development, other cities have great potential and latecomer advantages in terms of environmental improvement capabilities (67). Through the construction of national-level development zone, such cities can obtain the technological spillover effect and high-end production factor agglomeration associated with the establishment of development zones, and fully release the vitality of urban industrial transformation (68), thereby enhancing the city's ability to improve environmental quality by adjusting its industrial structure.

From the perspective of spatial effects, in view of the spatial agglomeration and spatial differences in my country's environmental policies, institutional environment and energy structure, when examining the impact of the establishment of national development zones on local carbon dioxide emissions, the carbon emissions of neighboring cities should also be taken into account (69). In other words, we should fully consider the spatial spillover effect of policy. With the rapid development of country's digital information technology, cloud computing platform, and big data applications, industries in different cities are more closely connected (70). The development zone policy can not only rely on the transformation of the local energy structure to enhance the green total factor productivity and improve the quality of the urban environment (63), but also have a significant impact on the carbon dioxide emissions of adjacent cities (71). Namely, although the local carbon dioxide emissions have been reduced, their cost may be higher energy consumption in neighboring cities and lower urban environmental quality. In particular, to reveal the theoretical mechanism vividly, we have draw the framework of empirical steps and reported in Figure 1.

After the creation of the initial group of development zones, the nation's policies have always maintained development as an essential objective. Pilot areas of development zones across China were the large-scale cities. It is clear that the development brought to the cities caused economic growth and betterment of science and technology. Consumption of resources is another causal factor for economic growth, the consumption of resources is at the expense of the environment, which will bring large emissions of greenhouse gases, such as the total carbon emissions of cities. Research holds that the development of science and technology can ease this occurrence. Theoretically, the use of advanced technology will lower the total carbon emissions of the city. In this regard, this paper considers development zone policies as a set of quasi-natural experiments. The paper also relies on the DID model to assess how development zone policies impact carbon emissions. For this research, pilot cities were listed as the experimental group while the non-pilot cities were the control group. Since the development zone pilots are conducted in different years, a multi-phase DID model will be used. The following is the formula for the multi-period DID model:

Where the subscripts i and t represent city and year, TCE is the total carbon emissions of the city, DZP represents the dummy variable about the status of the policy implementing, X is a series of control variables, λ_i denotes the individual effect, λ_t denotes the time fixed effect, α₁ is the DID estimator, indicating the net effect of policy on total carbon emissions, β is the coefficient of control variables, μ is the random disturbance term, α₀ is the constant term. DZP_it is the dummy variable, representing a city's development zone policy implementing status. More specifically, DZP_it= 1 if city i implemented the development zone policy during the sample period, and 0 otherwise.

In addition, we constructed a PSM-DID model for correlation robustness checks because the DID model is not ideal in addressing the problem of sample selection bias. In addition, the model can make the more actual test in line with the theory. Using the model, we can more accurately measure if development zone policies can effectively affect the city's total carbon emissions. The specific theoretical model is as follows:

Furthermore, a spatially extended form of the DID model to SDID model based on Equation (1) was used to analyze the spatial spillover effects of policy further. The following is the formulated form of this extended model:

Where W is the spatial weight matrix, W_{T, T}D_itrepresents the spatial spillover effects among pilot cities, W_{NT, T}D_itdesignates the spatial spillover result of pilot cities on non-pilot cities which neighbor pilot cities, W×X_it is the spillover effects of control variables. β₁ is the spatial coefficient of W_{T, T}D_it, β₂ is the spatial coefficient of W_{NT, T}D_it, and β₃ is the spatial coefficient of W×X_it.

Last but not least, it should be pointed out that the DID model satisfies the parallel trend test. In other words, levels of carbon emission from pilot and non-pilot cities must not differ systematically over time. According to the approach of Jacobson and Sullivan (72), by constructing a series of temporal dummy variables, a temporal analysis research framework is used to analyze the dynamic influence of the application of development zone policies on urban carbon emissions. Consequently, we create a dynamic analysis model as formulated below:

Among them, DZP represents the dummy variable of the years before and after the realization of the development zone policy. DZP₀ is the dummy variable of the year when the city enforces the development zone policy; DZP_t−n is the dummy variable of n years before the implementation of the development zone policy; DZP_t+n is the dummy variable of n years after the enforcement of the development zone policy; other variables are constant as in the case of Equation (1).

Panel datasets and IPE research reports of 285 cities in China provide the sample data. The data period was scanning from 2003 to 2020. Other data sources included several official national statistical documents like the China Urban Construction Statistical Yearbook, the China Urban Yearbook, and the China Statistical Yearbook. After the disruption, all minor indicators are adjusted to the constant price in 2003 as per the provincial price index.

It is important to note the regression analysis is conducted on the primary variables first then other overall variables in the results of the regression analysis are shown in Table 2. From column (1), it is found that the coefficient of the key variable DZP is −1.390. The coefficient reaches a significant level of 5%, indicating that the implementation of the development zone policy is conducive to the reduction of the total carbon emissions of the pilot cities. Column (2) of the table indicates that with the upsurge of the control variables, the coefficient of DZP and the significance test do not change considerably. These results mean that control variables have the very minimal effect on the total carbon emissions of the pilot cities. This conclusion is consistent with Gao et al. (76), whose research on the impact of development zones on carbon emissions, and the establishment of development zones has a positive impact on the city's carbon emission performance. Table 2 also shows that in the control variables, FD, FDI and the total carbon emissions of the pilot cities have the positive relationship of change. On the other hand, lnPGDP, lnDENSITY, IU, UR and the total carbon emissions of the pilot cities have an inverse relationship of change relating to the total carbon emissions of the pilot cities.

Several robustness tests were necessary to demonstrate the robustness of the results gotten from the regression tests on the core results. These robustness tests included PSM-DID model evaluation, endogeneity test, dynamic effect analysis, parallel trend test, placebo test, and other test removing municipalities that can prove robustness of regression analysis. Based on the results from a series of analysis the core results have strong robustness. In other words, the effect of development zone policies in pilot cities on the total carbon emissions is very significant.

According to relevant literature (32, 79), it is found that China's industrial carbon dioxide emissions have different emission standards due to different local policies and environmental requirements. Therefore, it is important to note that development zone policies may have varying effects from one city to the next. This phenomenon is caused by factors like large differences in the resource endowments of cities, different urban scales and different geographic locations. These variances are shown in Table 7, columns (1), (2), and (3) show the effect on the total carbon emissions after enforcing the development zone policy in the eastern, central and western pilot cities, respectively. Based on the analysis, development in the central region is balanced and implementation of the development zone policy at this point causes more effect on the decline of carbon emissions. The eastern region in China is relatively developed and the development zone policy is implemented early. Despite these efforts, the marginal effect of the policy is fading, from the current stage, the policy effect is not ideal. The western region in the country is least developed in terms of economy, science and technology, making it too backward in comparison. Therefore, it is essential to hasten the development of these sectors and eventual implementation of the development zone policy. As shown in the table, the implementation of the policy resulted in the slight increase in urban carbon emissions. Columns (4) represents the impact of the implementation of the development zone policy on the total carbon emissions in small cities, column (5) represents the impact of the implementation of the development zone policy on the total carbon emissions in medium cities and column (6) represents the impact of the implementation of the development zone policy on the total carbon emissions in large cities. The conclusion we made is that the larger the city scale, the more beneficial it is to reduce the total carbon emissions. Columns (7) represents the impact of the implementation of development zone policies in resource-based cities on the total carbon emissions. Column (8) represents the impact of the implementation of development zone policies in non-resource-based cities on the total carbon emissions. Conclusively, the enforcement of the development zone policy in non-resource-based cities has a very obvious effect on lowering the total carbon emissions of the city. Conversely, implementing the development zone policy increases the total carbon emissions in resource-based cities.

Possibly, there is a spatial connection between carbon emissions and neighboring cities. Neglecting this spatial correlation may cause inconsistent findings when comparing the theoretical results and actual results of the study. To prevent this inconsistency, the spatial impact of implementing development zone policy pilots and the spatial dependence of carbon emissions were taken into account in the model, and the SDID model was used for analysis. The results are shown in Table 8.

Column (1) and column (2) prove that the application of the development zone policy has a spatial effect on the total carbon emissions of neighboring cities. Absolute values of the coefficients of variables W_{NT, T}D, W_{T, T}D, and DZP, are very large and they all reach the 1% significance level. From column (1), it is conclusive that regardless of whether control variables are considered, the development zone policy has the spatial effect of the total carbon emissions of surrounding cities. On the other hand, the coefficient of DZP is significantly negative. In other words, the development zone policies can significantly reduce the total carbon emissions. The coefficient of W_{NT, T}D is significantly negative, indicating that the spatial effect of the development zone policy pilot cities reduces total carbon emissions in neighboring non-pilot cities. On the other hand, the coefficient of W_{T, T}D is significantly positive, indicating that the spatial effect of the pilot cities of development zone policy increases the total carbon emissions of the neighboring pilot cities. This condition may be implicated with the variances in economic scale, the policy implementation and the administrative levels between pilot cities and non-pilot cities. According to Guo et al. (52), whose research stands there are indeed geographical proximity effects and spatial spillover effects between cities in China. This effect is more pronounced between cities that are spatially closer.

To further improve the accuracy of the study, we factored if the upgrading and superposition of policies will still have the same impact on the total carbon emissions as the core results. The results of the regression analysis are shown in Table 9. The policy upgrading effect of the pilot city's upgrading to the national high-tech zone policy from the provincial high-tech zone policy is Column (1). DZP_up1 represents the core variable with its coefficient being −2.891 and a significance level of 5%. These results show that upgrading the high-tech zone policy considerably helps to lessen the total carbon emissions. The second column is the policy upgrading effect of the pilot cities' upgrading to the national-level economic development zone policy from the provincial-level economic development zone policy. DZP_up2 represents the core variable and its coefficient is −0.131. These results show that upgrading the economic development zone policy causes a minor decrease in the total carbon emissions.

The policy superposition effect of the pilot cities is Column (3). The superposition effect first applies to the national economic development zone policy and then to the national high-tech zone policy. DZP_sp1 represents the core variable. Its coefficient is −2.745 with a 10% significance level. These figures show that the superposition of such policies has a relatively clear lessening impact on the total carbon emissions of the pilot cities. Column (4) is the policy superposition effect of the pilot cities first executing the national high-tech zone policy and then executing the national economic development zone policy. The core variable is represented by DZP_sp2, and its coefficient is −0.665. These figures support that the superposition of such policies has a feeble impact on the total carbon emissions in the cities. Contrasting column (1), (2), (3), and (4), shows that the high-tech zone policy has more effect than the economic development zone policy in the cities. Contrasting the two columns (1) and (2), shows that the upgrade of the high-tech zone policy has a more significant and clear impact in terms of reducing the total carbon emissions compared to the upgrade of the economic development zone policy. Comparing the two columns (3) and (4), shows that shifting to a national high-tech zone from a national economic development zone can decrease the total carbon emissions of pilot cities more than shifting to a national economic development zone from a national high-tech zone.

By analyzing the results of the study, five main conclusions can be drawn: first, the implementation of the development zone policy can effectively reduce the total carbon emissions of the city; Second, the development zone policy that lags behind a single cycle can still significantly reduce the total carbon emissions of cities; Third, the implementation of development zone policies for cities, larger cities and non-resource cities in the central region has a better effect on carbon reduction than other cities; Fourth, there is a spatial effect on the impact of development zone policies on carbon emissions. Its spatial effect reduces the total carbon emissions of neighboring non-pilot cities and increases the total carbon emissions of neighboring pilot cities. Fifth, there are upgrading and overlapping effects in the implementation of relevant policies. Whether it is upgrading or superimposing, the carbon reduction effect of the high-tech zone policy is better than that of the economic development zone.

The following policy suggestions are put forward based on the above discussions: First, the development zone policy can effectively reduce the total carbon emissions, which means that the formulation of relevant policies can successfully achieve a balance between the ecological environment and economic development, which is a positive measure toward optimizing the ecological environment. Therefore, developing countries similar to China that face similar environmental optimization in the process of urbanization can vigorously promote the construction of development zones. Second, there is a single-period lag effect in the implementation of the development zone policy, which means that there is a single-cycle buffer time for the implementation of relevant policies. Therefore, the government should fully consider the time inconsistency of the policy and the final policy results when formulating relevant policies. The third suggestion is the selection of development zone policies suitable for local cities while considering the city's scale, resource endowment, and geographical location while promoting the inclusive development of the local cities in a targeted manner to evade the variance in the effect of development zone policies on carbon emissions. The western region of China is comparatively backward in technology. However, this region has abundant natural resources and a large urban area. The country's western development project has been advancing in recent years and it is arguable that the region is more conducive for the development zone policies. Science technology and natural resources are ample in the central region of China, thus it is necessary to choose and implement appropriate high-tech zone policies to coordinate the overall development and balance the requirements of development and economy. In the eastern part of China, there is a high population density, technological development and advanced economic development. Moreover, the majority of cities in eastern China are closer to the sea, hence, development of foreign trade is a primary advantage for these urban areas. With this in mind, local economic and technological development is arguably reliant on the development zone policies. Forth, the local government should try to reduce the impact of the spatial effect in the implementation of the development zone policy in response to the spatial effect among cities. The central government is responsible for strictly checking the policy effect of each region, establishing strict regulated rules, and issuing relevant deployment documents. The local governments are implicated with conducting implementations of the policies to categorize regional responsibilities and interests, further preventing any pollution transfer. Finally, according to the research results, the central government should vigorously support the construction of high-tech economic development zones in local areas as much as possible, take innovation as the driving force for sustainable development, and create a win-win development pattern of economic development and ecological protection.

Some of the limitations discovered during the research of this paper may encourage future research on relating subjects. To begin with, bias may have arose during the calculations of industrial carbon emissions because of the limited data on the subject. The aim is to attain an inclusive measure of the environmental efficiency. Thus, we included industrial carbon emissions as one of the undesirable outputs. Nonetheless, as per the “China Urban Statistical Yearbook”, we can only analyze the carbon dioxide emissions from three energy sources, that is, natural gas, electricity and liquefied petroleum gas. And so, the results can be altered or protracted in the future using newer or alternative data sources. The second limitation is that the endogeneity analysis lacked some sample data for the tests. The instrumental variable consists of the number of stations in a city and the product of the number of taxis. Unfortunately, we only managed to collect data for most cities, while we could not get data from other cities because of the limited data collection methods available and limited research. In other words, in all the 285 sample cities used for this study, there are missing data. It follows that this missing data needs to be collected to facilitate supplementary verification. Lastly, we use the dummy variable DZP to represent the core variable because the PITI (pollution source supervision information disclosure index) standard used by the 113 pilot cities from 2008 to 2012 is different from the PITI standard used by the 120 pilot cities from 2013 to 2020. The variance in PITI intensity between the two times is not definitely distinguished in terms of treatment. In future, more unified and reliable standards can enhance the study of the effect of national development zone policy on carbon emissions.