Can big data mitigating Chinese counties carbon emissions? Evidence from a Quasi-Natural Experiment
YangZechen1, LiXuelan2,*
1 College of Economics and Management, South China Agricultural University;yzcsheet@163.com
| Academic Editor: Firstname Lastname Received: date Revised: date Accepted: date Published: date Citation: To be added by editorial staff during production. Copyright: © 2025 by the authors. Submitted for possible open access publication under the terms and conditions of the Creative Commons Attribution (CC BY) license (https://creativecommons.org/licenses/by/4.0/). |
2 School of Management, Anhui Science and Technology University; lixuel@ahstu.edu.cn
* Correspondence: lixuel@ahstu.edu.cn; Tel.: 18755269730.
Abstract
The implementation of China’s big data strategy serves as a crucial catalyst for fostering high-quality development. Nevertheless, the potential correlation between big data development and carbon emission reduction has yet to be investigated, despite the pivotal role of data development in achieving county-level carbon emission reduction in China. This study aims to construct a multi-period difference-in-difference (Difference-in-difference method) method utilising macro panel data from 9,978 counties in China, spanning the period from 2010 to 2021. By leveraging the exogenous impact policy of the construction of China’s big data comprehensive experimental area, we assess the impact of policies implemented in these areas on county-level carbon emission reduction. The findings indicate that the implementation of policies pertaining to the establishment of big data integrated pilot zones can result in a notable reduction in county-level carbon emissions. Path analysis further demonstrates that the policies influence carbon emissions through two channels: promoting industrial upgrading and technological progress. These research results provide empirical evidence supporting accelerated development and application of big data, facilitating green development at the county level while contributing towards achieving China’s “double-carbon” goal.”
Keywords
Carbon emissions; Chinese counties; Big data; Quasi-Natural experiment; Difference-in-difference
- Introduction
The adverse impacts of extreme weather necessitate global regulation of greenhouse gases and mitigation of climate warming(Sun et al.,2022).
To address global environmental issues, such as pollution and climate change, the Chinese government has established policy objectives to peak carbon emissions by 2030 and achieve carbon neutrality by 2060. These goals have been incorporated into the overall layout for national ecological civilization construction. China’s transition from carbon peak to carbon neutrality is expected to take approximately 30 years, which is nearly half the time frame compared to developed countries. Consequently, China faces significant difficulties and challenges. China’s plans for reducing carbon emissions are mainly formulated at the national and provincial levels, specifying distinct requirements for different industries to achieve carbon emission targets at the provincial level, which leads to the neglect of the relevant reality of small spatial scales such as county level when formulating policies(Wang et al.,2024). Counties serve as vital transitional links between urban and rural areas within China’s administrative divisions (provinces, cities, counties, districts). They form an integral part of China ‘s urban system while supporting integrated development between urban and rural areas; thus making them fundamental units for studying carbon emissions. County is the core administrative entity and spatial basis of China’s industrial and economic expansion. (He et al.,2022 ) The contribution rate to China’s GDP is about 51.8 %. (NBS,2018). The county has undertaken most of the ‘ high-pollution and high-emission ‘ enterprises that are eliminated during industrial development processes. According to the latest statistics, China’s county carbon emissions account for 62 % of the country’s total carbon emissions. Therefore, it is necessary to analyze its emission reduction mechanism and experience from the county perspective. ( Yong-Yong et al., 2021 ).
In recent years, digital technology, digital economy and big data development have become the core forces leading the technological revolution and industrial transformation. Data is a new type of production factor in the form of digital, driven by digital technology, with modern Internet as an important carrier, through the transmission and use of information to improve modern production methods. Additionally, China’s big data industry has experienced rapid growth in recent years with an estimated scale reaching 1.57 trillion yuan by 2022, reflecting an annual increase of 18%(Yu et al.,2022). The swift development of big data has become the key to leading technological progress and profoundly impacting regional carbon emissions through transformative changes in traditional economic production and operation mechanisms(Hao et al.,2021).
According to the neoclassical economic growth model, technological progress can promote economic development. Economic development has externalities, while environmental pollution has negative externalities. Scholars have studied the relationship between digital technology and pollution emissions, and found that digital technology can affect environmental pollution from multiple dimensions. Digital technology can not only provide information technology support for environmental governance. ( Kwon et al., 2014 ; shin and Choi, 2015 ; hampton et al., 2013 ). At the same time, digital technology can also affect environmental pollution through multi-dimensional paths such as energy utilization, economic development and industrial structure upgrading. ( Dong et al., 2022 ). In addition, digital technology has reduced the dependence of many industries on energy use, thereby reducing pollution emissions ( Ishida, 2015 ). However, there are also scholars who believe that digital technology, digital economy, etc.will further promote pollution emissions. Wang ( 2022 ) believed that the popularity of ICT has promoted energy efficiency, further expanded the demand for energy use in manufacturing and other industries, and aggravated carbon emission pollution ( Wang et al., 2022 ). Salahuddin and Alam, ( 2016 ) believed that the rapid growth of Internet construction and network infrastructure will inevitably lead to the use of regional electricity, thus promoting the level of carbon emissions. Based on previous studies, it can be found that digital technology, digital economy and other industries that rely on the Internet, ICT and other industries have systematic research on carbon emissions, but data elements, as an important capital in the production function to promote economic development, are bound to have an impact on carbon emissions. Therefore, the development of big data has gradually attracted the attention of scholars. At present, research on big data mainly focuses on alleviating corporate financing constraints, creating new tasks, increasing labor income share and corporate green development ( Begenau et al., 2018 ; gardiner et al., 2018 ; duygan-Bump et al., 2015; Acemoglu and Restrepo，2019). However, the existing research does not involve the analysis of the relationship between big data and county carbon emissions.
Based on this premise, this paper takes the construction of big data comprehensive pilot area implemented in China since 2016 as a quasi-natural experiment to analyzes the impact of big data development on county-level carbon emissions at a smaller scale within administrative division. To address endogenous concern, this paper employs the multi-period DID method to construct the model and verifies the model results through various robustness testing methods.The marginal contribution of this paper are as follows: First, based on the county panel data, it evaluates the impact of big data development on reducing county-level carbon emissions, providing better empirical evidence for such county reduction. Secondly, from an industrial institutions and technological progress perspective, this paper analyzes specific mechanisms through which policies related to big data comprehensive pilot zone affect carbon emission reduction. The main sections of this paper are organized as follows : The second section is literature review and theoretical analysis; The third section describes the data used in this study along with variables and models employed ; The fourth section comprises the main body of this paper and various types of robustness tests, while the fifth section summarizes the entire text and proposes corresponding policy recommendations.
- Policy Background and Theoretical Analysis
2.1 Policy background
In the context of the rapid global informatization, big data has emerged as a crucial foundational strategic resource for nations. With the proliferation and advancement of information technologies such as the Internet, 5G, the Internet of Things, and cloud computing, we have entered an era dominated by big data. The ability to extract knowledge from vast amounts of data and convert it into productivity has become a pivotal factor in securing success in global competition. Like other countries around the world, China has been actively promoting the development of big data in recent years. Overall, our country’s progress in this field can be divided into three stages. The first stage is the preliminary stage, which took place before 2014 and primarily focused on discussing concepts and technologies related to big data. However, it failed to establish a comprehensive system. The second stage is the implementation stage, occurring between 2014 and 2019 when big data development became a national strategy. This period witnessed the orderly promotion of comprehensive pilot zones for big data and the introduction of various provincial policies. The third stage is the deepening stage, starting from 2019 when data officially became a new factor of production. It was explicitly recognized that data factors are crucial elements for digital economy development. In order to promote big data development, in 2015, The State Council issued an Action Plan outlining its significance in facilitating economic transformation and upgrading. In 2016, the state approved the establishment of a national Big Data comprehensive pilot zone in Guizhou province as an experimental area for implementing national strategies and promoting big data development. By 2022, China had established a total of ten big data pilot zones.
2.2 Research hypotheses
According to the relevant policy documents of the Chinese government on the big data comprehensive experimental area, the key projects of the big data experimental area include : big data system innovation, data resource sharing, big data innovation application and big data infrastructure construction. It can be seen from the above projects that the construction of big data comprehensive experimental area will have two effects on carbon emissions : first, the construction of big data experimental area will further coordinate digital infrastructure and innovate in integration and utilization, open sharing, industry application and so on. For enterprises, the availability of data and technology can help them optimize the production process and reduce unnecessary power consumption. Secondly, the innovative application of big data in the pilot area provides opportunities for the statistics, analysis, mining and prediction of power big data, which can promote the development of power big data and improve the intelligent decision-making ability of government and enterprises. Specifically, on the one hand, by building a power big data platform, government departments can monitor the power consumption of the whole society in an all-round way ; on the other hand, through the deep mining of power big data, enterprises can accurately calculate the power consumption of their production and operation, and analyze the prediction of power supply and demand, so as to reduce the power consumption of enterprises.
H1:Big data development contributes to carbon emission reduction at county-level.
Big data is the concentrated expression of technological progress in the new era. The development of big data is bound to cause technological progress and changes in production methods, thus changing the industrial structure and reducing carbon emission pollution. Based on this, this study expounds the specific path of big data development affecting carbon emission reduction from the perspective of technological progress and industrial upgrading.
2.2.1 technological progress
The endogenous growth theory posits that technological progress and its positive externalities enhance the efficiency of factor resource utilization, so that resources can be saved and recycled, and reduce county pollution emissions by reducing energy consumption and obtaining more output. The development of big data impacts carbon emission reduction through technological advancements. On the one hand, big data can expand the information search space(Kwon et al.,2014). Through the capture, discovery, storage, processing and analysis of data, innovation entities can quickly obtain structured or unstructured information resources related to innovation decision-making at a lower cost(Shin et al.,2015), reduce the information asymmetry of R&D innovation process and the uncertainty of technology application, and then improve the rate of return on innovation output and stimulate the internal motivation of innovation entities. On the other hand, the rapid development of big data has effectively improved the financial mismatch dilemma existing in the traditional financial system, which can better match financial resources with innovation activities, accelerate the financing speed of innovation subjects, reduce the financing cost of innovation investment, and then alleviate the financing problems in the innovation process, and provide basic guarantee for the smooth development of innovation activities(Wang et al.,2022). The improvement of technological level improves the efficiency of resource use and reduces carbon emission pollution.
2.2.2 upgrading of an industrial structure
According to the theory of environmental economics, the economic structure of a region determines the level and type of its resource consumption and environmental pollution. The adjustment of industrial structure can produce carbon emission reduction effect through optimal resources allocation. The development of big data impacts carbon emission reduction through the optimization of industrial structure : relying on the policy of big data comprehensive pilot area, various regions have attracted numerous big data and related enterprises, which can promote the development of information industry and drive the upgrading of industrial structure(Heo et al.,2019). For instance, Guizhou Province in China has implemented an ‘ enterprise integration ‘ plan to foster the development of big data industry and related sectors, which helps attract and nurture data-driven industries. The growth of the information industry facilitates a shift in production factors from low-efficiency sectors to high-efficiency ones, resulting in optimized and upgraded of industrial structure. At the same time, the concentration of big data also generates spillover effects on local enterprises, and drive the upgrading of industrial structure by promoting the transformation and upgrading of traditional industries. The continuous promotion of innovative applications of big data within pilot areas by transcending conventional production and manufacturing methods, can radiate and drive advancements across sectors, significantly enhancing production efficiency and optimizing processes. This is particularly impactful for the technological advancement and efficiency improvements within traditional industries. The upgrading of industrial structure optimizes the efficiency of resource allocation and reduces carbon emission pollution at county levels.
H2:Big data development can reduce county-level carbon emissions by promoting technological progress.
H3:Big data development can reduce county-level carbon emissions by promoting industrial structure livelihoods.
- Research Design
3.1. Measurement Model Setting
With the help of the policy exogenous impact of the construction of the national big data comprehensive pilot area, the impact of big data development on China ‘s county-level carbon emissions is studied. By dividing the samples into experimental group and control group, the following difference-in-differences model is constructed. The benchmark model is set as follows :
(1)
Among them,is the carbon emission intensity of county i in year t;
Whether it is in the national big data comprehensive test area in the t year of i county belongs to the test area with a value of 1, otherwise it is 0;
is the parameter to be estimated, indicating the impact of policy implementation on carbon emissions.
and
represent regional fixed effect and time fixed effect respectively;
is expressed as an error term;
is other control variables that affect carbon emissions, including : economic development level, traditional financial development level, etc.
3.2 Variables Selction
3.2.1 Explained variables
County carbon emissions. It is measured by the ratio of the county ‘s unit GDP to the total amount of carbon dioxide emitted by the county.
3.2.2 Core explanatory variables
The core explanatory variable is the dummy variable ( DID ) of the national big data comprehensive pilot area, and the development of big data is characterized by the intersection of the dummy variable ( Group ) of the city type ( experimental group, control group ) and the dummy variable ( Post ) of the policy time. Specifically, if a city or its province establishes a big data pilot area, it is the experimental group, the Group value is 1, the remaining cities are the control group, and the Group value is 0 ; for the time dummy variable Post, the value of the year after the implementation of the big data test area is 1, and the value of the remaining years is 0.
3.2.3 Control variables
In order to control the influence of other factors on the income gap between urban and rural areas in the county and reduce the estimation error, the per capita land area, economic level, financial dependence, financial development level and infrastructure construction level are selected as control variables for model construction and research. The definition of specific variables is shown in table 1.
Table 1. variable definition
| variable name | sign | variable definition |
| Carbon emission reduction | CARBON | The ratio of carbon emissions to GDP. |
| Establishment of big data experimental area | DID | The sample county is set up as a big data experimental area in a certain year, the variable value is 1, otherwise the value is 0. |
| Value of industrial output | IND | The ratio of the added value of the secondary industry to the total added value. |
| Economic development | GDP | The logarithm of per capita GDP. |
| Level of consumption | Consu | The ratio of social consumption to GDP. |
| Financial self-sufficiency rate | Fin | Ratio of fiscal expenditure to income. |
| Network | NET | Number of Internet processing households. |
| Telephone | TEL | The ratio of the number of fixed telephone users to the total population at the end of the year |
| Human capital | HUM | Ratio of the number of students in ordinary secondary schools to the total population at the end of the year |
| population density | DEN | The ratio of total population to administrative area at the end of the year |
| Number of enterprises | NOE | Number of industrial enterprises above designated size ( number ) |
3.3 Data sources
The data in this paper are unbalanced panel data of 730 counties from 2007 to 2021. The relevant data are derived from the ‘ China County Statistical Yearbook ‘, county statistical bulletins, county statistical bureaus, etc.
4. Analysis of Benchmark Empirical Results
4.1 descriptive statistic
Table 2 descriptive statistic
| Variable | Mean | Std | Min | Max |
| CARBON | 3.069 | 12.489 | 0.000 | 420.250 |
| DID | 0.205 | 0.403 | 0.000 | 1.000 |
| IND | 0.417 | 0.159 | 0.014 | 0.919 |
| GDP | 3.366 | 3.508 | 0.1549 | 91.941 |
| Consu | 3.401 | 1.979 | 0.076 | 35.535 |
| FIN | 0.357 | 0.257 | 0.0354 | 2.614 |
| NET | 10.660 | 0.686 | 5.3327 | 14.023 |
| TEL | 0.119 | 0.131 | 0.0001 | 4.125 |
| HUM | 0.053 | 0.032 | 0.0010 | 1.039 |
| DEN | 0.007 | 0.018 | 0.0000 | 0.369 |
| NOE | 4.086 | 1.080 | 0.0000 | 8.258 |
4.2 Basic characteristics
To study the impact of big data development on county carbon emissions, this paper constructs a regression model for analysis, and the regression results are shown in table 3. The results show that the estimation coefficient of big data development on urban carbon emission intensity is-0.523, which is significant at the level of 1 %, indicating that the development of big data has significantly promoted the decline of county carbon emission intensity, and preliminarily verified the content of hypothesis H1. Model 2 further incorporates relevant control variables, and finds that the estimated coefficient of the core explanatory variable is-0.292, which is significant at the level of 5 %, indicating that the development of big data can effectively reduce the intensity of urban carbon emissions, further confirming the theoretical hypothesis H1.
Table 3 Baseline regression
| Variable | (1)Carbon | (2)Carbon | (3)Carbon |
| DID | -0.523*** (0.144) | -0.288** (0.147) | -0.292** (0.147) |
| IND | -1.747** (0.588) | -1.739** (0.594) | |
| GDP | 0.069** (0.026) | 0.729** (0.026) | |
| Consu | -0.242*** (0.033) | -0.233*** (0.033) | |
| FIN | -1.635*** (0.325) | -1.808*** (0.329) | |
| TEL | -0.295 (0.448) | -0.297 (0.449) | |
| HUM | -5.153** (2.239) | -5.048** (2.261) | |
| DEN | 1.951 (6.375) | 1.574 (6.574) | |
| NOE | 0.301** (0.107) | 0.233** (0.109) | |
| Cons | 4.977*** (0.119) | 13.392*** (1.607) | 13.414*** (1.603) |
| County FE | Yes | No | Yes |
| Year FE | Yes | Yes | Yes |
| N | 9978 | 9978 | 9978 |
| R2 | 0.076 | 0.091 | 0.091 |
4.3 Balance Trend Test
The premise of the consistency of the results of the difference-in-difference model is that the experimental group and the control group meet the parallel trend hypothesis, that is, before the implementation of the pilot policy, the carbon trends of the pilot counties and non-pilot counties should be parallel. Therefore, this paper refers to the time research method proposed by Jacobson et al. (1993) to conduct a parallel trend test on the dynamic effects of the pilot policy, and constructs the model as follows :
(2)
Model ( 2 ) considers the relative time of the establishment of new rural financial institutions on the basis of Model ( 1 ). In order to avoid the problem of collinearity, the first year before the establishment of new rural financial institutions is taken as the benchmark group, and is the focus of parallel trend test. The coefficient represents the estimated value of the dummy variables before and after the establishment of new rural financial institutions. The specific parallel trend test results are shown in Figure 1.
Picture 1 parallel trend test
It can be seen from Figure 2 that after controlling the county fixed effect and the annual fixed effect, there is no significant difference in carbon emissions between the sample counties, and the parallel trend hypothesis is established. And with the passage of time, within 5 years after the establishment of the big data comprehensive experimental area, the mitigation effect of the establishment of the big data comprehensive experimental area on the county ‘s carbon emissions has gradually increased, indicating that the big data comprehensive experimental area can alleviate carbon emissions. This shows that the benchmark regression results of this model have certain robustness.
4.4 Placebo test
In order to exclude the possibility of random factors affecting county carbon emissions, this paper refers to the practice of scholars such as Wang (2024), and does a placebo test by randomly selecting the treatment group. Specifically, this paper randomly assigns the role of processing group and control group to the sample county, thus obtaining a randomly generated false big data comprehensive test area to set up dummy variables, and uses this variable to replace the big data comprehensive test area in the regression model to set up variables for regression. This process is repeated 500 times, and the regression coefficient and probability density of the variables set up in the false big data comprehensive test area are shown in Figure 3. It can be seen from Figure 3 that most of the regression coefficients of the variables set up in the false big data comprehensive experimental area are near 0, indicating that the impact of this variable is non-random, which provides further support for the research results that the establishment of the big data comprehensive experimental area will reduce the county ‘s carbon emissions.
Picture 2 Placebo Test Chart
Note : The horizontal axis is the year relative to the establishment time ( t ) of the new rural financial institutions ; the longitudinal axis is the estimated coefficient of the annual dummy variable before and after the establishment of the new rural financial institutions and the cross-multiplication term of the establishment of the new rural financial institutions when the base year is the year before the establishment of the new rural financial institutions ( t-1 ) ; the dotted line is the 95 % confidence interval of the estimated coefficient.
4.4 PSM-DID
When setting up big data experimental areas, we often take into account the status of financial institutions and the level of economic development in the county. Therefore, this study believes that the implementation of this policy is not a ‘ natural experiment ‘ in the true sense. This situation may violate the common trend hypothesis of the difference-in-difference method. Therefore, this paper adopts the method of PSM-DID to conduct robustness test to test whether there is a problem of sample selection bias. Based on this, this paper selects 8 matching variables such as per capita land area, communication level, economic level, financial dependence and financial development level for regression test. First, these matching variables were used for logit regression , and then the model were tested using nearest neighbor matching, kernel matching and radius matching. The test results are shown in Table 4. As shown in column (4) of Table 5, the estimated coefficient is significantly negative, which is consistent with the results obtained above. It shows that the establishment of big data pilot areas can significantly alleviate carbon emission pollution, which verifies the robustness of the regression results of the previous model.
Table 4 PSM-DID
| Variable | (1)Fixed effect | (2)PSM+DID | ||
| nearest neighbor matching | kernel matching | Radius matching | ||
| DID | -0.292** (0.147) | -0.304** (0.154) | -0.304** (0.154) | -0.304** (0.154) |
| Controls | Yes | Yes | Yes | Yes |
| County FE | Yes | Yes | Yes | Yes |
| Year FE | Yes | Yes | Yes | Yes |
| Cons | 13.414*** (1.603) | 13.347*** (1.669) | 13.347*** (1.669) | 13.347*** (1.669) |
| Obs | 9978 | 9685 | 9685 | 9685 |
| R2 | 0.091 | 0.087 | 0.088 | 0.088 |
4.5 Robustness Test
In addition, this study also carried out a variety of robustness tests, by replacing the dependent variables, replacing the dependent variables lagging one period, and further controlling the fixed effects of cities and provinces to carry out the robustness test. The specific results are shown in Table 5. As shown in Table 6 column (3) -column (6), the coefficients of DID are all negative and significant at the 5 % level, which further proves the reliability of the model results.
Table 5 Robustness test results
| Variable | (3) | (4) | (5) | (6) |
| DID | -0.018** (0.007) | -0.385** (0.147) | -0.327** (0.144) | -0.325** (0.141) |
| Controls | Yes | Yes | Yes | Yes |
| County FE | Yes | Yes | Yes | Yes |
| Year FE | Yes | Yes | Yes | Yes |
| Cons | 13.889*** (0.076) | 13.319*** (1.736) | 18.844*** (1.949) | 14.978*** (1.559) |
| Obs | 9978 | 9248 | 9978 | 9978 |
| R2 | 0.1869 | 0.089 | 0.1357 | 0.1558 |
5.Further Analysis
5.1 Mediation analysis
Studies have shown that the establishment of China ‘s big data comprehensive pilot area is conducive to improving regional technological innovation and upgrading regional industrial structure. The above research provides sufficient empirical basis for clarifying the development of big data on county-level carbon emission reduction research. However, there is no research on the specific mechanism of the establishment of a national big data comprehensive experimental zone affecting carbon emission reduction. Therefore, this section analyzes the specific path of big data development affecting county carbon emission reduction by establishing a mechanism model. The specific model is as follows :
(3)
(4)
In Equation (3) and Equation (4), and
represent industrial upgrading and technological progress respectively. Based on the limited data of counties, this paper selects the logarithm of the number of patent inventions in counties and the ratio of the added value of the tertiary industry to the total output value as the proxy variables of industrial upgrading and technological progress. The specific regression results are shown in Table6.
Table6 The guiding mechanism of big data development affecting carbon emission reduction
| Variable | Innovation | (2)Carbon | (3)UP | (4)Carbon |
| DID | 0.384*** (0.042) | -0.387** (0.177) | 0.013*** (0.001) | -0.290** (0.146) |
| Innovation | -0.129** (0.049) | |||
| UP | -7.267*** (1.096) | |||
| Controls | Yes | Yes | Yes | Yes |
| County FE | Yes | Yes | Yes | Yes |
| Year FE | Yes | Yes | Yes | Yes |
| Cons | 0.072 (0.166) | 5.925*** (0.695) | 0.598*** (0.004) | 11.246*** (0.254) |
| N | 7870 | 7767 | 10104 | 9978 |
| R2 | 0.599 | 0.062 | 0.845 | 0.093 |
Columns (1) and (2) of Table 6 show that the establishment of national big data comprehensive experimental area has a significant positive impact on technological progress, and after adding technological progress to the benchmark model, the establishment of big data comprehensive experimental area has a significant negative impact on carbon emission reduction and technological progress, indicating that technological progress as a mechanism affecting the development of big data on carbon emission reduction is established. The coefficients of industrial institutions and DID in column (3) and column (4) show that the livelihood of industrial structure is also established as a mechanism affecting the development of big data on carbon emission reduction.
5.2 Heterogeneity analysis
Since the national big data comprehensive experimental zone is constructed in batches, the policy design and specific implementation measures in different regions are different. Different policy intensity and heterogeneous government support may also lead to the heterogeneous impact of the construction of the experimental zone on carbon emissions. Therefore, it is necessary to examine the heterogeneous effects of different regional test areas on carbon emissions. At the same time, due to the geographical location and policy resource advantages, the impact of the policy pilot area on the eastern and western regions of China ‘s central and western regions and the eastern coastal regions may be different. Based on this, this paper conducts a heterogeneity test by grouping regression.
Table7 Heterogeneity analysis results table
| Variable | (1)the east region | (2)the western and middle regions |
| DID | -0.628 (0.301) | -0.319** (0.099) |
| Controls | Yes | Yes |
| County FE | Yes | Yes |
| Year FE | Yes | Yes |
| Cons | 16.609*** (3.374) | 10.875*** (1.047) |
| N | 4732 | 5235 |
| R2 | 0.065 | 0.2851 |
From the results of Table 7, it can be seen that there is indeed regional heterogeneity in the implementation of the policy of the big data experimental area. Column ( 1 ) is the impact of the policy of the big data experimental area on the carbon emissions of the eastern counties. Although the results are negative, they are not significant, indicating that the implementation of the policy of the big data comprehensive experimental area has no impact on the eastern counties. The possible explanation is that China has always adopted the strategy of giving priority to the development of the eastern coastal region since the reform and opening up. The eastern region, such as Shanghai, is likely to have a certain ability to apply digital technology before the implementation of the policy in the experimental area. Compared with the eastern region, the policy coefficient of the central and western pilot areas is negative, and it is significant at the level of 5 %, indicating that the policy of the big data pilot area has a significant inhibitory effect on the carbon emissions of the counties in the central and western regions of China. The possible explanation is that the implementation of the policy of the big data pilot area began in Guizhou, western China.As the first national-level big data comprehensive pilot area, the first-mover advantage is more obvious. It has achieved remarkable results in data resource management, sharing, integration, application and innovation, and has accelerated the construction of digital provinces under the guidance of big data. Therefore, the development of the pilot area plays a positive role in reducing carbon emissions.
6.Conclusions and Policy implications
Based on the theoretical analysis of the development of big data to enable carbon emission reduction, taking the construction of national big data comprehensive experimental area as a quasi-natural experiment, based on the sample data at the county level from 2007 to 2021, the multi-period difference-in-differences model was used to empirically test the impact of big data development on county carbon emission reduction. The study found that the development of big data can significantly reduce the carbon emission intensity of China ‘s counties. Mechanism analysis shows that the development of big data mainly achieves carbon emission reduction by promoting technological progress and industrial upgrading. This study uses county data to refine the study of regional carbon emissions and provide empirical evidence for policy implementation.
First, we should fully implement the big data strategy and promote the deep integration of big data and all walks of life based on the construction practice of the pilot area. The empirical results show that the construction of the pilot area can significantly reduce the level of carbon emissions in the county, which is an important positive external effect of the big data strategy in the ecological environment and enriches the current hot digital economy. Therefore, under the policy background of carbon peak carbon neutralization and ‘ digital China ‘ formulated by the Chinese government, the construction of the pilot area should be taken as an important starting point to promote the development of the digital economy, promote the innovation and development of big data and other related industries, make full use of China ‘s data scale advantages, enrich data development and application scenarios, stimulate the potential of big data, constantly spawn new formats and new models, and realize the simultaneous improvement of quality and application level. In the areas that have been established as test areas, it is necessary to establish a long-term mechanism, consolidate the new kinetic energy of economic development, and amplify the radiation-driven role of the test area. In addition, it is necessary to strengthen the summary of the work results since the construction of the pilot area, and provide successful experience that can be replicated and promoted for the vigorous development of the national big data industry and the healthy operation of the data factor market.
Second, actively explore multiple paths to reduce carbon emissions in the pilot area, and promote the application of data innovation as the core to improve the effect of policy implementation. In the process of the construction of the pilot area, we should take digitalization as the guide to promote the development of digital inclusive finance and improve financial efficiency ; we should actively guide the integration of inclusive finance and green finance to help technological innovation and green development. At the same time, it is necessary to further promote the construction of a new generation of information infrastructure, promote the application of emerging technologies such as big data, cloud computing, and artificial intelligence, and give full play to the role of data as an innovation engine by building a digital innovation platform to promote technological innovation. In addition, for enterprises, they should provide more perfect service support for emerging enterprises, and at the same time, further precise measures should be taken to promote the development of big data integration, continuously integrate data, technology and scenarios in the real economy, promote the growth of productivity and drive the development of other industries, and promote the optimization and upgrading of industrial structure.
Third, China objectively has significant regional development differences. The government needs to fully consider the differences in the county ‘s own endowments and implement poorly according to local conditions.
Funding:
1.Key research Project of Anhui Federation of Social Science (2023CX094).
2.Key Research Project of Humanities and Social Sciences of Anhui ProvincialDepartment of Education (2022AH040230,2022AH051615,2023AH051829).
3.Anhui Provincial Planning Office Philosophy and Social Science Planning Project : Research on Farmers ‘ Climate Response Behavior and Green Production Transformation from the Perspective of ‘ Emission Reduction ‘ Synergy (AHSKYY2023D11)
References
- Sun, J., Dong, F., 2022. Decomposition of carbon emission reduction efficiency and potential for clean energy power: evidence from 58 countries. J. Clean. Prod. 363, 132312.
- Yu, B., Fang, D., Kleit, A.N., Kun, S., 2022a. Exploring the driving mechanism and the evolution of the low‐carbon economy transition: lessons from OECD developed countries. World Econ. 45, 2766–2795.
- Hao, Y., Ba, N., Ren, S.Y., et al., 2021. How does international technology spillover affect China’s carbon emissions? A new perspective through intellectual property protection. Sustain. Prod. Consum. 25, 577–590.
- He, Q.; Wei, F.; Deng, X.; Kong, F.; Li, C.; Yan, Z.; Qi, Y. Spatiotemporal Pattern of Carbon Productivity and Carbon Offset Potential in Chinese Counties. Sci. Total Environ. 2022, 846, 157153.
- Wang, X.; Yu, H.; Wu, Y.; Zhou, C.; Li, Y.; Lai, X.; He, J. Spatio-Temporal Dynamics of Carbon Emissions and Their Influencing Factors at the County Scale: A Case Study of Zhejiang Province, China. Land 2024, 13, 381.
- NBS. China Statistical Yearbook (County-Level); China Statistics Press: Beijing, China, 2018.
- S. Yong-Yong, M.A. Bei-Bei, D. Lan-Hai, X. Dong-Qian, X. Si-You, W. Peng-Tao
- Spatial-temporal pattern and formation mechanism of county urbanization on the Chinese Loess PlateauJ. Mt. Sci., 18 (2021), pp. 1093-1111.
- Begenau, J, Farboodi, M. and Veldkamp, L.,2018,“Big Data in Finance and the Growth of Large Firms”,Journal of Monetary Economics, 97: 71-87.
- Gardiner, A., Aasheim, C., Rutner, P. and Williams, S.,2018,“Skill Requirements in Big Data: A Content Analysis of Job Advertisements”, The Journal of Computer Information Systems, 58(4): 374-384
- Duygan-Bump, B., Levkov, A. and Montoriol-Garriga,J.,2015,“Financing Constraints and Unemployment: Evidence from the Great Reces⁃ sion”, Journal of Monetary Economics, 75: 89-105.
- Acemoglu, D. and Restrepo, P., 2019,“Automation and New Tasks: How Technology Displaces and Reinstates Labor”, Journal of Economic Perspectives, 33(2): 3-30.
- Kwon, O., Lee, N., Shin, B., 2014. Data quality management, data usage experience and acquisition intention of big data analytics. Int. J. Inf. Manag. 3, 387–394.
- Shin, D.H., Choi, M.J., 2015. Ecological views of big data: perspectives and issues. Telematics Inform. 32, 311–320.
- Wang, J.D., Dong, X.C., Dong, K.Y., 2022. How does ICT agglomeration affect carbon emis sions? The case of Yangtze River Delta urban agglomeration in China. Energy Econ. 111, 106107.
- Heo, P.S., Lee, D.H., 2019. Evolution of the linkage structure of ICT industry and its role in the economic system: the case of Korea. Inform. Technol. Dev. 25 (3), 424–454.
- Jacobson, L.S.; LaLonde, R.J.; Sullivan, D.G. Earnings losses of displaced workers. Am. Econ. Rev. 1993, 83, 685–709.
- Wang, L., & Shao, J. (2024). Digital economy and urban green development: A quasi-natural experiment based on national big data comprehensive pilot zone. Energy & Environment, 235-260.
- Salahuddin, M., Gow, J., 2016. The effects of internet usage, financial development and trade openness on economic growth in South Africa: a time series analysis. Telematics Inform. 33 (4), 1141–1154.
- Kwon, O., Lee, N., Shin, B., 2014. Data quality management, data usage experience and acquisition intention of big data analytics. Int. J. Inf. Manag. 3, 387–394.
- Shin, D.H., Choi, M.J., 2015. Ecological views of big data: perspectives and issues. Telematics Inform. 32, 311–320.
- Hampton, S., Strasser, A., Tewksbury, J., et al., 2013. Big data and the future of ecology. Front. Ecol. Environ. 11, 156–162.
- Dong, F., Li, Y.F., Gao, Y.J., et al., 2022. Energy transition and carbon neutrality: exploring the non-linear impact of renewable energy development on carbon emission efficiency in developed countries. Resour. Conserv. Recycl. 177, 106002.
- Dong, F., Li, Y.F., Gao, Y.J., et al., 2022. Energy transition and carbon neutrality: exploring the non-linear impact of renewable energy development on carbon emission efficiency in developed countries. Resour. Conserv. Recycl. 177, 106002.
- Dong, F., Li, Y.F., Qin, C., et al., 2022. Information infrastructure and greenhouse gas emission performance in urban China: a difference-in-differences analysis. J. Environ. Manag. 316, 115252.
- Wang, J.D., Dong, X.C., Dong, K.Y., 2022. How does ICT agglomeration affect carbon emissions? The case of Yangtze River Delta urban agglomeration in China. Energy Econ. 111, 106107differences analysis. J. Environ. Manag. 316, 115252.