Can Digitalization Improve the High-quality Development of Manufacturing? An Analysis Based on Chinese Provincial Panel Data

Abstract

Digitalization plays a pivotal role in enhancing manufacturing quality development and upgradation, helping translate transformation goals. The available literature has explored the impact of high-quality development in manufacturing, but the impact of digitalization and the underlying mechanism still needs to be systematically studied. This study aims to verify the impact of digitalization on the high-quality development of manufacturing employing dynamic panel data and a quantile regression model on China’s provincial panel data between 2011 and 2020. The results indicated that high-quality manufacturing development showed a fluctuating rising trend with insignificant regional differences, with no apparent trend with a significant difference in the eastern region compared to the central and western regions. The nonlinear effect analysis results revealed that digitalization significantly promoted high-quality development in manufacturing with a more prominent effect at 50% quantile. Digitalization showed significant and heterogenous results in eastern region data compared to other regions, cementing the importance of digitalization in high-quality manufacturing development. This research also empirically enriches the discussion on the relationship between digitalization and the high-quality manufacturing development of industries providing meaningful insight into policy implications for improving the high-quality level of the manufacturing industry and achieving sustainable economic development.

Introduction

The high-quality development of manufacturing is the core driving force for economic transformation and development and is regarded as an integral pillar in the realization of economic modernization. In the new wave of scientific and technological revolution, the “reindustrialization strategy” of developed countries and the “kinetic energy transformation strategy” of developing countries depict that the importance of manufacturing has been re-understood globally. In recent years, China’s manufacturing industry has developed and grown by leaps and bounds (see Appendix Fig. 6). However, with the changes in the domestic and foreign economic development situation, the problem of deep-seated development of manufacturing has gradually become prominent. The high input, output, and pollution development model led to low resource utilization efficiency, severe environmental pollution, and even serious duplication of investment (Wang and Li, 2021; Udemba et al., 2020). In the current scenario, it could be deduced that though China’s manufacturing is enormous yet not strong and comprehensive (Wang et al., 2022). The sustainable development principle is linked to reduced pollution emissions (Mele and Magazzino, 2020), which led to redefining the priorities in China by putting forward new requirements and making new arrangements to achieve the goal of high-quality manufacturing development under the new situation. “Made in China 2025,” proposed in 2015, planned an action route for the future development of China’s manufacturing, aiming to help make China a manufacturing power. The 2018 Central Economic Work Conference listed the promotion of high-quality development of manufacturing as the top priority, and the 2019 government work report once again emphasized promoting the high-quality development of manufacturing (Yang and Zhang, 2020).

With the in-depth advancement of a new round of technological revolution and industrial transformation, the deep integration of a new generation of information technology represented by 5G networks, big data, and artificial intelligence with various industries is becoming a new engine for industrial optimization and upgrading. As a new type of economic and social form, digital development is steadily improving in China. According to estimates, from 2016 to 2020, China’s digitalization index increased from 100 to 601, making it the second-largest digital economy in the world. The development scale of the digital economy and its proportion in GDP has increased with each passing year (see Appendix Fig. 7). Digital development can promote the optimization and reorganization of the original production factors, improve the service structure and operation management process of products, and play a pivotal role in the high-quality development of manufacturing. The Chinese government attaches great importance to the role of digitalization in transforming and upgrading the manufacturing industry. The 2019 Central Economic Work Conference focused on the transformation and upgrading of manufacturing. It proposed vigorously developing the digital economy to achieve the goal of high-quality manufacturing development. However, the literature on the relationship between digitization and high-quality manufacturing development is not systematic. Based on China’s provincial panel data from 2011 to 2020, this study employed a dynamic panel data and quantile regression model and discusses whether digitalization promotes the high-quality development of manufacturing, with an additional investigation of the inherent mechanism of influence. Compared with the available research, the marginal contribution of this paper is mainly reflected in the following three aspects, i.e., it discusses the high-quality development of manufacturing from the perspective of digitization by using dynamic panel data and a quantile regression model to empirically analyze the causal relationship between digitization and high-quality development of manufacturing industry, providing a new idea for enhancing the core competitiveness of manufacturing, followed by an in-depth analysis of the influence mechanism of digitalization on the high-quality development of manufacturing. Based on the empirical results, it also provides valuable policy suggestions for decision-making sectors to achieve the goal of high-quality development of the manufacturing industry. Although this study tried to address many issues, some limitations should be considered in the future, i.e., due to availability, only provincial data was in this study. In the future, research should focus on micro-level enterprises. Furthermore, only linear and nonlinear regression models were used to explore the relationship between digitalization and the high-quality development of manufacturing, thus necessitating research on the spatial effect relationship between them in the future. Since the measurement of high-quality development of the manufacturing industry in this research varied, widely, more scientific and practical measurement methods are planned for future work.

This study’s results showed that the high-quality development of the manufacturing industry is in a fluctuating yet negligible upward trend, with significant regional differences. The overall upward trend of digitalization and regional differences were not apparent, yet digitalization had a significant role in promoting the high-quality development of the manufacturing industry. Further analysis of nonlinear effects showed that digitalization’s impact on the manufacturing industry’s high-quality development was at the 50% quantile. At this stage, digitalization’s role in promoting the manufacturing industry’s high-quality development is mainly achieved by promoting economic growth.

The remaining paper is structurally divided into sections. The relevant literature is reviewed in “Literature Review and Research Hypothesis,” while the research design is presented in “Research Design.” The empirical results and discussion followed by the research conclusion summarization are given in “Results,” while “Conclusions and Policy Implications” presents policy implications. This paper also further discusses the causal relationship between digitalization and the high-quality development of the manufacturing industry from the theoretical analysis aspect based on the measurement of the level of digitalization and the high-quality development of the manufacturing industry, in addition to the transmission path of digitalization to the high-quality development of the manufacturing industry from the two aspects of economic growth effect and technology progress effect, and finally draws research conclusions and policy implications. The specific framework is shown in Fig. 1.

Fig. 1.

Research framework

Literature Review and Research Hypothesis

Literature Review

As the leading economic body of a country or region, manufacturing is the key and focus of economic development. Comprehensively promoting the quality reform of manufacturing, making the transition from “Made in China” to “Created in China” is currently a key topic of discussion in academia. The high-quality development of manufacturing is an important strategic task for China’s economic development, which has been reported to improve the country’s core competitiveness significantly (Han et al., 2022; Wang and Li, 2021). The existing research on the high-quality development of manufacturing mainly focuses on measuring the high-quality development of manufacturing and identifying its influencing factors (Ma et al., 2018; Wang et al., 2022; Wang and Shi, 2022). The measurement methods for the high-quality development of manufacturing are mainly divided into single index and comprehensive index system methods. The former is mainly measured from the value-added rate, total factor productivity, and green total factor productivity. Wang and Shi (2022) comprehensively measured the high-quality development of manufacturing from the aspects of green development efficiency and technical structure; its essence still belongs to the category of single indicator measurement. The high-quality development of manufacturing is multi-dimensional, and a single index only reflects one aspect, so the evaluation method for constructing an index system is more reasonable (Jin, 2018). When constructing a multi-dimensional indicator system, existing research is mainly based on the perspective of manufacturing production input and output (Liang and Luo, 2019; Zhang et al., 2018; Schleich et al., 2017; Leoncini et al., 2019). Wang et al. (2022) measured the high-quality development of manufacturing in 30 provinces in China by constructing an index system employing the entropy method. The results showed that the fluctuation of high-quality manufacturing is rising, with significant regional differences. Li (2019) measured the high-quality development of manufacturing by constructing a multi-dimensional evaluation index system based on industry-level data from China’s manufacturing industry. Similarly, Han (2022) designed an evaluation index system for the high-quality development of manufacturing from various aspects.

Different scholars have also researched the factors affecting the high-quality development of manufacturing. For instance, Wang and Li (2021) analyzed the impact of capacity utilization efficiency on the high-quality development of manufacturing. They reported that capacity utilization efficiency had a significant positive impact on the high-quality development of manufacturing, with an inverse effect regarding environmental regulation. Wang et al. (2022), based on the quasi-natural experiment of China’s carbon emission trading pilot, showed that although China’s carbon emission trading pilot policy has improved the level of high-quality development of manufacturing, the policy effect had a certain lag. Li (2019) found that the foundation of industry development is a prerequisite for the high-quality development of manufacturing, while economic development and the natural environment are conducive to the high-quality development of manufacturing, which was also supported by the study conducted by Wang and Shi (2022). Tian et al. (2021) examined the high-quality development of China’s manufacturing from the perspective of import competition, and the results indicated that import competition generally inhibited the high-quality development of China’s manufacturing, while the impact of import competition in developed economies and import competition in developing countries was completely different. Du and Hong (2021) discussed the policy implications of high-quality development of China’s manufacturing under the new development pattern of “Dual Circulation.”

China is currently undergoing a “digitalization” process of transforming from a traditional economy to a digital one, which is envisaged to have a comprehensive and significant impact on the entire economic system. According to data from the China Academy of Information and Communications Technology, by 2020, China’s digital economy will reach 39.2 trillion Yuan, accounting for about 39% of GDP. Therefore, the development of digitalization has attracted widespread attention in the academic community. Existing studies mainly focus on measuring the digital level and its impact (Guo et al., 2022; Liu et al., 2022; Guo and Xu, 2021; Busulwa et al., 2022; Orłowski et al., 2022; Magazzino, Mele, et al., 2021). Li et al. (2022) measured the development level of the digital economy from the two dimensions of internet development and digital financial inclusion adopting the principal component analysis method to obtain the comprehensive development indicators of the digital economy. Fan and Wu (2021) constructed a digital indicator system from production digitalization, consumption digitalization, and circulation digitalization aspects. The weights of indicators are mainly determined by the principal component analysis and expert scoring method, and finally, the measurement of the digitalization level is completed. Guo and Xu (2021) empirically studied the impact of digital transformation on business operations and financial performance. The findings showed that the intensity of digital transformation boosts business performance, with a “U”-shaped relationship with profit-oriented financial performance. Liu et al. (2022) explained the theoretical mechanism of urban digital transformation on the recovery of the COVID-19 epidemic based on the background of the global response to the COVID-19 epidemic and tested its effectiveness through empirical analysis. The study conducted by Guo et al. (2022) started with the concept of enterprise digital transformation and theoretically expounded the role of digital transformation in breaking industry and regional monopoly.

Through the review and summary of the above literature, it is found that there is negligible literature on digital for the high-quality development of manufacturing. Although Chao et al. (2012) study was based on the entire manufacturing chain, they explain the theoretical mechanism of the impact of new digital infrastructure on the high-quality development of manufacturing from a theoretical aspect. However, digital development does not only refer to digital infrastructure; its research needs to analyze digitalization’s impact on manufacturing's high-quality development comprehensively. Furthermore, Liu and Hui (2021) discussed the digital economy's linear and nonlinear effect and constraint mechanism on the high-quality development of China’s manufacturing. But in-depth research has yet to be done on its internal theoretical mechanism. Therefore, with the rapid development of digitalization, it is imperative to study it in-depth for the high-quality development of manufacturing. Based on this, this paper analyzes the impact of digitalization on the high-quality development of manufacturing from the theoretical and empirical levels to provide decision-making references for government decision-making sectors.

Research Hypothesis

Digital development spans time and space, data creation, and interconnection (Han et al., 2020) and has a significant role in promoting the quality, efficiency, and power changes of manufacturing. The effect of digitalization on the quality transformation of manufacturing is mainly reflected in the improvement of product and service quality. On the one hand, digitalization expands the adequate supply of manufacturing. It improves product quality by making full use of digital technology. It integrates the final product design into the consumer’s ideas through quantitative analysis of user needs and consumption patterns. Secondly, the traditional manufacturing mode has been changed through the realization of customized manufacturing through the digital production workshop enabling the realization of the goal of expanding adequate supply. On the other hand, digitization has reconstructed the development pattern of manufacturing, improved service quality, and provided the necessary industrial support for servitization manufacturing. Based on advanced digital technology and the modern service industry after digital transformation, the online services of manufacturing enterprises are more efficient, and the added value of manufacturing products has increased. Digitization breaks the limitation of time and space, realizes the effective allocation of production factors such as human resources, technology, and materials under the adjustment of the market, and ensures the full utilization of resource factors (Ma and Ni, 2020). It can be embedded in the manufacturing process to promote agile manufacturing and lean production of enterprises, thereby significantly improving the production efficiency of manufacturing. The driving force change of digitalization to the high quality of manufacturing is mainly reflected in the changes in key production factors and infrastructure, which in turn accelerates the formation of new industries and the upgrading of traditional industries. Data has become a new production factor that drives the high-quality development of manufacturing; by using the industry’s massive data collected by smart manufacturing, the loss of market operations is reduced. Additionally, the required industrial resources are integrated and processed through the data information network, which realizes the real-time transmission of information. Digitization accelerates user data flow, overcomes the scarcity and homogeneity of internal resources, and spawns new industries and formats. Therefore, the first hypothesis H1 of this paper, is proposed.

• H1. Digitalization has a significant role in promoting the high-quality development of manufacturing.

Digitalization combines the Internet and cloud computing, promotes the deep integration of information technology and economic development, and fosters new momentum for economic growth. In digitization, data, as a new production factor, increase entrepreneurial opportunities and promotes economic growth (Magazzino, Porrini, et al., 2021). The promotion of digitalization to economic growth is mainly reflected in the promotion of industrial digitalization, the expansion of the digital economy, and the construction of digital infrastructure. Specifically, digitization mainly affects economic growth by improving production efficiency and reducing production costs. From the perspective of improving production efficiency, the productivity of enterprises that adopt intelligent digital equipment is significantly improved compared to enterprises lacking it, thereby obtaining higher benefits. Secondly, from the perspective of reducing production costs, improving production efficiency is conducive to allocating fixed costs and thus accelerates the marginal cost of production. The rapid development of digital logistics saves transportation time and costs and helps promote economic growth. However, the level of economic development, as the core basis for the high-quality development of manufacturing, is crucial in realizing the high-quality development of manufacturing. Therefore, this paper proposes hypothesis H2.

• H2. The effect of economic growth is an important way for digitalization to affect the high-quality development of manufacturing.

Digitalization will bring disruptive innovations to the Chinese economy, bringing specific changes and innovations in production, management, and organization (Guo and Luo, 2016). However, existing studies have yet to reach a consistent conclusion on the impact of digitalization on technological progress. On the one hand, it is believed that although digitalization has improved scale efficiency, it has not significantly impacted technological progress. On the other hand, research also concluded that digitalization has a significant role in promoting technological progress. This paper believes that digitalization has broken through a certain time and space and enhanced the ability to summarize and organize scattered information. According to “Metcalfe’s Law,” society’s technological progress is promoted through information sharing and dissemination. Therefore, this paper proposes a third hypothesis H3.

• H3. The effect of technological progress is another important way in which digitalization affects the high-quality development of manufacturing.

Research Design

Measurement of High-quality Development of Manufacturing

Based on the availability and quantification of data, this paper constructs an evaluation index system for high-quality development of manufacturing based on reference to existing research (Wang et al., 2022), as shown in Table 1. In this paper, the entropy TOPSIS method is used to measure the high-quality development of manufacturing. The specific measurement process refers to existing research (Wang et al., 2021; Yu, 2021). The calculation steps are as follows: (1) construct judgment matrix P = (a_ij)_m × n, and carry out standardized processing; (2) calculate the proportion x_ij of the index a_ij, x_ij = a_ij/ ∑ a_ij; (3) calculate the entropy value e_j, e_j =  − k ∑ x_ij ln(x_ij), k = 1/ ln (m); (4) calculate the difference coefficient g_j, g_j = 1 − e_j; (5) calculate the weight w_jof the index a_ij, w_j = g_j/ ∑ g_j; (6) calculate the high-quality development of manufacturing HM_i of the ith city, ${\text{HM}}_i={\sum }_{j=1}^m{w}_{\text{ij}}{a}_{\text{ij}}$.

Table 1.

Evaluation index system for high-quality development of manufacturing

Target	Subsystem	Indicator measurement method
High-quality development of manufacturing	Economic benefits	Gross industrial output value/average number of employees (+)
		Total profit/main business income (+)
		Industrial value-added gross regional product (+)
	Innovation driven	R&D internal expenditure/main business income (+)
		Full-time equivalent of R&D personnel (+)
		Number of valid invention patents (+)
	Green development	New product sales revenue/main business income (+)
		Industrial solid waste production/industrial value-added (−)
		Total energy consumption in manufacturing/industrial added value (−)
	Structural optimization	Comprehensive utilization of industrial solid waste/production of industrial solid waste (+)
		Main business income of high-tech industry/main business income (+)
		High-tech industry new product sales revenue/new product sales revenue (+)

“+” and “−” indicate the attributes of the indicator. “+” means indicator is positive, while “−” means indicator is negative. “/” means division symbol

Measurement of Digitalization

There are few quantitative studies on digitalization, and no unified measurement standard has been formed. Some existing studies use the internet penetration rate and the number of internet broadband access users per capita to measure the level of digitalization development (Habibi and Zabardast, 2020). However, digitization is a complex system that drives the evolution of economic and social structures, and a single indicator cannot accurately and comprehensively reflect the level of digitalization. Because of this, some literature used the measurement indicators of the digital economy to characterize digitalization (Jovanovic et al., 2018), while some studies constructed an indicator system for measurement evaluation from different perspectives (Zhou et al., 2020). Therefore, considering the availability and quantification of data, this paper comprehensively measures the digitalization index from the level of digital infrastructure, digital application, and digital industry development. The specific measurement indicators are shown in Table 2. Similarly, the entropy TOPSIS method is also used to calculate digitalization.

Table 2.

Evaluation index system for digitalization

Target	Subsystem	Indicator measurement method
Digitalization	Digital infrastructure	Internet broadband access port (+)
		Office switch capacity (+)
		Mobile telephone exchange capacity (+)
		Mobile phone base station (+)
		Optical cable line length (+)
		Number of sites (+)
	Digital application	Mobile SMS traffic (+)
		Total postal service (+)
		Total telecom business (+)
		Courier business income (+)
		E-commerce sales (+)
		Mobile phone penetration (+)
	Digital industry development	Fixed investment in information transmission, software, and information technology services/total investment in the whole society (+)
		Employment of information transmission, software, and information technology services/urban employment (+)

“+” indicates the attributes of the indicator. “+” means indicator is positive. “/” means division symbol

Empirical Model

In order to test the impact of digitalization on the high-quality development of manufacturing, this paper incorporates digitalization into the research framework of high-quality development of manufacturing for analysis. Drawing on the Zhang et al. (2021) method, the following panel data model is constructed.

$${\text{HM}}_{\text{it}}={\alpha }_0+{\alpha }_1{\text{DG}}_{\text{it}}+\sum {\beta }_i{\text{control}}_{\text{it}}+{\mu }_i+{\nu }_t+{\epsilon }_{\text{it}}$$ 1

In Eq. 1, i represents each province (1, 2, …, 30), t represents time (2011, 2012, …, 2020). The explained variable HM represents the high-quality development of manufacturing, and the core explanatory variable DG. Here, control represents the control variable, α and β are the coefficient matrix of the core explanatory variable and the control variable, respectively, μ is the fixed effect coefficient, and ε_it is the random error term.

Although the fixed effects in the static panel data model can alleviate the problem of variable endogeneity, it is difficult to reflect the dynamic changes in the high-quality development of manufacturing. Using the dynamic panel data model cannot only better deal with the endogenous problem, but also observe the dynamic effect of high-quality development of manufacturing. Based on this, this paper uses the generalized moment estimation method (GMM) for model analysis. GMM estimation methods mainly include differential generalized moment estimation (DIF-GMM) and system generalized moment estimation (SYS-GMM). The system GMM method overcomes the weak instrumental variable problem of the differential GMM method, solves the endogenous problem, and improves estimation efficiency (Bond, 2002). The following dynamic panel regression model is constructed.

$${\text{HM}}_{\text{it}}={\alpha }_0+{\alpha }_1{\text{HM}}_{it-1}+{\alpha }_2{\text{DG}}_{\text{it}}+\sum {\beta }_i{\text{control}}_{\text{it}}+{\mu }_i+{\nu }_t+{\epsilon }_{\text{it}}$$ 2

where, HM_{i t − 1} represents the high-quality development of manufacturing with a lag of one period, and the rest of the variables are the same as Eq. 1.

It should be pointed out that the above measurement model mainly examines the impact of digitalization on the expectation of high-quality development conditions in the manufacturing, which is essentially a mean regression and is easily affected by extreme values, in order to accurately describe the complete statistical characteristics of the conditional distribution and effectively capture the impact of digitalization in the extreme areas of high-quality development of manufacturing. Therefore, based on the practice of Chao et al. (2012), the following quantile regression model is constructed.

$${\text{HM}}_{\text{it}}\left(\tau \right)={\alpha }_0\left(\tau \right)+{\alpha }_1\left(\tau \right){\text{DG}}_{\text{it}}+\sum {\beta }_i\left(\tau \right){\text{control}}_{\text{it}}+{\mu }_i+{\nu }_t+{\epsilon }_{\text{it}}$$ 3

In Eq. 3, τ (0<τ<1) represents the different points of the conditional distribution, which are 0.1, 0.25, 0.5, 0, 75, and 0.9, respectively. The core coefficient α₁(τ) reveals the marginal impact of digitalization on the high-quality development of manufacturing at different quantiles.

In order to further verify H2 and H3, this paper constructs the following mediation effect model based on Wen et al. (2004).

$${\text{EG}}_{\text{it}}\left({\text{SI}}_{\text{it}}\right)={\alpha }_0+{\alpha }_1{\text{DG}}_{\text{it}}+\sum {\beta }_i{\text{control}}_{\text{it}}+{\mu }_i+{\epsilon }_{\text{it}}$$ 4

$${\text{HM}}_{\text{it}}={\alpha }_0+{\alpha }_1{\text{HM}}_{it-1}+{\alpha }_2{\text{DG}}_{\text{it}}+a_3{\text{EG}}_{\text{it}}\left({\text{SI}}_{\text{it}}\right)+\sum {\beta }_i{\text{control}}_{\text{it}}+{\mu }_i+{\nu }_t+{\epsilon }_{\text{it}}$$ 5

where, EG is economic growth; SI is technological progress; other variables and regression coefficients are consistent with Eq. 2.

Data and Variables

Data Source

Based on the availability and consistency of data, this paper selects the provincial panel data from 2011 to 2020 (excluding Hong Kong, Macao, Taiwan, and Tibet) as the basic data for this study. The data required for the high-quality development of the manufacturing industry in this paper are derived from China Industrial Statistical Yearbook, China Statistical Yearbook, China Science and Technology Statistical Yearbook, and China Energy Statistical Yearbook. The original data required for digitization are derived from the China Statistical Yearbook and the National Bureau of Statistics, and some of the missing data are supplemented by the corresponding yearbook of the corresponding region. The control variable data are mainly obtained from the China Statistical Yearbook and the National Bureau of Statistics.

Variable and Definition

The explained variable, the high-quality development of manufacturing (HM), and the core explanatory variable digitalization (DG) in this paper are calculated from the above and will not be repeated here. Economic growth (EG) and technological progress (SI) are mediation variables. Based on existing research, economic growth is expressed by per capita GDP (Zhao et al., 2021; Sun and Huang, 2020). Existing research generally measured technological progress from the perspective of input and output (Kontolaimou et al., 2016; Wurlod and Noailly, 2018). However, improving efficiency through technological progress is the fundamental goal of technological progress, so this paper uses the number of effective invention patents as a substitute variable for technological progress.

According to the existing studies, some provincial characteristics were added as control variables in the regression analysis of the model to alleviate the bias of omitted variables as much as possible (Wang et al., 2022). Therefore, this paper incorporates industrial structure upgrading (SU), urbanization level (UR), marketization level (MR), government support (GS), and foreign direct investment (FDI) as control variables into the model. Referring to the method of Yu and Wang (2021), an industrial structure upgrading index is constructed. Existing classical literature about urbanization (UR) measurement methods generally includes the population ratio, urban land use index, rural urbanization index, and adjustment coefficient method. This paper adopted the indicator method of the proportion of the urban population in the total population to measure the level of urbanization (Wei, 2019). For the measurement of the marketization level, we adopted the method proposed by Fan et al. (2011) and used the marketization index as a surrogate variable for the marketization level. Government support was measured by the proportion of fiscal general public budget expenditure in GDP, and FDI was measured by the ratio of regional FDI to local GDP (Hu et al., 2020).

Data Descriptive Statistics

Table 3 summarizes the data characteristics of each variable. At the same time, in order to solve the problem of heteroscedasticity, data processing mostly adopts the method of natural logarithm and ratio. It can be seen from Table 3 that the average HM is 1.362, the maximum value is 2.075, and the minimum value is only 1.132, indicating that there are large differences in the high-quality development of manufacturing between regions. Likewise, digitalization is at a low level and there are large differences between regions. All other variables were within the expected range, and there were no significant outliers.

Table 3.

Descriptive statistics of the main variables

Variable	Mean	Max	Min	Std. dev.	P25	P50	P75	P95	Obs.
HM	1.362	2.075	1.132	0.163	1.245	1.321	1.438	1.693	300
DG	1.264	2.100	1.031	0.189	1.132	1.201	1.354	1.590	300
EG	5.617	16.492	1.643	2.698	3.766	4.884	6.682	11.206	300
SI	8.982	12.984	4.534	1.568	8.031	8.965	10.082	11.316	300
SU	2.374	2.835	2.166	0.129	2.286	2.361	2.427	2.680	300
UR	0.590	0.896	0.350	0.122	0.505	0.572	0.649	0.867	300
MR	6.795	11.400	2.330	1.999	5.200	6.695	8.095	10.230	300
GS	0.251	0.643	0.110	0.103	0.184	0.227	0.291	0.436	300
FDI	0.378	1.839	0.048	0.367	0.141	0.206	0.493	1.284	300

Obs. stands for the observations of the variables, Mean refers to the average value of the variables, Std. dev. represents standard deviation, Min and Max indicate the minimum and maximum values of the variables, respectively

Results

Calculation Results

This paper estimated the high-quality development of China’s manufacturing based on the entropy TOPSIS method. As can be seen from Fig. 2, although the high-quality development of manufacturing was in a fluctuating upward trend, the increase was negligible, depicting that the high-quality development of manufacturing still faces many obstacles. For example, technological innovation needs to meet the requirements for high-quality manufacturing development, and core technologies lag behind developed countries. In some areas, there is an excessive development of the tertiary industry with the service industry as the primary industry, showing a situation of “deviation from the real to the virtual” and ignoring the high-quality development of manufacturing. There were significant regional differences in the high-quality development of manufacturing, i.e., the high-quality development level of manufacturing in the eastern region was higher than that in the central and western regions, showing a development trend of “strength in the east and weakness in the west,” which was consistent with the report of Wang et al. (2022). The major reasons observed were that the eastern region had a sound industrial base, economic strength, and a strong capacity for technological innovation. It had long been a pacesetter for China’s openness (Lin and Zhou, 2022), putting the development quality of the manufacturing industry at a high level. The economic, ecological, and environmental constraints in the central and western regions have caused the high-quality development of manufacturing to lag behind the eastern regions.

Fig. 2.

Average level of high-quality development of manufacturing

Likewise, we measured digitalization according to the entropy TOPSIS method. As shown in Fig. 3, the overall improvement of digitalization in 2011–2020 was not noticeable, and the level of digitalization development in 2020 was only about 1.3% higher than that in 2011. From a regional perspective, the digitalization level in the eastern region was significantly higher than in the central and western regions. The main reason could be that digital construction requires a lot of human, material, and financial resources, and areas with weak economic strength need more resources to develop digitalization vigorously. Moreover, digitization requires a lot of digitization-related professionals and technologies. The central and western regions’ geographical location and natural environment have natural disadvantages and no advantages in talent introduction and investment promotion. The strong economic development in the eastern region has provided a solid foundation for digital infrastructure and applications. For example, the proportion of online shopping in Jiangsu, Zhejiang, and Shanghai in the eastern region ranked first in the country.

Fig. 3.

Average level of digitalization

Empirical Results

Variable Stationarity Test

In order to avoid regression and ensure the unbiasedness and validity of the results, this paper conducts a stationarity test on the variables. From the test results, it can be seen that each variable is a horizontal series stationary, which can be perform regression analysis. The test results are shown in Table 4.

Table 4.

Results of unit root test

Variables	LLC	HT	IPS
HM	−4.4283***	−24.6546***	−8.7413***
DG	−7.7978***	−31.4789***	−10.0145***
SU	−9.1082***	−31.9870***	−9.8951***
UR	−19.0295***	−37.8036***	−10.9775***
MR	−1.3362*	−27.3272***	−9.1540***
GS	−4.9488***	−31.4838***	−9.6727***
FDI	−16.7464***	−34.7669***	−10.4075***

***p<0.01, **p<0.05, *p<0.1

Benchmark Regression Results

Before estimating the model, this paper fits the linear relationship between the high-quality development of manufacturing and digitalization. Figure 4 shows that the linear fitting slope was more significant than 0, roughly depicting a linear relationship between high-quality manufacturing development and digitalization. Considering that there may be endogenous problems in the model, we used the system GMM method proposed by Arellano and Bover (1995) and Blundell and Bond (1998) to estimate the econometrics and parameters of the panel data model. System GMM is divided into one-step and two-step system GMM estimations. Compared to the one-step method, the two-step method is less susceptible to the interference of heteroscedasticity. Under limited sample conditions, the standard error of the two-step method may be biased. This paper estimates the model by selecting robust standard error and then corrects the two-step GMM estimation. However, some scholars have criticized the problem of weak instrumental variables in the system GMM method. So, it is vital to test the robustness of the estimation results. The stepwise regression method can eliminate the variables that cause multicollinearity. Accordingly, this paper uses the stepwise addition of control variables to regress the model. To obtain robust regression results, the OLS test results are reported. The regression results of the model are shown in Table 5.

Fig. 4.

Linear fitting

Table 5.

Benchmark regression results

Variables	OLS	OLS	OLS	OLS	SYS-GMM	SYS-GMM	SYS-GMM	SYS-GMM
L.HM					0.039 (0.069)	0.041 (0.034)	−0.020 (0.026)	−0.051 (0.039)
DG	0.482*** (0.042)	0.430*** (0.047)	0.286*** (0.056)	0.297*** (0.056)	0.558*** (0.021)	0.425*** (0.054)	0.275*** (0.048)	0.219*** (0.101)
SU		√	√	√		√	√	√
UR		√	√	√		√	√	√
MR			√	√			√	√
GS			√	√			√	√
FDI				√				√
_cons	0.752*** (0.053)	1.444*** (0.210)	1.513*** (0.211)	1.416*** (0.210)	0.601*** (0.118)	0.869*** (0.163)	0.999*** (0.140)	0.803*** (0.206)
AR (1)					−2.2812 (0.0225)	−2.423 (0.0154)	−2.5076 (0.0122)	−2.5738 (0.0101)
AR (2)					−0.5113 (0.6091)	−0.9478 (0.3432)	−0.2360 (0.8134)	0.1754 (0.8608)
Sargan test					9.8169 (1.0000)	9.4125 (1.0000)	9.7559 (1.0000)	9.0941 (1.0000)
R2	0.3077	0.3652	0.4107	0.4168
Obs.	300	300	300	300	300	300	300	300

∗p < 0.1, ∗∗p < 0.05, ∗∗∗p < 0.01

As shown in Table 5 AR (1) and AR (2), the regression results showed a first-order serial correlation between the disturbance items, with no second-order serial correlation. The corresponding Sargan test P values were high, confirming that the original hypothesis, i.e., “all instrumental variables are valid,” cannot be rejected, and the over-identification test is passed. It can be explained that the model regression results are reliable and effective. From the perspective of the regression coefficient, the regression coefficient of the lag period of the high-quality development of manufacturing was not significant, indicating superposition effect of the high-quality development of manufacturing was not noticeable. Regardless of whether control variables are added or not, the estimated coefficient of digitalization was significantly positive and passed the 1% significance level, indicating that digitization has a significant role in promoting the high-quality development of manufacturing, thereby verifying H1.

On the one hand, digitalization has promoted the expansion of adequate supply in manufacturing and improved product quality. On the other hand, digitalization has reconstructed the development pattern of manufacturing and improved service quality. Moreover, data has become a new production factor that drives the high-quality development of the manufacturing industry. Using industry big data collected by intelligent manufacturing realizes real-time information transmission, reduces market operation losses and is conducive to improving industrial collaborative innovation. Finally, as one of the digitalization expressions, the new digital infrastructure promotes the high-quality development of the manufacturing industry by promoting the transformation of functions. The development of new digital infrastructure has accelerated the circulation of data among various users, overcome the scarcity and homogeneity of internal resources, and spawned new industries and formats by continuously expanding the temporal and spatial boundaries covered by data-sharing platforms, which is in good agreement to the study of Chao et al. (2012). As a result, with the development of digitalization, the high-quality development of manufacturing has been promoted.

Robustness Test

To obtain more robust empirical results, we performed robustness tests from the following three aspects: Firstly, due to the uniqueness of the four municipalities, Beijing, Tianjin, Shanghai, and Chongqing were removed from the overall sample, and the remaining sample size was re-substituted into the model for regression testing. Secondly, we eliminated the effect of outliers, winsoring the highest and lowest 1% of all continuous variables, and re-estimated the benchmark model. Thirdly, we substitute digital infrastructure and digital application into empirical models to test the robustness of benchmark models. Finally, there could be random error terms that do not meet the model assumptions in the above model estimation. The obtained results may be biased if the parameter estimation is directly performed. Generalized least squares (FGLS) effectively resolve the problems of serial correlation and heteroscedasticity caused by cross-sectional data and obtain more effective estimation results (Sun and Liu, 2021). Nonetheless, due to the nature of the data in this paper, the standard deviation of the FGLS method may not effectively reflect its variation. In this case, panel correction standard error estimation (PCSE) should be used for correction to obtain a more accurate estimation result (Wang and Shao, 2022). The results of the above robustness test reveal that the signs of the core explanatory variable has not significantly changed, indicating that our research conclusions have desirable robustness (Table 6).

Table 6.

Estimation results of robustness test

Variables	(1)	(2)	(3)	(4)	(5)
L.HM	−0.167*** (0.054)	−0.203** (0.082)	−0.022 (0.038)	−0.041 (0.056)
DG	0.702** (0.334)	0.279*** (0.034)	0.510*** (0.097)	0.613*** (0.100)	0.297*** (0.075)
Control variables	√	√	√	√	√
_cons	0.602** (0.274)	1.463*** (0.303)	0.824*** (0.074)	1.273*** (0.346)	1.416*** (0.300)
AR (1)	−1.5689 (0.1167)	−2.5321 (0.0113)	−2.4683 (0.0136)	−2.3066 (0.0211)
AR (2)	0.4108 (0.6812)	−1.6842 (0.0921)	−0.1308 (0.8959)	−0.5796 (0.5622)
Sargan test	8.9014 (1.0000)	5.3024 (1.0000)	9.6920 (1.0000)	9.1607 (1.0000)
R2					0.4285
Obs.	300	300	3000	300	300

∗p < 0.1, ∗∗p < 0.05, ∗∗∗p < 0.01

Endogeneity Test

In order to further deal with the possible endogeneity problem and ensure the robustness and reliability of the empirical research conclusions in this paper, in Table 7, we use the instrumental variable method (IV) to regress the model. In this paper, the selection of instrumental variables is based on a relatively conventional practice, and the first-order lag term of endogenous variables is used as its own instrumental variable. The regression results show that the weak instrumental variable test, over-identification test and non-recognition test are passed, and the regression results of the impact of digitalization on the high-quality development of manufacturing industry are basically consistent with the conclusions of benchmark regression, indicating that the core conclusions of this paper are still relatively robust after considering endogeneity.

Table 7.

Estimation results of endogeneity test

Variables	IV - 2SLS
	The first stage	The second stage
DG		0.391*** (0.065)
L.DG	0.750*** (0.052)
Control variables	√	√
_cons	−0.209 (0.165)	1.546*** (0.245)
Kleibergen-Paap rk LM		31.033 (0.0000)
Shea’s partial R2		0.5842
Cragg-Donald Wald F		205.808 (0.0000)
Obs.	270	270

∗p < 0.1, ∗∗p < 0.05, ∗∗∗p < 0.01

Quantile Regression Results

The above basic econometric models mainly describe digitalization’s effect on manufacturing’s high-quality development in the mean range, ignoring its tail state characteristics in the extreme range. Digitalization’s impact on manufacturing’s high-quality development may be nonlinear. To accurately describe the asymmetric impact of digitalization on the high-quality development of manufacturing, we can effectively capture the tail characteristics of the distribution of digitalization and high-quality development of manufacturing. This paper next used quantile regression to estimate the quantile equations of the high-quality development of manufacturing affected by digitalization at the 0.1, 0.25, 0.5, 0.75, and 0.9 quantiles, respectively. From the regression results in Table 8, it can be seen that the fitting coefficients of digitalization are all significantly positive, indicating that digitalization has a significant positive impact on each quantile of high-quality development of manufacturing. However, with the increase of quantiles, the fitting value of the digitalization coefficient presents an inverted “U”-shaped structure, indicating that when the level of digitalization is low, as the level of digitalization development increases, its role in promoting high-quality development of manufacturing gradually increases. When the level of digitalization develops to a higher level, its role in promoting the high-quality development of manufacturing is weakened. The changing trend of the digitalization regression coefficient at different quantiles is shown in Fig. 5.

Table 8.

Quantile regression results

Variables	10%	25%	50%	75%	90%
DG	0.220*** (0.077)	0.217*** (0.070)	0.372*** (0.069)	0.337*** (0.053)	0.291*** (0.094)
Control variables	√	√	√	√	√
_cons	0.938*** (0.170)	0.858*** (0.185)	1.318*** (0.199)	1.541*** (0.284)	1.886*** (0.462)
R2	0.2573	0.3027	0.3399	0.3448	0.2944
Obs.	300	300	300	300	300

∗p < 0.1, ∗∗p < 0.05, ∗∗∗p < 0.01

Fig. 5.

Variation of quantile regression coefficients

Heterogeneity Analysis

Due to the vast territory of China, the economic base, industrial base, and resource endowment between regions are quite different. The provinces in the eastern region have taken the lead in entering the late stage of industrialization, while the provinces in the central and western regions are still in the middle stage of industrialization, and there is an obvious gradient gap (Huang, 2018). At the same time, the process of industrialization is often accompanied by the development of industrial transformation. Compared with the eastern region, the central and western regions are lagging behind in terms of technological changes such as automation, informatization, networking, and intelligence, as well as in resource allocation and institutional reform. Therefore, it is necessary to examine the heterogeneous impact of digitalization on the high-quality development of manufacturing in different regions. From the test results in Table 9, it can be seen that digitalization promotes the high-quality development of manufacturing in the eastern and central regions, but inhibits the high-quality development of manufacturing in the western region. The main reason is that the eastern economy is developed, which provides a good foundation for digital development, while the central and western regions have inherent shortcomings in digital development, so digitalization has not played a positive role in promoting the high-quality development of the manufacturing industry.

Table 9.

Heterogeneity test results

Variables	Eastern	Central	Western
L.HM	−0.132*** (0.046)	0.025 (0.069)	−0.162 (0.210)
DG	0.266*** (0.041)	0.217** (0.088)	−0.114 (0.321)
Control variables	√	√	√
_cons	0.528 (0.427)	0.874* (0.489)	2.731*** (1.027)
AR (1)	−1.9290 (0.0537)	−1.8553 (0.0636)	−1.1979 (0.2310)
AR (2)	1.1098 (0.2671)	0.3346 (0.7379)	−0.8965 (0.3699)
Sargan test	7.8402 (1.0000)	8.9027 (0.9978)	3.6920 (1.0000)
Obs.	120	90	90

∗p < 0.1, ∗∗p < 0.05, ∗∗∗p < 0.01

Mechanism Test

The first two columns in Table 10 are the test results of the economic growth effect channel. The fitting coefficient of digitalization to economic growth was significantly positive (as shown in the first column), indicating that the development of digitalization is conducive to improving economic growth. In the second column, the fitting coefficient of economic growth was significantly positive, indicating that economic growth can significantly promote the high-quality development of manufacturing. At the same time, the fitting coefficient of digitalization was not significant, indicating that economic growth plays a complete mediating role in the process of digitalization affecting the high-quality development of manufacturing, thereby verifying H2. The third and fourth columns are the technical progress effect channel test results. As shown in the third column, the fitting coefficient of digitalization to technological progress was negative, indicating that digitalization has not promoted technological progress. In the fourth column, the fitting coefficient of technological progress was significantly positive, indicating that technological progress can significantly promote the high-quality development of manufacturing. However, the fitting coefficient of digitalization was also significantly positive, with the absolute value more significant than that of the benchmark regression, indicating that the mediating role of technological progress in the digitalization process affecting the high-quality development of manufacturing has not appeared. The main reason might be that the realization of technological progress takes longer and faces more constraints. The effect of digitalization on technological progress can only be manifested after a long period of accumulation.

Table 10.

Mechanism test results

Variables	EG	HM	SI	HM
DG	1.166** (0.463)	0.352 (0.331)	−1.787*** (0.600)	0.270*** (0.085)
EG		0.100*** (0.012)
SI				0.011** (0.005)
Control variables	√	√	√	√
_cons	−5.724*** (0.787)	1.044*** (0.075)	8.807*** (1.021)	0.746*** (0.239)
Obs.	300	300	300	300

∗p < 0.1, ∗∗p < 0.05, ∗∗∗p < 0.01

Conclusions and Policy Implications

In the era of rapid development of digitalization, its importance to the high-quality development of China’s manufacturing is self-evident. Based on expounding the high-quality development of manufacturing driven by digitalization, this paper empirically examined the impact and mechanism of digitalization on the high-quality development of manufacturing. The main conclusions are as follows: (1) The high-quality development of manufacturing is in a fluctuating upward trend, but the increase is negligible, with significant regional differences, showing a development trend of “strength in the east and weakness in the west.” The overall upward trend of digitalization was not noticeable. From a regional perspective, the digitalization level in the eastern region was significantly higher than in the central and western regions. (2) Digitalization has a significant role in promoting the high-quality development of manufacturing. Through nonlinear effect analysis, it can be seen that digitalization’s role in manufacturing’s high-quality development was more evident at the 50% quantile. (3) There was significant regional heterogeneity in the promotion of digitalization to the high-quality development of manufacturing, and the promotion of the high-quality development of manufacturing in the eastern region was most significant. Still, it inhibited the high-quality development of manufacturing in the western region. (4) At this stage, the promotion of digitalization to the high-quality development of manufacturing is mainly achieved by promoting economic growth. However, the mediating role of technological progress has not been revealed.

Based on these findings, we put forward the following policy recommendations: (1) promote digital infrastructure construction and pragmatically develop the foundation for the industrial internet by further improving digital infrastructure such as satellite internet and industrial internet, and by building a digital trading platform and a shared ecological platform through the efficient use of digital elements and digital technology; (2) improve the deep integration of digital technology and manufacturing, and stimulate the digital network effect to improve the numerical control rate of equipment and the ability to apply information services in manufacturing enterprises to promote the production efficiency of the manufacturing industry; (3) implement a differentiated development strategy to promote the coordinated development of the manufacturing industry between regions. The eastern region should actively promote the development of digital industries and improve the supply capacity of network platforms and digital information service capabilities. The governments of the central and western regions should increase financial support, improve infrastructure construction such as optical fiber cables and broadband access ports, strengthen the introduction and cultivation of high-quality talents, and provide a suitable environment for the integrated development of digitalization and manufacturing. (4) Furthermore, there is improvement of economic development, strengthening of technological leadership, and promotion of the deep integration of technological and industrial chains. Economic growth, as the material basis for the high-quality development of manufacturing, must unswervingly develop the regional economy and escort the high-quality development of manufacturing. As the core driving force for the high-quality development of manufacturing, technological progress plays a vital role. First, it is necessary to systematically deploy major innovation carriers, speed up the construction of high-tech and innovation zones in various regions, and ensure complete digital infrastructure. Secondly, we must focus on improving major R&D platforms, actively cultivate national or provincial manufacturing innovation centers, and strengthen the construction of large-scale co-creation and sharing laboratories.

Although this study addresses many issues, some limitations should be considered in the future. Due to availability, we only used provincial data in this study. In the future, research should focus on micro-level enterprises. Secondly, we only used linear and nonlinear regression models to explore the relationship between digitalization and the high-quality development of manufacturing, and further research is needed on the future spatial effect relationship between them. In addition, the measurement of high-quality development of the manufacturing industry in the existing research varies widely. Therefore, we plan to use more scientific and effective measurement methods.

Appendix

See Figs. 6 and 7.

Fig. 6.

Scale and proportion of manufacturing added value. Source: China Statistical Yearbook

Fig. 7.

Scale and proportion of digital economy. Source: China Academy of Information and Communications Technology

Author contribution

Lianghu Wang: conceptualization, methodology, software, investigation, writing—original draft, writing—review and editing. Jun Shao: supervision, writing—review and editing. All authors read and approved the manuscript.

Funding

This work was supported by the Chongqing Municipal Education Commission Innovation Project (No. CYS18080); the National Social Science Foundation of China (No. 22&ZD095).

Declarations

Competing interest

The authors declare no competing interests