Does the Integrated Development of High-End, Intelligent, and Green Manufacturing in China Influence Regional Dual Control of Carbon Emissions?—An Analysis Based on Impact Mechanisms and Spatial Effects

Abstract

Against the backdrop of the “dual carbon” goals, the transformation and upgrading of the manufacturing industry play a crucial role in achieving the dual objectives of controlling both total carbon emissions and carbon intensity. This study first defines the connotation of the integrated development of high-end, intelligent, and green manufacturing (referred to as “Three Modernization”) and constructs a conceptual framework illustrating its impact mechanisms on carbon emission dual control, with particular emphasis on intermediary pathways such as technological progress and energy structure optimization. Subsequently, based on provincial panel data from China covering the period 2009–2023, this paper employs fixed effects and spatial Durbin models to empirically examine the impact of the integrated development of “Three Modernization” on total carbon emissions and carbon emission intensity. The results show that the integration of high-end, intelligent, and green manufacturing significantly suppresses carbon emissions, with a more pronounced effect observed in economically developed regions. Mechanism tests further reveal that technological innovation enhances the application capacity of low-carbon technologies, while an increased share of clean energy usage effectively reduces reliance on fossil fuels, thereby indirectly facilitating the realization of the dual control targets. The spatial effect analysis indicates that the integration of “Three Modernization” exhibits significant spatial spillover effects, whereby regional synergies contribute to improved carbon reduction performance in neighboring areas. Furthermore, threshold model analysis confirms a notable nonlinear relationship moderated by technological complexity: when technological complexity is at a lower level, the emission-reduction effect of the “Three Modernization” integration is more substantial; however, once a certain threshold is exceeded, the marginal abatement effect diminishes, suggesting that in high-technology phases, the carbon-reduction efficiency of additional technological inputs declines. This nonlinear pattern indicates an inverted U-shaped relationship between the “Three Modernization” integration and carbon emission control. Therefore, differentiated dual control policies should be formulated to promote the region-specific integration of high-end, intelligent, and green development in manufacturing. This should be accompanied by continuous enhancement of technological innovation and green technology adoption, along with energy structure optimization, to ensure the sustainability of both total carbon emission control and intensity reduction.

1. Introduction

In the report of the 20th National Congress of the Communist Party of China, General Secretary Xi Jinping pointed out that high-quality development of the manufacturing industry is crucial to the high-quality development of our economy. We must promote the high-end, intelligent, and green development of the manufacturing industry. Currently, the manufacturing industry is not only a vital support for driving China’s high-quality economic development but also a major source of carbon emissions. Its transformation and upgrading are crucial for achieving the “dual carbon” goals. To implement the decisions and deployments of the Party Central Committee and the State Council, and accelerate the establishment of a “dual control” mechanism for carbon emissions in both total volume and intensity, the General Office of the State Council in the “Work Plan for Accelerating the Establishment of a Carbon Emission Dual Control System” clearly proposed the need to comprehensively establish a dual control system for carbon emissions. Therefore, systematically studying the role mechanism in the integrated development of the manufacturing industry’s “Three Modernization” (high-end, intelligent, green) on dual carbon control is not only of significant theoretical value, but also of practical significance for advancing China’s green transformation.

Currently, research on the integrated development of the manufacturing industry’s “Three Modernization” mainly focuses on the measurement of development levels and analysis of influencing factors. Peng Diyun (2020) constructed an evaluation system to measure the relationship between consumption upgrading and the high-end development of the manufacturing industry, using the PVAR model to analyze their dynamic interactions [1], but did not involve the dimensions of intelligence and greening. Wang Xiangjin (2018) studied the paths of high-end upgrading and carbon reduction in the manufacturing industry from a global value-chain perspective [2], but its path analysis was relatively static and lacked an integrated perspective. Research on green development is also relatively independent. Fan Qiufang (2023) used social network analysis to explore the spatial characteristics of energy-use greening [3], while Zhang Youzhi (2021) and Xiong Xiaolian (2023) used PSM-DID and synthetic control methods to assess the impact of carbon trading policies on the greening efficiency in the manufacturing industry [4,5], but did not integrate greening into the overall transformation framework of the manufacturing industry.

In addition, research on the dual control of carbon emissions has recently expanded to the dimensions of digital economy and technological innovation. Guo Lixiang (2024) found that the digital economy has an inverted U-shaped effect on carbon emissions [6], and Wang Jiao (2024) and Ma Yuefeng (2024) further analyzed the chain mediation effect of digital development on regional dual carbon performance and carbon emission intensity [7,8], but did not discuss the internal structural upgrading in the manufacturing industry. Foreign scholars focus more on the causal mechanisms between manufacturing digitalization and carbon reduction. Zhao and Sun (2024) and Li and Wang (2023), among others, empirically validated the role of intelligent manufacturing in promoting carbon reduction [9,10], and Geng and Liu (2024) explored the transformation paths from the perspective of green innovation [11], but overall, the analysis is still more focused on a single dimension.

Regarding the policy transition from energy consumption dual control to carbon emission dual control, research has mainly focused on indicator construction and policy simulation. For example, Cui Xiaowei (2024) constructed a “dual carbon” index model to analyze regional carbon neutrality paths [12], and Tang Lang (2024) conducted a comparative analysis from the perspective of policy evolution [13], but there is a lack of systematic discussions combined with the transformation of the manufacturing industry. Madu (2022) introduced a system dynamics model to analyze the long-term impact of the green transformation of the manufacturing industry on carbon dual control [14], but its simulation results lack an analysis of regional heterogeneity.

In summary, although existing literature has explored the manufacturing industry’s “Three Modernization” and carbon emission issues from various perspectives, there are still several shortcomings: First, there is limited research on the integrated development of “high-end, intelligent, and green manufacturing”; most existing studies focus on individual dimensions such as high-end, intelligent, and green, lacking a comprehensive perspective. Second, the role mechanisms of the integrated development of “Three Modernization” on dual carbon control are not deeply revealed, especially in terms of empirical testing of regional heterogeneity and spatial spillover effects. Third, there is insufficient identification of nonlinear relationships, and dynamic marginal changes in policy effects have not been considered. To fill these research gaps, this paper takes the Chinese manufacturing industry as the research object, constructs an index system for the integrated development of the “high-end, intelligent, and green manufacturing” and uses fixed effect models and spatial Durbin models to systematically assess their impact mechanisms on the total volume and intensity of carbon emissions, revealing regional differences and spatial spillover characteristics. The research results are expected to provide theoretical support for optimizing the green transformation paths in the manufacturing industry and empirical evidence for the regional fine-tuning of “dual control” policies.

2. Theoretical Analysis and Research Hypotheses

2.1. Definition of Key Concepts

2.1.1. Definition of the Integrated Development of High-End, Intelligent, and Green Manufacturing

High-end manufacturing refers to enhancing the technological sophistication, added value, and international competitiveness in the manufacturing sector through technological innovation, product upgrading, and management optimization [15]. It emphasizes technological advancements, including innovations in product technology, production processes, and management models, in order to produce high-value-added products with strong market competitiveness. Through technological innovation, enterprises can rapidly adapt to market dynamics, improve product quality, and maintain a competitive edge in the global market. Intelligent manufacturing refers to the application of information technology, automation, and internet technologies to achieve intelligent, automated, and networked production processes, thereby improving production efficiency [16]. Green manufacturing centers on energy conservation and emission reduction, adopting energy-efficient technologies and clean production processes to reduce energy consumption and emissions, enhance resource utilization efficiency, and lower production costs. It emphasizes recycling and resource recovery to ensure sustainable resource use [17].

The integrated development of high-end, intelligent, and green manufacturing (referred to as “Three Modernization” of manufacturing) seeks to integrate advanced technologies and innovative practices into the manufacturing sector. This approach aims to enhance the technological content and added value of products while simultaneously achieving the intelligent, automated, and green transformation of the production process. The integrated development framework emphasizes not only technological innovation and product quality enhancement but also the efficient use of resources and the minimization of environmental impacts. By incorporating strategies such as intelligent manufacturing, energy conservation, emission reduction, and resource recycling, this integrated approach seeks to enhance market competitiveness and drive transformation and upgrading in the manufacturing industry. Under this integrated development paradigm, the establishment of comprehensive environmental management systems and corporate social responsibility frameworks is crucial. This strengthens the monitoring and management of environmental impacts during the production process, aiming to achieve the simultaneous realization of economic, social, and environmental benefits.

2.1.2. Definition of the Dual Control of Carbon Emissions

The dual control of carbon emissions refers to the regulatory approach that simultaneously controls both the total carbon emissions and the carbon emission intensity [18]. Total carbon emissions control aims to set caps on the total amount of greenhouse gases that can be emitted by a region, industry, or enterprise within a specified time period, ensuring that carbon emissions do not grow uncontrollably. Carbon emission intensity control focuses on reducing carbon emissions per unit of economic output in order to achieve emission reduction targets. This approach emphasizes the decoupling of economic growth from carbon emissions, thereby driving improvements in energy efficiency and the optimization of industrial structures.

The core of the dual control framework lies in promoting economic growth while simultaneously reducing carbon emissions through technological innovation, energy efficiency improvements, and structural transformations. By establishing regional and sectoral carbon emissions caps, optimizing energy usage, and effectively preventing carbon emissions rebound, this framework encourages the development of a low-carbon economy and supports the achievement of sustainable green transformation and carbon neutrality goals.

2.2. Impact Mechanism of Integrated Development of Manufacturing “Three Modernization” on Dual Control of Carbon Emissions

The theory of industrial upgrading emphasizes that the optimization and upgrading of industrial structures contribute to carbon emission reduction. By improving technological capabilities and optimizing production processes, manufacturing industries can transition from high-pollution, high-energy consumption sectors to low-carbon, intelligent, and green industries, thereby achieving more efficient energy use and reducing carbon emissions [19,20]. The theory of technological diffusion posits that the widespread dissemination and adoption of technology can effectively promote the diffusion of low-carbon technologies, thereby reducing carbon emissions. In particular, the diffusion of green technologies enables manufacturing industries to enhance energy efficiency, reduce resource consumption, and promote the widespread adoption of low-carbon production methods [21,22].

Therefore, the integrated development of “Three Modernization” in the manufacturing industry may have a significant impact on dual carbon emission control. First, the integrated development of “Three Modernization” may have a direct impact on dual carbon control. Manufacturing enterprises are gradually transitioning toward technology-intensive, high-value-added sectors, reducing their reliance on energy-intensive industries, enhancing production efficiency, and minimizing energy waste. Second, the integrated development of “Three Modernization” indirectly reduces carbon emissions by promoting technological innovation and optimizing the energy consumption structure. This encourages enterprises to increase investment in research and development, thereby accelerating the application of low-carbon technologies. As a result, production efficiency is enhanced while emissions are reduced. Finally, the integrated development of “Three Modernization” optimizes the energy consumption structure by reducing reliance on fossil fuels and increasing the share of renewable energy, thereby facilitating the dual control of both carbon emissions total volume and intensity. The mechanism through which the integrated development of “Three Modernization” impacts dual carbon control is illustrated in Figure 1.

2.2.1. Direct Impact of the Integrated Development of “Three Modernization” on Dual Control of Carbon Emissions

The integrated development of “Three Modernization” has a direct and profound impact on the dual control of carbon emissions through high-end manufacturing, intelligent manufacturing, and green manufacturing.

First, high-end manufacturing implies a shift toward technology-intensive and high-value-added industries. By promoting innovations and breakthroughs in key technologies, such as new energy materials and low-carbon technologies, it enhances the technological capabilities of the entire industry chain. Furthermore, high-end manufacturing leads to a deep adjustment of the industrial structure, gradually reducing reliance on traditional high-energy-consuming and high-pollution sectors. This transition drives the manufacturing industry from a resource-consuming model to a technology-driven model, thereby significantly reducing carbon emission intensity. This transformation not only reduces the energy and resources consumed per unit of output, but also minimizes waste and unnecessary emissions through efficient production methods.

Second, the development of intelligent manufacturing relies on digital technologies, industrial internet, and artificial intelligence. By automating and smartening production processes, it greatly improves both production efficiency and energy utilization efficiency. Intelligent manufacturing technologies can precisely manage and optimize energy consumption during the production process and monitor and adjust emission data in real-time, ensuring that each production step operates at the optimal energy consumption level. This intelligent management reduces energy waste on the one hand and enhances the enterprise’s ability to monitor carbon emissions on the other, helping businesses better meet the requirements for controlling both total carbon emissions and emission intensity. In addition, intelligent technologies, through data analysis and forecasting capabilities, help enterprises optimize future production layouts, further reducing their carbon footprint.

Finally, green manufacturing directly addresses environmental issues within the manufacturing sector. By promoting clean production technologies and the application of renewable energy, it achieves a low-carbon and efficient use of energy. Green development not only requires enterprises to reduce harmful emissions during the production process, but also optimizes resource use through life-cycle management, emphasizing environmental protection and resource recycling from design, production, to waste disposal. Specifically, green manufacturing has a significant contribution to reducing total carbon emissions by improving energy efficiency and promoting renewable energy, which reduces reliance on fossil fuels and lowers overall carbon emissions. In this process, the concept of green manufacturing gradually permeates all stages of industry, driving low-carbon transformation and sustainable development in manufacturing.

Based on this, the following hypothesis is proposed:

Hypothesis 1.

The integrated development of “Three Modernization” has a significant negative impact on both total carbon emissions and carbon emission intensity. That is, the higher the level of integration in “Three Modernization”, the lower the total carbon emissions and carbon emission intensity.

2.2.2. Indirect Impact of the Integrated Development of “Three Modernization” on Dual Control of Carbon Emissions

(1) Mediating Effect of Technological Innovation

The integrated development of high-end, intelligent, and green manufacturing promotes the dual control of carbon emissions indirectly by enhancing technological innovation. Specifically, high-end manufacturing introduces advanced technologies, intelligent manufacturing applies industrial internet and artificial intelligence, and green manufacturing promotes clean production technologies. Together, they drive the development and application of low-carbon technologies, optimize energy use, reduce pollutant emissions, and improve resource utilization efficiency. Therefore, technological innovation plays a key mediating role between the integrated development of “Three Modernization” and the dual control of carbon emissions.

Based on this, the following hypothesis is proposed:

Hypothesis 2.

Technological innovation plays a significant mediating role between the integrated development of “Three Modernization” and the dual control of carbon emissions. Specifically, the integrated development of “Three Modernization” reduces total carbon emissions and carbon emission intensity indirectly by enhancing technological innovation.

(2) Mediating Effect of Energy Consumption Structure

The integrated development of “Three Modernization” also exerts a significant indirect impact on dual control of carbon emissions by improving technological innovation and optimizing the energy consumption structure, displaying a strong mediating effect. High-end manufacturing drives the manufacturing sector toward high-value-added and low-energy consumption directions, reducing reliance on traditional energy-intensive industries. Intelligent manufacturing improves energy utilization efficiency through technologies such as industrial internet and artificial intelligence. Green manufacturing promotes the application of clean energy and drives the energy structure toward a green, low-carbon transformation.

Based on this, the following hypothesis is proposed:

Hypothesis 3.

The energy consumption structure plays a significant mediating role between the integrated development of “Three Modernization” and the dual control of carbon emissions. Specifically, the integrated development of “Three Modernization” reduces total carbon emissions and carbon emission intensity indirectly by optimizing the energy consumption structure.

3. Variables Description, Model Construction, and Data Sources

3.1. Variables Description

3.1.1. Dependent Variables

Total Carbon Emissions (TCE) and Carbon Emission Intensity (CEI). The measurement of total carbon emissions follows the method in the study by Lü Jiehua (2024) [23]. First, relevant data are collected, including the consumption of different energy types (such as coal, oil, natural gas, etc.) and industrial carbon emissions data. Then, using national or international standards for emission factors, the consumption of different energy types is converted into carbon dioxide emissions. The emission factor refers to the amount of CO₂ produced per unit of energy consumed, typically expressed in tons of CO₂ per unit of energy (such as tons of coal). Finally, the total carbon emissions are calculated using the following Formula (1):

$$TCE={\displaystyle \sum \left(E_i\times E{F}_i\right)}$$

(1)

where $E_i$ represents the consumption of different energy types and $E{F}_i$ represents the corresponding emission factors.

The calculation of carbon emission intensity is based on the total carbon emission, as shown in Formula (2):

$$CEI=TCE/GDP$$

(2)

where TCE is the total carbon emissions, and GDP represents the regional gross domestic product, typically measured in 10,000 RMB. Carbon emission intensity is generally expressed in tons of CO₂ per 10,000 RMB of GDP.

3.1.2. Core Explanatory Variables

The core explanatory variable in this study is the level of integrated development of the manufacturing industry’s “Three Modernization”. Based on the connotation of “Three Modernization”, this study selects 11 indicators from three dimensions: high-end manufacturing, intelligent manufacturing, and green manufacturing to construct an indicator system for measuring the level of integrated development of China’s manufacturing “Three Modernization”. Capital intensity and industrial efficiency serve as secondary indicators for measuring the level of high-end manufacturing development, while the intelligent manufacturing foundation and intelligent manufacturing achievements serve as secondary indicators for measuring the level of intelligent manufacturing. Pollution emissions and environmental regulation are secondary indicators for measuring the level of green manufacturing development. The specific indicator system is shown in

Table 1. Comprehensive evaluation indicator system for the integrated development of “Three Modernization” in manufacturing.

Primary Indicator	Secondary Indicator	Tertiary Indicator	Indicator Explanation	Indicator Attribute
High-end Manufacturing (HE)	Capital Intensity	Manufacturing Capital–Labor Ratio (%)	Capital stock/Labor input	+
		Manufacturing Capital–Output Ratio (%)	Capital stock/Total output value	+
		Share of Fixed Assets in Manufacturing (%)	Total fixed assets/Total output value	+
	Industrial Efficiency	Industrial Cost–Profit Ratio (%)	Industrial profit/industrial cost	+
Intelligent Manufacturing (IT)	Foundation of Intelligence	Proportion of Research Personnel (%)	Research personnel/Total personnel	+
		Investment in Technological Innovation	R&D research and experimental funding	+
		Proportion of Automated Equipment Investment (%)	Automated equipment assets/Total assets	+
	Achievements of Intelligence	Degree of Technological Marketization	Technology market transaction volume	+
		New Product Output Ratio (%)	Output value of newly developed products/Regional GDP	+
Green Manufacturing (GI)	Pollution Emissions	Total Sulfur Dioxide Emissions	Reflects the level of air pollution	−
	Environmental Regulation	Industrial Pollution Control Intensity (%)	Total investment in industrial pollution control/Regional GDP	−

(note: in the “Indicator Attribute” column, “+” denotes a positive attribute, while “−” denotes a negative attribute).

To further calculate the integrated development level of China’s manufacturing “Three Modernization”, the entropy weight TOPSIS method is employed. The relevant calculation steps are as follows:

(1)

Standardization of Indicators:

$$\mathrm{Positive} \mathrm{indicators}:X_{tij}^{\prime }={\displaystyle \frac{X_{tij}-\min X_{tj}}{\max X_{tj}-\min X_{tj}}}$$

(3)

$$\mathrm{Negative} \mathrm{indicators}:X_{tij}^{\prime }={\displaystyle \frac{\max X_{tj}-X_{tij}}{\max X_{tj}-\min X_{tj}}}$$

(4)

In Formulas (1) and (2), the standardized values are represented by $X_{tij}^{\prime }$, where $X_{tij}$ represents the value of the $i$-th indicator in the $j$-th year province, and $\max X_{tj}$ and $\min X_{tj}$ represent the maximum and minimum values of the $j$-th indicator across all provinces for the corresponding year, respectively.

(2)

Quantification of Indicators and Calculation of Weights:

$$p_{tij}=X_{tij}^{\prime }/{\displaystyle \sum _{t=1}^{\theta }{\displaystyle \sum _{i=1}^m{X}_{tij}^{\prime }}}$$

(5)

(3)

Calculation of Information Entropy:

$$e_j=-{\displaystyle \frac{1}{\ln m}}{\displaystyle \sum _{i=1}^m{p}_{tij}}\ln p_{tij}$$

(6)

(4)

Calculation of Variability Coefficients:

$$a_j=1-e_j$$

(7)

(5)

Calculation of Indicator Weights:

$$W_j=a_j/{\displaystyle \sum _{j=1}^n{a}_j}$$

(8)

(6)

Calculation of Comprehensive Scores:

$$s_{ti}={\displaystyle \sum _{j=1}^{n}W}\times X_{tij}^{\prime }$$

(9)

The comprehensive index results for the integrated development of “Three Modernization” manufacturing in China are presented in

Table 2. Comprehensive index for the integrated development of “Three Modernization” manufacturing in China.

Region	2009	2011	2013	2015	2017	2019	2021	2023
Beijing	0.793	0.774	0.870	0.857	0.847	0.735	0.772	0.775
Tianjin	0.329	0.327	0.310	0.291	0.286	0.364	0.401	0.459
Hebei	0.212	0.185	0.214	0.240	0.230	0.231	0.272	0.197
Shanxi	0.118	0.151	0.186	0.234	0.138	0.213	0.187	0.198
Inner Mongolia	0.084	0.157	0.157	0.099	0.156	0.153	0.287	0.163
Liaoning	0.214	0.280	0.237	0.320	0.315	0.317	0.353	0.334
Jilin	0.173	0.173	0.241	0.146	0.170	0.189	0.123	0.123
Heilongjiang	0.143	0.162	0.229	0.159	0.205	0.229	0.170	0.167
Shanghai	0.526	0.484	0.470	0.455	0.448	0.540	0.536	0.591
Jiangsu	0.473	0.427	0.454	0.468	0.476	0.504	0.584	0.607
Zhejiang	0.432	0.369	0.402	0.418	0.394	0.480	0.512	0.539
Anhui	0.283	0.221	0.276	0.292	0.379	0.353	0.413	0.473
Fujian	0.309	0.338	0.261	0.245	0.297	0.318	0.340	0.272
Jiangxi	0.098	0.176	0.109	0.275	0.235	0.179	0.212	0.271
Shandong	0.274	0.266	0.355	0.336	0.304	0.329	0.440	0.467
Henan	0.136	0.144	0.074	0.215	0.095	0.166	0.268	0.135
Hubei	0.318	0.268	0.351	0.326	0.367	0.400	0.372	0.420
Hunan	0.284	0.285	0.307	0.236	0.252	0.351	0.270	0.424
Guangdong	0.485	0.475	0.435	0.433	0.518	0.460	0.644	0.632
Guangxi	0.184	0.173	0.208	0.196	0.203	0.115	0.211	0.138
Hainan	0.079	0.161	0.144	0.181	0.186	0.207	0.194	0.173
Chongqing	0.309	0.300	0.284	0.373	0.314	0.295	0.363	0.286
Sichuan	0.220	0.266	0.241	0.214	0.276	0.335	0.356	0.262
Guizhou	0.183	0.147	0.267	0.099	0.178	0.179	0.116	0.118
Yunnan	0.195	0.170	0.175	0.308	0.090	0.198	0.101	0.233
Shaanxi	0.166	0.180	0.257	0.223	0.203	0.282	0.257	0.285
Gansu	0.126	0.169	0.176	0.118	0.169	0.174	0.227	0.217
Qinghai	0.160	0.137	0.220	0.142	0.182	0.093	0.144	0.179
Ningxia	0.224	0.200	0.085	0.145	0.184	0.151	0.191	0.260
Xinjiang	0.175	0.152	0.167	0.097	0.136	0.215	0.079	0.122

(note: due to space limitations in this paper, only data for select years are presented here).

3.1.3. Mechanism Variables

Technological Innovation Level (TIL) and Proportion of Clean Energy Consumption (CES). This study adopts the method outlined by Lu Jingyu (2024) [24], and based on data published by the State Intellectual Property Office of China, it examines relevant patents from the Chinese manufacturing industry from 2009 to 2023. These patents are used as a metric for assessing the technological innovation level in the “Three Modernization” integration of manufacturing. For the clean energy consumption ratio, based on the manufacturing terminal energy consumption data and the International Energy Agency’s classification of clean energy, the ratio of clean energy consumption to total energy consumption is used as the mechanism variable. Clean energy is defined to include natural gas, liquefied natural gas, and thermal energy. As different energy types have different measurement units, energy consumption is standardized and converted into an equivalent amount of standard coal for statistical purposes.

3.1.4. Control Variables

Referring to the study by Guo Lixiang (2024) [6], this paper selects economic development level (FDL) and industrial structure (ISD) as control variables. Considering that marketization level, government regulation, environmental regulation, scientific and technological support, and public attention may also affect the “Three Modernization” integrated development of the manufacturing industry as well as the dual control of carbon emissions, additional control variables are introduced, including market factors (MAR), government factors (GAC), environmental regulation (ENV), and public attention (PAB). Specifically, the economic development level is measured by the logarithm of per capita GDP, industrial structure is represented by the logarithm of the number of manufacturing enterprises, marketization is measured by the ratio of the secondary industry to GDP, government factors are measured by the proportion of local fiscal expenditure to GDP, environmental regulation is indicated by the comprehensive utilization rate of industrial solid waste, scientific and technological support is measured by the proportion of R&D expenditure to GDP, and public attention is represented by the logarithm of annual per capita water consumption.

The descriptive statistical results of the variables are presented in

Table 3. Descriptive statistics of variables.

Variable		Definition	Maximum Value	Minimum Value	Mean Value	Standard Deviation
Explained Variables	Total Carbon Emissions (TCE)	Total industrial carbon emissions (million tons)	7.712	3.189	5.645	0.791
	Carbon Emission Intensity (CEI)	The ratio of total carbon emissions to industrial GDP	2.929	0.333	1.446	0.598
Core Explanatory Variables	“Three Modernization” Integration Level (HIG)	Composite index of “Three Modernization” integration level	0.885	0.074	0.283	0.153
Mediating Variables	Technological Innovation Level (TIL)	Number of invention patents (count), log-transformed	14.820	5.429	10.096	1.700
	Clean Energy Share (CES)	The ratio of clean energy consumption to total energy usage	0.832	0.101	0.269	0.113
Instrumental Variables	Coal Share in Primary Energy (CEC)	The ratio of coal consumption to primary energy consumption	0.830	0.100	0.383	0.165
Control Variables	Economic Development Level (FDL)	Logarithm of GDP per capita	13.204	9.085	10.878	0.721
	Industrial Structure (ISD)	Logarithm of the number of manufacturing enterprises	12.083	5.814	8.968	1.263
	Marketization Level (MAR)	The ratio of secondary industry output to GDP	0.615	0.158	0.437	0.088
	Government Regulation (GAC)	Proportion of local fiscal expenditure to GDP	0.695	0.087	0.247	0.015
	Environmental Regulation (ENV)	Comprehensive utilization rate of industrial solid waste	1.108	0.269	0.665	0.196
	Public Attention (PAB)	Logarithm of per capita annual water consumption (tons)	8.503	5.082	6.098	0.599
	Energy Consumption (EEC)	Logarithm of electricity consumption by manufacturing enterprises (10,000 kWh)	17.774	11.634	15.578	1.050

.

3.2. Model Construction

3.2.1. Benchmark Regression Model

To investigate the impact of the “Three Modernization” integration of manufacturing on the “Dual Control of Carbon Emissions”, this study constructs a provincial threshold-time, two-way fixed-effects model:

$$TC{E}_{it}=\alpha +\beta HI{G}_{it}+\gamma C{V}_{it}+CityFE+YearFE+{\epsilon }_{it}$$

(10)

$$CE{I}_{it}=\alpha +\beta HI{G}_{it}+\gamma C{V}_{it}+CityFE+YearFE+{\epsilon }_{it}$$

(11)

In the above formula, $CV$, $CityFE$, and $YearFE$ are the set of control variables, province and region fixed effects, year fixed effects, and $\epsilon$ is a random disturbance term.

3.2.2. Impact Mechanism Model

To further explore the mechanism through which technological innovation level and the proportion of clean energy consumption in the “Three Modernization” integration of manufacturing affect the “Dual Control of Carbon Emissions”, this study draws on the research by Jiang Ting (2022) [25] and constructs the following mechanism effect model:

$$TI{L}_{it}={\beta }_0+{\beta }_{1}HI{G}_{it}+{\beta }_{i}C{V}_{it}+CityFE+YearFE+{\epsilon }_{it}$$

(12)

$$CE{S}_{it}={\beta }_0+{\beta }_{1}HI{G}_{it}+{\beta }_{i}C{V}_{it}+CityFE+YearFE+{\epsilon }_{it}$$

(13)

In these formulas, ${\beta }_1$ is the marginal estimated coefficient for the integration development level of “Three Modernization” in the manufacturing industry to the mechanism variable.

3.2.3. Spatial Econometric Model

Given that the “Three Modernization” integration of manufacturing has spatial spillover effects, it is hypothesized that the dual control of carbon emissions in one region may influence other regions. To explore the spatial effects of the “Three Modernization” integration of manufacturing on the dual control of carbon emissions, this study refers to the research by Sun Chang (2019) [26] and employs a spatial Durbin model (SDM) with two-way fixed effects as the spatial econometric model for this study.

The spatial Durbin model (SDM) extends traditional spatial econometric models by incorporating spatial lags of both the explained and explanatory variables, enabling a comprehensive depiction of spatial spillover effects. This model not only considers the impact of a region’s own factors on the outcome variable, but also includes the influence of surrounding regions, thus more accurately reflecting the interconnections and spatial interactions between regions. The SDM is suitable for examining the complexity of spatial transmission paths of policy, economic, or technological factors and effectively identifies the differences between direct and indirect (spillover) effects.

The specific model is as follows:

$$\begin{array}{l}TC{E}_{it}={\gamma }_0+\rho W\times TC{E}_{it}+{\gamma }_{1}HI{G}_{it}+{\lambda }_{1}W\times HI{G}_{it}+{\gamma }_{i}C{V}_{it}\\ +{\lambda }_{i}W\times C{V}_{it}+CityFE+YearFE+{\epsilon }_{it}\end{array}$$

(14)

$$\begin{array}{l}CE{I}_{it}={\gamma }_0+\rho W\times CE{I}_{it}+{\gamma }_{1}HI{G}_{it}+{\lambda }_{1}W\times HI{G}_{it}+{\gamma }_{i}C{V}_{it}\\ +{\lambda }_{i}W\times C{V}_{it}+CityFE+YearFE+{\epsilon }_{it}\end{array}$$

(15)

In these formulas, $W$, $W\times HI{G}_{it}$, $W\times TC{E}_{it}$, and $W\times CE{I}_{it}$ are the spatial weight matrix, the spatial lag term of the integration development level of “Three Modernization” in the manufacturing industry, the spatial lag term of the total carbon emissions, the spatial lag term of the carbon emission intensity, and the spatial lag term of the carbon emission intensity, respectively.

3.2.4. Panel Threshold Model

The effect of the “Three Modernization” integration in manufacturing on the dual control of carbon emissions may be influenced by technological complexity. Therefore, this study constructs a panel threshold model to explore the nonlinear dynamic effect of the “Three Modernization” integration on the dual control of carbon emissions, as shown in the following panel threshold model:

$$\begin{array}{l}TC{E}_{it}={\alpha }_0+{\alpha }_{1}TC{X}_{it}\times I\left(TCX\le \theta \right)+{\alpha }_{2}TC{X}_{it}\times I\left(TCX>\theta \right)\\ +{\alpha }_{i}CV+CityFE+YearFE+{\epsilon }_{it}\end{array}$$

(16)

$$\begin{array}{l}CE{I}_{it}={\alpha }_0+{\alpha }_{1}TC{X}_{it}\times I\left(TCX\le \theta \right)+{\alpha }_{2}TC{X}_{it}\times I\left(TCX>\theta \right)\\ +{\alpha }_{i}CV+CityFE+YearFE+{\epsilon }_{it}\end{array}$$

(17)

In the above formula, TCX represents the threshold variable for technological complexity, $\theta$ is the threshold value, $I\left(*\right)$ is the indicator function, and ${\alpha }_1$ and ${\alpha }_2$ are the impact coefficients for different threshold intervals.

3.3. Data Sources

This study selected the 2009 to 2023 panel data from 30 provinces in China. Given the continuity and availability of the data, Tibet and the Hong Kong, Macau, and Taiwan regions are excluded. Specifically, Tibet’s economy is primarily based on agriculture and resource extraction, with a relatively small manufacturing sector. Energy consumption and carbon emission levels in Tibet significantly differ from other provinces. Furthermore, due to Tibet’s unique geographical environment and socio-economic development level, its carbon emissions show weaker correlations with those of other provinces. Excluding Tibet’s data helps ensure the reliability of the research results. The Hong Kong, Macau, and Taiwan regions have considerable differences in industrial structure, especially as Hong Kong and Macau are highly internationalized service-oriented economies, with industries focused on finance, tourism, and high-end services. Their carbon emissions differ greatly from those of China’s mainland provinces, which are primarily based on manufacturing. Taiwan has an independent economic system, and its industrial development and energy structure differ from the mainland’s, which is why it is also excluded from the data analysis. Excluding data from these regions does not affect the generalizability of the research conclusions.

The relevant data sources include the China Statistical Yearbook, China Urban Yearbook, China Ecological Environment Statistical Yearbook, China Environmental Statistical Yearbook, Patent Statistical Yearbook Compilation, Technology Fund Input Statistical Bulletin, and the China Energy Statistical Yearbook. Some data have missing values, which were filled using linear interpolation.

4. Empirical Analysis

4.1. Baseline Regression

Using STATA16 software, a baseline regression analysis was conducted on the panel data. According to the baseline regression results in

Table 4. Baseline regression results.

	(1) Benchmark Regression		(2) Benchmark Regression
	TCE	CEI	TCE	CEI
HIG	−2.099 *** (−12.35)	−2.473 *** (−17.23)	−1.317 *** (−7.42)	−1.140 *** (−6.90)
Control Variables	NO	NO	YES	YES
Province Fixed	YES	YES	YES	YES
Year Fixed	YES	YES	YES	YES
Constant Term	−0.847 *** (−2.35)	2.145 *** (46.53)	0.345 *** (0.83)	3.840 *** (9.87)
N	450	450	450	450
Adj R2	0.677	0.397	0.768	0.649

Note: *** indicate 1% significance levels. The data in brackets are t values.

, the integrated development of “Three Modernization” in the manufacturing industry has a significant negative impact on both total carbon emissions and carbon emission intensity.

In the model without control variables, the regression coefficient for total carbon emissions is −2.099, indicating that as the level of integration of “Three Modernization” in the manufacturing industry increases, total carbon emissions significantly decrease. After including control variables, the regression coefficient for total carbon emissions changes to −1.317. Similarly, the integrated development of “Three Modernization” in the manufacturing industry also exhibits a significant negative relationship with carbon emission intensity. In the model without control variables, the regression coefficient is −2.473. After the inclusion of control variables, although the coefficient increases to −1.140, it remains strongly significant, suggesting that the integrated development of “Three Modernization” not only effectively reduces total carbon emissions, but also lowers carbon emission intensity per unit of economic output.

As the level of “Three Modernization” in manufacturing improves, the industry’s resource consumption and negative environmental impact decrease, further promoting the effect of energy conservation and emission reduction. The introduction of control variables enhances the explanatory power of the model, particularly for the regression model of total carbon emissions, indicating that the control variables play a crucial role in the model’s explanation.

With the inclusion of fixed effects for provinces and years, the regression results are robust and reflect the impact of regional and temporal variations on carbon emissions. These regression results suggest that the integrated development of “Three Modernization” in the manufacturing industry plays a key role in achieving the dual control of carbon emissions, thus supporting Hypothesis 1.

4.2. Robustness Test

To make the research findings more robust, this study replaces the dependent variables, substituting total carbon emissions and carbon emission intensity with the carbon emissions per unit of manufacturing enterprise. This helps to further examine whether the conclusions still hold after replacing the dependent variables. According to the data in column (1) of

Table 5. Robustness test.

	(1) Replacing Dependent Variable	(2) Instrumental Variable Method
	CEC	TCE	CEI
HIG	−1.836 *** (−9.47)	−20.158 *** (−3.57)	−24.986 *** (−2.14)
Control Variables	YES	YES	YES
Province Fixed	YES	YES	YES
Year Fixed	YES	YES	YES
Constant Term	4.862 *** (10.64)	−8.464 *** (−3.01)	0.048 (0.02)
N	450	450	450
Adj R2	0.769	0.435	0.540

Note: *** indicate 1% significance levels. The data in brackets are t values.

, after replacing the dependent variables, the development of “Three Modernization” in manufacturing has a significant negative impact on the carbon emissions per unit of manufacturing enterprises. Specifically, the regression coefficient for the level of “Three Modernization” in manufacturing is −1.836, and it is significant at the 1% level. These findings suggest that as the level of “Three Modernization” improves, the carbon emissions per unit of manufacturing enterprises decrease significantly. Therefore, the integration of “Three Modernization” in manufacturing has a significant effect on improving resource utilization efficiency and reducing carbon emissions. Additionally, the introduction of control variables significantly enhances the explanatory power of the model, indicating that the model has strong explanatory power regarding changes in the carbon emissions per unit of manufacturing enterprises.

Carbon emission dual control and the integration of “Three Modernization” in manufacturing may interact, so it is necessary to address endogeneity issues. To manage the potential endogeneity problem, this study uses the instrumental variable method to ensure the robustness of the results. The coal consumption share of primary energy (CEC) in each province is selected as the instrumental variable for the level of “Three Modernization” in manufacturing. Compared to the results of replacing the dependent variables, the regression results from the instrumental variable method show that the impact of “Three Modernization” in manufacturing on carbon emissions is strengthened. According to the data in column (2) of Table 5, the effect of “Three Modernization” in manufacturing on total carbon emissions is significantly negative at the 1% level, with a regression coefficient of −20.158, indicating that under the instrumental variable method, the integration of “Three Modernization” in manufacturing also has a significant inhibitory effect on total carbon emissions. For carbon emission intensity, the impact of “Three Modernization” in manufacturing is also significantly negative at the 1% significance level, with a regression coefficient of −24.986. This suggests that the level of “Three Modernization” in manufacturing has a strong negative impact on reducing carbon emission intensity, and as the level of integration of “Three Modernization” improves, carbon emission intensity will decrease.

4.3. Heterogeneity Analysis

(1) Regional Heterogeneity Analysis

To study the differences in the impact of integrating “Three Modernization” in manufacturing on the dual control of carbon emissions across different regions, this study follows the research of Fang Yan (2024) [27] and divides China into four major regions: eastern, central, western, and northeastern. Heterogeneity analysis is then conducted for each region. According to the regression results in columns (1) to (4) of

Table 6. Heterogeneity test results.

	(1) Regional Heterogeneity
	(1) Eastern Region		(2) Central Region		(3) Western Region		(4) Northeastern Region
Variable	TCE	CEI	TCE	CEI	TCE	CEI	TCE	CEI
HIG	−0.902 *** (−3.24)	−0.851 *** (−2.23)	−1.937 *** (−4.33)	−1.108 *** (−2.60)	−2.216 *** (−4.69)	−1.830 *** (−4.12)	−1.006 *** (−2.93)	−0.806 ** (−2.23)
Control Variables	YES	YES	YES	YES	YES	YES	YES	YES
Province Fixed	YES	YES	YES	YES	YES	YES	YES	YES
Year Fixed	YES	YES	YES	YES	YES	YES	YES	YES
Constant Term	−1.046 * (1.64)	4.352 *** (7.21)	0.405 (0.59)	3.277 *** (5.03)	−0.773 (−1.01)	3.882 *** (5.40)	−0.147 (−0.22)	2.452 *** (4.71)
N	150	150	90	90	165	165	45	45
Adj R2	0.922	0.723	0.604	0.638	0.768	0.699	0.941	0.762
	(2) Economic Development-Level Heterogeneity				(3) Urban-Type Heterogeneity
	(1) Economically Developed Regions		(2) Economically Less Developed Regions		(1) Resource-Based Cities		(2) Non-Resource-Based Cities
Variable	TCE	CEI	TCE	CEI	TCE	CEI	TCE	CEI
HIG	−1.239 *** (−3.97)	−0.755 *** (−2.39)	−1.200 *** (−3.43)	−1.212 *** (−3.70)	−1.892 *** (−7.49)	−1.812 (−7.86)	−0.735 ** (−2.57)	−0.061 (−0.14)
Control Variables	YES	YES	YES	YES	YES	YES	YES	YES
Province Fixed	YES	YES	YES	YES	YES	YES	YES	YES
Year Fixed	YES	YES	YES	YES	YES	YES	YES	YES
Constant Term	1.667 *** (2.07)	4.123 *** (5.08)	−0.115 (−0.22)	3.676 *** (7.62)	−0.460 (−0.10)	3.391 *** (8.45)	0.951 (0.94)	2.991 *** (2.84)
N	150	150	300	300	420	420	30	30
Adj R2	0.844	0.625	0.751	0.508	0.758	0.656	0.856	0.839
	(4) Industrial Structure Heterogeneity
	(1) Comprehensive Manufacturing Strong Provinces		(2) High-End Manufacturing and Tech Innovation Centers		(3) Traditional Manufacturing Base		(4) Characteristic Industries and Emerging Manufacturing
Variable	TCE	CEI	TCE	CEI	TCE	CEI	TCE	CEI
HIG	−1.562 *** (−2.21)	−2.348 *** (−4.32)	−0.892 *** (−3.75)	−0.038 (−0.15)	−0.311 ** (−0.68)	−1.947 *** (−4.07)	−0.513 * (−1.79)	−0.127 (−0.07)
Control Variables	YES	YES	YES	YES	YES	YES	YES	YES
Province Fixed	YES	YES	YES	YES	YES	YES	YES	YES
Year Fixed	YES	YES	YES	YES	YES	YES	YES	YES
Constant Term	0.216 (0.22)	1.304 * (1.72)	−0.492 (−0.91)	0.501 (1.00)	1.342 * (1.77)	0.750 (1.18)	−3.981 *** (−5.86)	−0.232 (−0.39)
N	75	75	105	105	165	165	105	105
Adj R2	0.780	0.585	0.870	0.424	0.548	0.324	0.902	0.665

Note: ***, **, and * indicate 1%, 5%, and 10% significance levels, respectively. The data in brackets are t values.

on regional heterogeneity, the integration of “Three Modernization” in manufacturing has a significant negative impact on both total carbon emissions and carbon emission intensity, and this impact is significant at the 1%, 5%, and 10% levels. This suggests that the integration of “Three Modernization” in manufacturing in the eastern region can effectively reduce carbon emissions and carbon emission intensity. The regional heterogeneity analysis reveals that the impact of integrating “Three Modernization” on carbon emissions and carbon emission intensity varies significantly across regions.

In the western region, integrating “Three Modernization” has a strong inhibitory effect on both total carbon emissions and carbon emission intensity, indicating that despite the region’s relatively traditional industrial structure, recent efforts to promote industrial upgrading through advanced, intelligent, and green methods have achieved significant carbon reduction results, likely due to policy support and faster diffusion of green technologies. In the central region, integrating “Three Modernization” also has a clear inhibitory effect on carbon emissions, particularly concerning carbon emission intensity, showing that this region has made solid progress in improving energy efficiency per unit of output.

Although the impact of “Three Modernization” on total carbon emissions in the northeastern region is relatively limited, it still has a certain inhibitory effect on carbon emission intensity, suggesting that there is still potential for energy structure optimization and the transformation of old industrial bases. In contrast, while the eastern region has a higher level of integration of “Three Modernization”, its impact on total carbon emissions and intensity is not as strong. This may be due to the fact that the region’s industries have already reached a stable and high-end level, leaving limited room for marginal improvements, or because the integration of “Three Modernization” has not fully translated into carbon reduction benefits in the context of a high concentration of energy-intensive industries.

(2) Heterogeneity Analysis of Economic Development Level

To study the impact of integrating “Three Modernization” in manufacturing on total carbon emissions and carbon emission intensity at different levels of economic development, this study follows the research of Song Jiaying (2023) [28] and divides China into economically developed and economically less developed regions for heterogeneity analysis. The results show that integrating “Three Modernization” in manufacturing has a significant negative impact on both total carbon emissions and carbon emission intensity in both economically developed and less developed regions, but with notable regional differences.

According to column (1) of Table 5 on the economic development-level heterogeneity, the regression coefficients for total carbon emissions and carbon emission intensity in economically developed regions are −1.239 and −0.755, respectively, indicating that the integration of “Three Modernization” in manufacturing has achieved significant results in reducing carbon emissions and lowering carbon emission intensity in these regions. In column (2), for economically less developed regions, the regression coefficients for total carbon emissions and carbon emission intensity are −1.200 and −1.212, respectively, showing that integrating “Three Modernization” in manufacturing also has a suppressive effect on carbon emissions in less developed areas.

The heterogeneity regression results suggest that the effect of “Three Modernization” in manufacturing on total carbon emissions is stronger in economically developed regions than in economically less developed regions. This is because developed regions possess higher technological capabilities, ample innovation resources, and better policy support systems, enabling these regions to more effectively promote industrial transformation and upgrading, thereby improving resource utilization efficiency, optimizing energy structures, and reducing carbon emissions and intensity.

(3) Urban-Type Heterogeneity Analysis

There are significant differences in resource endowments, economic structures, and environmental pressures between resource-based cities and non-resource-based cities, and these differences may lead to varying carbon reduction effects from integrating “Three Modernization” in manufacturing in different types of cities. To further explore the impact of “Three Modernization” on carbon emissions in cities with different resource endowments, this study refers to the “National Sustainable Development Plan for Resource-Based Cities” by the State Council and divides China into resource-based and non-resource-based cities for heterogeneity analysis.

Column (1) of Table 6 on urban-type heterogeneity shows the regression results for integrating “Three Modernization” in manufacturing in resource-based cities. The regression coefficients for total carbon emissions and carbon emission intensity are −1.892 and −1.812, respectively, indicating that integrating “Three Modernization” in manufacturing in these cities can effectively reduce carbon emissions. In column (2), for non-resource-based cities, the regression coefficients for total carbon emissions and carbon emission intensity are −0.735 and −0.061, respectively, showing that integrating “Three Modernization” in manufacturing in these cities can also effectively reduce carbon emissions.

Heterogeneity analysis indicates that the control effect on both total carbon emissions and carbon emission intensity is significantly stronger in resource-based cities than in non-resource-based cities. This is because resource-based cities, having long relied on resource extraction and extensive development, have accumulated high carbon emission baselines and greater environmental pressures, which force them to more actively promote the transformation of their manufacturing sectors through high-end, intelligent, and green upgrades as part of their sustainable development efforts. With increased policy support and technological investments, “Three Modernization” has rapidly advanced in resource-based cities, resulting in more significant carbon reduction space and marginal effects. In contrast, non-resource-based cities have more diverse industrial structures, with some regions already achieving a high level of industrial upgrading, resulting in lower carbon emission intensity. As a result, the marginal improvement effect of “Three Modernization” in these cities is relatively limited. Moreover, resource-based cities often experience stronger policy interventions and resource integration during the green transformation process, which further amplifies the carbon emission reduction effects of “Three Modernization”.

(4) Heterogeneity Analysis of Industrial Structure

The integration of “Three Modernization” in manufacturing shows significant inhibitory effects on the dual control of carbon emissions, exhibiting heterogeneity under different industrial structures. Provinces with comprehensive manufacturing industries and high-end manufacturing and technology innovation centers have significantly reduced total carbon emissions and carbon emission intensity through technological innovation and green transformation. In contrast, traditional manufacturing bases and regions with specialized industries have relatively limited carbon reduction effects due to weak technological foundations and short industrial chains.

Therefore, promoting green development of manufacturing should be tailored to local conditions with differentiated policies. It is recommended to strengthen green and low-carbon technology transformation; improve energy utilization efficiency; accelerate the green and low-carbon transformation and upgrading of traditional industries; optimize industrial structures; promote the green, low-carbon, and high-quality development of emerging industries; and establish a comprehensive policy system for green development. These measures will help promote the high-quality development of manufacturing and achieve the dual control targets for carbon emissions.

4.4. Impact Mechanism Analysis

Table 7. Impact mechanism test results.

Variable	(1) TCE	(2) CEI	(3) TIL	(4) CES	(5) TCE	(6) CEI	(7) TCE	(8) CEI
HIG	−1.317 *** (−7.42)	−1.140 *** (−6.90)	0.399 *** (1.96)	0.272 *** (7.96)	−1.334 *** (−7.48)	−1.051 *** (−6.58)	−0.648 *** (−3.87)	−0.544 *** (−3.45)
TIL	-	-	-	-	−0.043 *** (−1.03)	−0.223 *** (−6.02)	-	-
CES	-	-	-	-	-	-	−2.455 *** (−11.26)	−2.188 *** (−10.66)
Control Variables	YES	YES	YES	YES	YES	YES	YES	YES
Province Fixed	YES	YES	YES	YES	YES	YES	YES	YES
Year Fixed	YES	YES	YES	YES	YES	YES	YES	YES
Constant Term	0.345 *** (0.83)	3.840 *** (9.87)	−5.207 *** (−10.83)	−0.258 *** (−3.20)	0.567 (1.21)	2.677 *** (6.35)	−0.288 *** (−0.77)	3.276 *** (9.32)
N	450	450	450	450	450	450	450	450
Adj R2	0.768	0.649	0.934	0.576	0.768	0.675	0.820	0.721

Note: *** indicate 1% significance levels. The data in brackets are t values.

presents the results of the mechanism tests. The regression results in columns (3) and (4) indicate that the “Three Modernization” integrated development of the manufacturing industry has a significant positive impact on the improvement of technological innovation level and the use of clean energy, with regression coefficients of 0.399 and 0.272, respectively. These findings further verify the validity of the relevant hypothesis, namely that the enhancement of the “Three Modernization” integrated development level in manufacturing can significantly boost technological innovation and optimize the energy consumption structure.

Columns (5) and (6) show the regressions of the “Three Modernization” integration level and the technological innovation level on total carbon emissions and carbon emission intensity. The regression coefficients of technological innovation level on total carbon emissions and carbon emission intensity are both negative, at −0.043 and −0.223, respectively. This result suggests that the “Three Modernization” integration in manufacturing can indirectly facilitate effective carbon emission control by enhancing regional technological innovation capacity, thereby supporting Hypothesis 2.

Columns (7) and (8) display the regression results of the “Three Modernization” integration level and the proportion of clean energy use on total carbon emissions and carbon emission intensity. The regression coefficients of clean energy use on total carbon emissions and carbon emission intensity are both negative, at −2.455 and −2.188, respectively, indicating that clean energy use also significantly contributes to the dual control of carbon emissions, thereby confirming Hypothesis 3.

4.5. Spatial Effect Test

Considering the substantial interregional interactions, solely relying on the regression coefficients of the spatial Durbin model may lead to bias. Therefore, this paper employs the partial differential method proposed by Wu Junqian (2023) [29] to decompose the estimation results. The results show that integrating “Three Modernization” in manufacturing has a significant direct and indirect suppressive effect on total carbon emissions and carbon emission intensity. Specifically, the direct effects are −0.698 and −1.313, indicating that the local “Three Modernization” integration significantly promotes the dual control of carbon emissions. The indirect effects are −3.207 and −1.081, suggesting that neighboring regions also benefit from the improvement in the “Three Modernization” integration level. The overall emission reduction effects are −3.905 and −2.394, highlighting the significant impact under the combined regional and cross-regional effects.

To ensure the robustness of the regression results, this study replaced the spatial adjacency matrix with spatial economic distance and spatial geographic distance matrices. According to the regression results from columns (4) to (9) in

Table 8. Spatial econometric SDM model effect decomposition.

	Based on the Spatial Adjacency Nesting Matrix
	(1) Direct Effect		(2) Indirect Effect		(3) Total Effect
Variable	TCE	CEI	TCE	CEI	TCE	CEI
HIG	−0.698 *** (−2.16)	−1.313 *** (−8.30)	−3.207 *** (−6.49)	−1.081 *** (−3.61)	−3.905 *** (−6.616)	−2.394 *** (−7.071)
Control Variables	YES	YES	YES	YES	YES	YES
Province Fixed	YES	YES	YES	YES	YES	YES
Year Fixed	YES	YES	YES	YES	YES	YES
N	450	450	450	450	450	450
Adj R2	0.785	0.736	0.785	0.736	0.785	0.736
Log_likelihood	−478.701	−143.485	−478.701	−143.485	−478.701	−143.485
	Based on the Spatial Economic Distance Matrix
	(4) Direct Effect		(5) Indirect Effect		(6) Total Effect
Variable	TCE	CEI	TCE	CEI	TCE	CEI
HIG	−0.869 *** (−2.30)	−0.822 *** (−2.24)	−2.369 *** (−2.40)	−1.479 *** (−1.48)	−3.238 *** (−3.067)	−2.301 *** (−2.160)
Control Variables	YES	YES	YES	YES	YES	YES
Province Fixed	YES	YES	YES	YES	YES	YES
Year Fixed	YES	YES	YES	YES	YES	YES
N	450	450	450	450	450	450
Adj R2	0.790	0.715	0.790	0.715	0.790	0.790
Log_likelihood	−146.402	−88.620	−146.402	−88.620	−146.402	−88.620
	Based on the Spatial Geographic Distance Matrix
	(7) Direct Effect		(8) Indirect Effect		(9) Total Effect
Variable	TCE	CEI	TCE	CEI	TCE	CEI
HIG	−0.008 *** (−0.07)	−0.275 *** (−1.54)	−0.133 *** (−0.37)	−0.911 *** (−2.26)	−0.141 *** (−0.377)	−1.186 *** (−2.689)
Control Variables	YES	YES	YES	YES	YES	YES
Province Fixed	YES	YES	YES	YES	YES	YES
Year Fixed	YES	YES	YES	YES	YES	YES
N	450	450	450	450	450	450
Adj R2	0.855	0.450	0.855	0.450	0.855	0.450
Log_likelihood	278.257	67.316	278.257	67.316	278.257	67.316

Note: *** indicate 1% significance levels. The data in brackets are t values.

, the regression coefficients for the “Three Modernization” integration in manufacturing on total carbon emissions and carbon emission intensity are negative and significant, indicating that the integration of China’s “Three Modernization” manufacturing has a significant suppressive effect on the dual control of carbon emissions, along with spillover effects.

Overall, under the three spatial weight matrices mentioned above, the indirect effects of the “Three Modernization” integration in manufacturing on the dual control of carbon emissions are greater than the direct effects, suggesting that the suppressive effects of integrating “Three Modernization” in neighboring regions on total carbon emissions and carbon emission intensity are stronger than those within the local area. Therefore, integrating “Three Modernization” in manufacturing has a strong spatial spillover effect on the control of total carbon emissions and carbon emission intensity.

4.6. Nonlinear Relationship Test

Table 9. Nonlinear relationship test.

(1) U-Shape/Inverted U-Shape Relationship Test			(2) Threshold Model Test
Variable	TCE	CEI	TCE	CEI
HIG	−2.441 *** (−4.54)	−2.449 *** (−4.91)	-	-
HIG2	1.327 *** (2.22)	1.543 *** (2.78)	-	-
TCX (<0.48)	-	-	−0.831 *** (−3.27)	-
TCX (≥0.48)	-	-	−0.428 *** (−2.48)	-
TCX (<0.48)	-	-	-	−0.756 *** (−4.08)
TCX (≥0.48)	-	-	-	−0.297 * (−1.36)
Control Variables	YES	YES	YES	YES
Province Fixed	YES	YES	YES	YES
Year Fixed	YES	YES	YES	YES
Constant Term	0.276 (0.66)	3.760 *** (9.71)	0.141 *** (0.54)	1.638 *** (30.93)
N	450	450	450	450
F-statistic	168.45 ***	95.58 ***	190.69 ***	71.19 ***
Adj R2	0.771	0.655	0.697	0.467

Note: ***, * indicate 1% and 10% significance levels, respectively. The data in brackets are t values.

provides a detailed test of the nonlinear relationship between the integration of “Three Modernization” in the manufacturing industry and the total carbon emissions and carbon emission intensity. The results show that there exists a significant “inverted U-shaped” relationship between the integration of “Three Modernization” and both total carbon emissions and carbon emission intensity. From the regression results in column 1, it can be seen that the regression coefficients for the integration of “Three Modernization” on total carbon emissions and carbon emission intensity are −2.441 and −2.449, respectively, while the squared terms for total carbon emissions and carbon emission intensity have regression coefficients of 1.327 and 1.543, further confirming the existence of the “U-shaped” nonlinear relationship.

In the initial stage of the “Three Modernization” integration in manufacturing, as the integration level gradually increases, both total carbon emissions and carbon emission intensity show a significant decreasing trend. During this stage, the “Three Modernization” integration in manufacturing brings about significant carbon emission reductions by improving resource utilization efficiency and technological levels. However, as the level of integration further increases, this suppressive effect gradually weakens. When the “Three Modernization” integration reaches a high level, total carbon emissions and carbon emission intensity begin to rise again, presenting a typical “U-shaped” curve. This phenomenon suggests that when the “Three Modernization” integration in manufacturing enters a higher level, although technological capabilities and automation levels are significantly improved, the accompanying increase in technological complexity leads to higher energy consumption, which may trigger a rebound effect in carbon emissions. Therefore, there is a significant “inverted U-shaped” relationship between the “Three Modernization” integration in manufacturing and carbon emission control, indicating a nonlinear mechanism of action. Specifically, at different stages of development, the “Three Modernization” integration in manufacturing will have distinctly different effects on carbon emission control. In the early stages of the integration, it helps to promote carbon emission control, but at higher levels, the effect weakens, showing a complex changing trend.

Column 2 of Table 9 presents the test results of the threshold model, indicating that technological complexity (TCX) exhibits a significant nonlinear characteristic in its effect on carbon emissions, i.e., there is a threshold effect. Specifically, when the level of technological complexity is below 0.48, its suppressive effect on total carbon emissions and carbon emission intensity is more significant, with regression coefficients of −0.831 and −0.756, respectively. However, when technological complexity is above or equal to 0.48, this negative impact significantly weakens, with corresponding coefficients decreasing to −0.428 and −0.297. This result suggests that in the early stages with lower technological innovation capacity, increasing technological complexity can effectively drive production method optimization and energy efficiency improvements, leading to more noticeable carbon emission reductions. This is likely due to the fact that in the low-level stages of manufacturing, there are generally weak technological foundations, low energy efficiency, and high pollution intensity. Therefore, the introduction and increase in technological complexity act as a “corrective” push for emissions reduction, offering significant reduction space and high marginal benefits. However, once technological complexity surpasses a certain threshold, its effect on carbon emissions weakens due to diminishing returns. On the one hand, as technological levels improve, enterprises have often already completed initial energy-saving and emissions-reduction transformations, with new technological investments focusing more on optimizing production efficiency and developing high-end products, thus weakening the direct improvement effect on carbon emission intensity. On the other hand, some high-tech activities themselves may involve higher energy and resource consumption, such as precision manufacturing and materials research. Without green guidance, this could potentially lead to a rebound effect on carbon emissions. Additionally, the increase in technological complexity often requires substantial capital investment and long development cycles, so its emissions reduction effects may not be fully realized in the short term, possibly weakening the overall suppressive effect.

5. Conclusions and Policy Recommendations

5.1. Conclusions

Based on the integration of “Three Modernization” in manufacturing and the estimation of total carbon emissions and carbon emission intensity, this study explores the impact mechanism of the “Three Modernization” development on regional carbon emission control. The following conclusions are drawn:

(1)

The integration of “Three Modernization” in manufacturing has a significant negative impact on carbon emissions. Both total carbon emissions and carbon emission intensity show a marked inhibitory effect.

(2)

The effects of carbon emission control vary significantly due to regional differences, economic development levels, resource dependence, and industrial structure. Specifically, integrating “Three Modernization” in manufacturing in the eastern region, economically developed areas, resource-based cities, and high-end manufacturing centers has achieved more significant results in controlling total carbon emissions and carbon emission intensity, while the reduction effects are relatively weaker in the central, western, and northeastern regions, less developed areas, non-resource cities, and traditional manufacturing bases.

(3)

The improvement of technological innovation significantly enhances the low-carbon technology application capacity of manufacturing, thereby indirectly reducing carbon emissions. The increased proportion of clean energy usage significantly promotes the green development of manufacturing, reducing its dependence on fossil fuels and playing an important role in dual control of carbon emissions.

(4)

Spatial analysis results show that integrating “Three Modernization” in manufacturing has a greater impact on the dual control of carbon emissions in neighboring regions than in the local region, indicating the presence of spatial spillover effects.

(5)

There is a nonlinear “inverted U-shape” relationship between the integration of “Three Modernization” in manufacturing and carbon emission control. In the early stages of integration, it helps promote dual control of carbon emissions. However, once the integration reaches a certain level, the effect of carbon emission control may show diminishing returns or even rebound. These findings are of great significance for future policy formulation and academic research. Policymakers should develop differentiated strategies for integrating “Three Modernization” based on regional differences, promote technological innovation and clean energy applications, strengthen regional cooperation, and enhance the overall effect of carbon emission control. Academic research should further explore the nonlinear impact mechanism of the “Three Modernization” integration to provide theoretical support for achieving carbon peak and carbon neutrality goals.

5.2. Policy Recommendations

Based on these conclusions, the following policy recommendations are proposed:

(1)

In the eastern regions, establish “Green Manufacturing Demonstration Zones” and pilot “Zero Carbon Factories”. Local governments, in collaboration with industry associations, should provide special subsidies and tax incentives for low-carbon technology transformation, improve regional carbon trading platforms, strengthen cross-regional green supply chain construction, and promote low-carbon information sharing. For the central, western, and northeastern regions, the government should set up “Central and Western Manufacturing Upgrading Special Funds” to increase infrastructure investment, such as smart grids and clean energy projects, and encourage technological cooperation and talent training between these regions and advanced enterprises in the eastern regions to help local businesses transition toward high value-added, low-energy consumption production.

(2)

In economically developed areas, strict carbon emission regulatory measures should be enforced. Focus on supporting enterprises’ independent research and development of low-carbon technologies, providing green loans, low-interest loans, and R&D subsidies, and establishing a comprehensive carbon emission monitoring system. In economically underdeveloped regions, more policy guidance and financial support are needed. Low-interest loans, tax reductions, and special subsidies should be used to encourage enterprises to introduce advanced low-carbon technologies from abroad while strengthening industry–university research cooperation and accelerating the localization of green technologies to ensure that low-carbon transformation of enterprises is synchronized with regional economic development.

(3)

For resource-based cities, it is recommended to develop a “Green Transformation Plan for Resource-Based Cities” to promote energy-saving renovations and green upgrades in traditional high-energy-consuming industries, and increase investments in recycling facilities. For non-resource cities, intelligent manufacturing and digital management systems should be vigorously promoted. Establish low-carbon intelligent manufacturing demonstration zones and foster green technology sharing through regional alliances. Additionally, for comprehensive manufacturing powerhouses and high-end manufacturing centers, policies should focus on supporting enterprises’ independent innovation and equipment upgrades. For traditional manufacturing bases and special industrial provinces, specialized energy-saving renovation plans and cross-industry integration development pilot projects should be implemented to improve the low-carbon level of the overall industrial chain.

(4)

A cross-regional low-carbon collaborative innovation platform should be established to form complementary mechanisms for technology, capital, and information among the eastern, central, western, and northeastern regions. A unified national dynamic monitoring and evaluation system for manufacturing industry carbon emissions should be constructed. Through regular data releases and third-party assessments, policies can be adjusted in a timely manner. Furthermore, public participation and information transparency should be strengthened, promoting green finance and environmental certification to ensure that regional cooperation and policy enforcement are coordinated, thus enhancing the overall effect of cross-regional carbon emission control.

(5)

Given the inverted U-shape relationship between the integration of “Three Modernization” and carbon emission control, local governments should determine the optimal integration level and establish a dynamic evaluation mechanism based on quantitative indicators to regularly monitor carbon emission effects. At the same time, phased policy adjustment plans should be introduced. In the early stages, low-carbon technologies and green transformations should be vigorously promoted. Once the optimal level is reached, related subsidies should be gradually reduced to prevent carbon emission rebound due to excessive investment, ensuring the long-term and stable achievement of the carbon emission dual control targets.