Digital economy, binary factor mismatch and sustainable economic development of coastal areas in China

Abstract

In order to promote the sustainable economic development, it is critical employ the digital economy to solve the mismatch dilemma of land and marine factors in coastal areas. It analyzed the influencing mechanisms between the digital economy, land and labor factor mismatch and coastal economic sustainable development using network development and new economic growth theories. The intermediary and regulating effect models were used for empirical tests using panel data from 11 Chinese coastal provinces (city or district) between 2009 and 2018. Results found that: (1) Digital economy promoted the sustainable development of land and marine binary economies in coastal areas; (2) Digital economy improved the factor mismatch of land and marine binary economies, which further affected the sustainable economic development; (3) Market integration is conducive to alleviating land and marine factor mismatch and strengthening the optimization effect of the digital economy on the factor mismatch. This research provides a new perspective for clarifying the mechanism of the digital economy on sustainable economic development, as well as a reference for the realization of rational allocation of factor resources and sustainable economic development by taking factor mismatch of land and marine binary economies and market integration as the intermediary variables and regulatory variables.

Highlights

• •Digital economy promotes the sustainable development of binary economy in China's coastal areas.
• •Digital economy improves binary economic factor mismatches.
• •Market integration strengthens the above positive effect.

1. Introduction

The prosperity of the digital economy inject a new impetus for the green development of economy in the past decade [1]. The advantages of digital economy, such as factors, carrier platform, and technological innovation [2,3], have become the keys to optimizing the allocation of land and marine factors in the spatial and temporal dimensions, as well as the spatial layout of traditional land and marine industries [4]. After all, the dilemma of resource shortage in the land industry and the abundance but low utilization of marine resources has not been solved. China has a long coastline. The Opinions of the Central Committee of the Communist Party of China and the State Council on Building a More Comprehensive Market-oriented Allocation System and Mechanism for Factors, as well as the “14th Five Year Plan for the Development of the Digital Economy”, has proposed to accelerate the cultivation of a data factor market, enhance the value of data resources, and highlight the key position of data elements in the social economy. The implementation of multiple provincial-level sea area use rights trading service specification documents releases signals marketization reforms in those areas. Therefore, optimizing the allocation of land and marine resources through the digital economy and promoting sustainable economic development is of great significance.

The digital economy's impact on sustainable economic development has been under hot debate [5]. The Solo Paradox exemplifies the negative correlation viewpoint. Acemoglu et al. [6] demonstrated that the efficiency of labor and capital invested in the IT industry decreased. Sharpe et al. [7] also concluded that the digital economy had no significant productivity growth effect and that the digital economy's productivity growth effect could be realized when the digital economy progressed from the “installation stage” to the “mature development stage”. As for the positive correlation viewpoint, Bohlin [8] and Thompson et al. [9] reported that the development of the telecommunications and mobile communication industries showed a significant positive correlation with economic growth. Ivus et al. [10] and Dale et al. [11] also found a significant positive correlation between E-commerce applications, information technology investment, and economic growth. Most positive correlations opinions agreed that data were essential in the digital economy [12,13] because it subverted the traditional enterprises by breaking the market structure through mechanism innovation at the micro level, promoted the trading market closer to complete competition at the intermediate level, and optimized the resource allocation at the macro level, thus achieving sustainable economic development [14,15]. Several studies found that digital economy had a nonlinear relationship with economic growth. Consumer economy [16], financial economy [17] and other studies discovered a “U” type nonlinear relationship. Moreover, Solomon and Klyton [18] distinguished between the impact of individual, business, and government ICT usage on growth and showed that only individual usage has a positive impact. Therefore, the relationship between digital economy and economic growth is dynamic and changes with economic evolution.

The deepening land and marine industrial specialization implies greater land and marine factors differentiation and increased transfer barriers and costs. The factor mismatch has long existed in space, industries, and sectors of land and marine economies (binary economies). The traditional concept of “seeing the marine by land and determining the marine by land” has caused the structural supply-and-demand imbalance between the binary economies, reducing the possibility of labor and capital factors realizing the optimal balance through the market mechanism and dragging down the overall factor allocation efficiency of the economic system. Factor supplies exceed the land economy demand, weakening factor endowment advantage and diminishing scale return [19]. However, barriers, such as a high access threshold for marine factors and a relatively poor working environment, hinder the free flow of factors between the binary economies, leading to the coexistence of “labor shortages”, insufficient financing and R&D resources in the marine economy, and excessive factor resource allocation in the land economy [20,21]. The structural imbalance of factor allocation in the binary economies [22] restricts the ME's driving ability to land economy and the sustainable economic development of coastal areas.

Extensive researches on the impact of marketizaion on the economic have been carried out. Scholarship defines marketization in varying ways and privatization and competition are the key merits of marketization [23]. The spatial mobility of resources such as labor [23], capital, land [24], energy [25], coastlines, and carbon emissions rights [26] has attracted much attention. Most scholars agree that marketization has a positive effect on economic integration [27]. This is mainly because the degree of market perfection affects the threshold and transaction costs of resource spatial mobility, thereby affecting the industrial supply chain, value chain [28], and even the willingness of market entities to cooperate across regions.

To summarize, numerous studies have been conducted on the digital economy's impact on economic sustainable development and factors configuration in binary economies; however, there is still room for improvement. One is neither sufficient nor strong evidence has been provided on how integrating land and marine factors with the digital economy affects sustainable economic development. Therefore, the direction of policy adjustment is unclear. Second, research on the marketization level's impact was insufficient, which is important in integrating the digital economy and real economy.

The coastal area economy was divided into two economic subsystems, i.e., the binary economies, and the influence mechanism of digital economy, factor mismatch of the binary economies, and green total factor productivity (GTFP) in coastal areas were analyzed. GTFP and factor mismatch of the binary economies in China's coastal areas were measured using the panel data from 2009 to 2018. The digital economy's impact on the factor mismatch and the GTFP, the intermediary effect of factor mismatch on the digital economy's impact on the GTFP, and the marketization level's regulating effect were tested using the intermediary effect and the regulating effect models.

The possible marginal contributions of this paper include, First, an innovative research perspective. The digital economy, the factor mismatch of binary economies, and the GTFP of binary economies are incorporated into one analysis framework, advancing the research on the digital economy's direct and intermediary effects on the GTFP. Second, the research on the regulating effect of marketization level has expanded the research on the digital economy's impact on sustainable economic development. Third, using the binary economies in China's coastal areas as the research object, this study surpasses the single economy conception in the existing research and meets the requirements for implementing the marine economy strategy.

2. Theoretical analysis and research hypothesis

2.1. The influence mechanism of the digital economy on sustainable economic development

The digital economy is a new form of the eco-friendly economy [29], and its positive effect on sustainable economic development could be as follows: First, the digital economy's universal applicability and embedding function and the data factor input reduce the demand for resource-intensive factors and the environmental impact of economic activities. Second, the digital economy enables industrial digitization by integrating into the R&D, production and sales activities, promoting industrial transformation and upgradation, especially improving the technology- and knowledge-intensiveness of traditional industries and industrial structure [30,31]. Moreover, the digital economy reduces industrial energy consumption, alleviates the problems of “high pollution, high energy consumption, and high emission” and the deterioration of the ecological environment, realizes the green transformation of production, and improves the overall GTFP. Third, the use of digital technology promotes the intensity of green environmental supervision [32,33], strengthens the dissemination of the concept of green environmental protection, and thus creates a green and friendly social ecology to ensure green economic development.

Conversely, the digital economy has a negative effect on GTFP. The application of digital technology requires adjustment of production factors in all production departments, and the marginal utility of these factors is decreasing, causing a weakening positive effect of digital economy on GTFP. Owing to the characteristics of self-expansion and external effects of digital economy [34], and the failing law of diminishing marginal utility, the role of digital economy in improving GTFP will be highlighted again when digital technology reaches a critical scale that can play the external effects [35]. For example, the development of the Internet has a significant double threshold effect on efficiency [36]. When the network population shifts from low to medium and then to large scale, the development of the Internet positively impacts efficiency, but the coefficients first decrease and then increase [37].

In summary, the digital economy's effect on sustainable economic development is uncertain and varies with the degree of industrial digitization. As the digital economy has entered a mature development period, the development of the low-carbon industry has also achieved remarkable results [38,39]. Therefore, the following hypothesis is proposed.

Hypothesis 1

Digital economy positively affects the GTFP of binary economies.

2.2. Factor mismatch has an intermediary effect on the digital economy's impact on sustainable economic development

The high permeability and synergy of digital economy improve the efficiency and accuracy of factor allocation [40], optimizing factor allocation, reducing energy input and environmental pollution, and contributing to sustainable economic development. Therefore, the factor mismatch's intermediary effect on the impact of digital economy on sustainable economic development is mainly manifested in the following:

Furthermore, the digital economy affects the factor allocation of land and marine binary economic systems. First, digital economy affects the factor allocation efficiency of binary economies and alleviates the problem of factor mismatch. From the micro perspective, the digital economy relies on its advantages of digitization, intelligence, informatization, and networking to improve the factor allocation efficiency and production efficiency of traditional factors such as labor and capital [29,41]. More excess factors increase the likelihood of factor flow between binary economies. Second, digital economy promotes the coordination of binary economies and indirectly affects factor mismatch. Network effects of the digital economy [42,43] and integrating data factor with factor markets expand the allocation network. With the development of the industrial digitization process, the information flow of digital platforms, the logistics of the Internet of Things, the labor and capital flows in factor markets, and the mutual penetration of binary economies [44], i.e., the digitization of production factors strengthens the connection between binary economies and creates a coordinated network economic system. Driven by bilateral network externalities, digital economy plays its role of permeability and synergy, expands the boundary of the industrial chain division of labor, and the binary economies' connection, promotes the factor configuration, and decreases factor mismatch degree. Third, the application of digital technology alleviates the externality of information asymmetries between binary economies, reduces transaction costs, eliminates inter-industrial barriers, and improves the coordination between binary economies. The factor flow between binary economies stimulates inter-industrial factor market competition, alleviates inter-industrial factor allocation distortion, promotes production factor flow, and optimizes the factor allocation of binary economies.

Moreover, the allocation of land and marine binary economic factors is closely related to the sustainable development of the regional economy. First, the rational allocation of factors in a binary economy, and the input and application of new factors, such as data factors, can improve the allocation efficiency of factors in the binary economy, reduce the input of energy, and improve energy efficiency, promoting the optimization of regional green efficiency. Second, a rational factor allocation in binary economies causes energy prices to converge, thus stabilizing energy costs and improving GTFP. Third, optimizing the factor allocation efficiency in binary economies can reduce unexpected outputs such as pollutants, thus contributing to sustainable economic development.

Therefore, the following hypothesis is proposed.

Hypothesis 2

Factor mismatch has an intermediary effect on the digital economy's impact on sustainable economic development in binary economies.

2.3. The regulating effect of marketization on the digital economy's impact on factor mismatch

The establishment of digital economy and a unified national market slows the restriction of market segmentation on the flow and allocation of factors and has a regulating effect which is manifested in the following aspects:

First, marketization improvement, especially the factor market integration, improves the pricing mechanism of the factor market, which helps to improve the inter-regional production factor liquidity, adjusting the supply structure of regional production factors and narrowing the inter-regional factor income gap so that the factor price can reflect its marginal output value [45,46]. Therefore, in the factor market, the digital economy can achieve a high degree of industry agglomeration under the platform economy, overcome the space limitations, realize regional integration of factor supply and demand, and expand the factor allocation spatial network.

Second, the factor market competition will further optimize the factor market mechanism to realize inter-industrial interaction, competition, and coordination. Simultaneously, digital technology can reduce the conditions for realizing industrial agglomeration [47,48], promote the realization of information sharing between binary economies, and weaken their industrial boundaries as the Internet industry agglomeration deepens.

Third, the marketization level affects the competitiveness of market subjects. In an efficient market competition, market entities should apply digital, information, and automation technologies, promote and accelerate R&D activities and transformation, realize automatic and intelligent production and business, optimize their competitive strategies, and reduce homogeneous competition.

In conclusion, marketization construction can strengthen the internal connection between digital economy and factor mismatch. The following hypothesis is therefore proposed.

Hypothesis 3

Marketization has a regulatory effect on digital economy's impact on the factor allocation of binary economies.

3. Research design

3.1. Empirical model setting

3.1.1. Analysis of the direct effect of the digital economy on GTFP in coastal areas

Based on the aforementioned theoretical analysis and using binary economies in coastal areas as the research objects, the benchmark regression models (1) and (2) of the direct effect of digital economy on GTFP were set respectively:

$$LGtf{p}_{it}={\partial }_0+{\partial }_{11}Dig{e}_{it}+{\partial }_{21}{X}_{it}+{\delta }_t+{\mu }_i+{\epsilon }_{it}$$ (1)

$$MGtf{p}_{it}={\partial }_0+{\partial }_{12}Dig{e}_{it}+{\partial }_{22}{X}_{it}+{\delta }_t+{\mu }_i+{\epsilon }_{it}$$ (2)

where the degree of sustainable economic development was represented by GTFP; LGtfp_it was the GTFP of the land economy in year t of the province (city or district) i; MGtfp_it was the GTFP of marine economy in year t of the province (city or district); Dige_it was the digital economic development level in coastal provinces (city or district) i in year t; X_it was the control variables, including market index (Mar) [49], industrial concentration level (Indc) [50], foreign investment level (Fdi) [51], level of government fiscal expenditure (Gov) [52] and human capital stock (Edu) [53]; δ_t was the fixed effect of time; μ_i was the fixed effect of the region; and ε was a random disturbance term.

3.1.2. Intermediary effect model of factor mismatch

The intermediary variables land economic factor mismatch (Lr_it) and marine factor mismatch (Mr_it) were added to the models (1) and (2), respectively, to investigate the intermediary role of factor mismatch in digital economy's impact on GTFP in binary economies, and the models (3)–(6) were set as follows:

$$L{r}_{it}={\alpha }_0+{\alpha }_{11}Dig{e}_{it}+{\alpha }_{21}{X}_{it}+{\delta }_t+{\mu }_i+{\epsilon }_{it}$$ (3)

$$M{r}_{it}={\alpha }_0+{\alpha }_{12}Dig{e}_{it}+{\alpha }_{22}{X}_{it}+{\delta }_t+{\mu }_i+{\epsilon }_{it}$$ (4)

$$LGtf{p}_{it}={\beta }_0+{\beta }_{11}Dig{e}_{it}+{\beta }_{21}{X}_{it}+{\beta }_{31}M{r}_{it}+{\beta }_{41}L{r}_{it}+{\delta }_t+{\mu }_i+{\epsilon }_{it}$$ (5)

$$MGtf{p}_{it}={\beta }_0+{\beta }_{12}Dig{e}_{it}+{\beta }_{22}{X}_{it}+{\beta }_{32}M{r}_{it}+{\beta }_{42}L{r}_{it}+{\delta }_t+{\mu }_i+{\epsilon }_{it}$$ (6)

Lr_it and Mr_it were the factor mismatch degrees of land and marine economies in year t in the province (city or district) i, respectively.

3.1.3. The regulating effect model for marketization

To determine how the degree of marketization regulates digital economy's impact on the GTFP of binary economies in coastal areas, the degree of regulatory variable marketization (Mar_it) and the cross term of marketization with digital economy (Dige_it) were added to the models (3)–(6) and set the regulating effect models (7)–(10) as follows:

$$L{r}_{it}={\alpha }_0+{\alpha }_{11}Dig{e}_{it}+{\alpha }_{21}Ma{r}_{it}+{\alpha }_{31}Ma{r}_{it}*Dig{e}_{it}+{\alpha }_{41}{X}_{it}+{\delta }_t+{\mu }_i+{\epsilon }_{it}$$ (7)

$$M{r}_{it}={\alpha }_0+{\alpha }_{12}Dig{e}_{it}+{\alpha }_{22}Ma{r}_{it}+{\alpha }_{32}Ma{r}_{it}*Dig{e}_{it}+{\alpha }_{42}{X}_{it}+{\delta }_t+{\mu }_i+{\epsilon }_{it}$$ (8)

$$LGtf{p}_{it}={\beta }_0+{\beta }_{11}Dig{e}_{it}+{\beta }_{21}Ma{r}_{it}+{\beta }_{31}Ma{r}_{it}*Dig{e}_{it}+{\beta }_{41}{X}_{it}+{\beta }_{51}M{r}_{it}+{\beta }_{61}L{r}_{it}+{\delta }_t+{\mu }_i+{\epsilon }_{it}$$ (9)

$$MGtf{p}_{it}={\beta }_0+{\beta }_{12}Dig{e}_{it}+{\beta }_{22}Ma{r}_{it}+{\beta }_{32}Ma{r}_{it}*Dig{e}_{it}+{\beta }_{42}{X}_{it}+{\beta }_{52}M{r}_{it}+{\beta }_{62}L{r}_{it}+{\delta }_t+{\mu }_i+{\epsilon }_{it}$$ (10)

3.2. Definition of variables and data sources

3.2.1. GTFP of the binary economies in coastal areas

Based on the system theory and considering the undesired output influence of TFP, the GTFP growth rate in the land economy (LGtfp_it) and GTFP growth rate in marine economy (MGtfp_it) were calculated using panel data from 11 coastal provinces (city or district) between 2008 and 2018 and applying the Directional Distance Function (DDF) based Global Malmquist-Luenberger index model [54]. Using 2008 as the base period, the comparable efficiency, i.e., the GTFP of binary economies between 2009 and 2018, was used to represent the sustainable development level of land and marine economies, respectively. The evaluation index system is shown in Table 1.

Table 1.

Green Total factor productivity of binary economies in coastal areas.

System-level	Functional layer	State layer	Meaning	Index layer	Data sources
Land economy	Investment index	Land	Land for land construction	Urban, the county seat, urban construction land and cultivated land area	Urban and rural construction statistical Yearbook, Guo Tai'an Database and China Statistical Yearbook
		Capital	Land capital stock	Investment in fixed assets in the whole land area	China Statistical Yearbook
		Labour force	Land employment	Number of people employed in the three industries	China Statistical Yearbook
		Technology	Land technology investment	Number of patent applications accepted	The Statistical Annual Report of the State Intellectual Property Office
	Output indicators	Economic output	Land economic GDP	Gross economic product	China Statistical Yearbook
		Undesired output	Land pollution emissions	The discharge amount of industrial wastewater	China Statistical Yearbook
The Marine economy	Investment index	Sea area	Sea area	The sea area	The China Marine Statistical Yearbook
		Capital	The stock of marine capital	Investment in the total social fixed assets in the sea areas	The China Marine Statistical Yearbook
		Labour force	The number of employed persons involved in the sea	The number of employed people involved in the sea	China Statistical Yearbook
		Technology	Investment in marine science and technology	Number of patent applications accepted	The China Sea Areas Statistical Yearbook
	Output indicators	Economic output	The GDP of the marine economy	Gross economic product	The China Marine Statistical Yearbook
		Undesired output	Pollution emissions from sea areas	Industrial wastewater discharged into mass	The Bulletin of China's Marine Ecological Environment State

3.2.2. Digital economy level (Dige)

The digital economy development level measurement is abundant. Existing studies have diverse evaluation index systems due to differences in research perspectives and objectives. Zhao [55], Huang [56] and Liu [57] measured the digital economy development level from the perspectives of Internet development and digital finance. Zhao [55] measured the digital economy development level using the number of Internet users. Yang [58] developed a comprehensive evaluation index system based on digital industrialization and industrial digitalization. Other scholars [[59], [60], [61]] used China's “Internet +” digital economy Index compiled by Tencent Research Institute.

Given the digital economy's collaboration, subversion, and reshaping functions on land-marine factors, four types of indices were used to build the digital economy development level index system: the digital economy development foundation, data sharing media, digital industrialization, and industrial digitalization. The index system is shown in Table 2. The Topsis entropy method [62] was utilized for dimension reduction and the evaluation of evaluation.

Table 2.

Index system of digital economic development level in coastal areas.

Objective	First level indicators	Secondary indicators	Data sources
Digital economy	Development foundation	Mobile phone exchange capacity	Guo Tai'an database
		Local switch capacity
		Long-distance optical cable line length
		The Internet broadband access port
	Data sharing media	Number of mobile phone users
		Total telecom business
		Express business volume
	Digital industrialization	Number of software and information technology service enterprises
		Revenue from software products
	Industrial digitization	Expenditure on technical and technological transformation of regional above-designated industrial enterprises
		Expenditure by industrial enterprises in each region to purchase domestic technology funds
		The number of websites	Statistical Report on the Development of the Internet in China
		Internet penetration rate

3.2.3. Mismatching of land and marine factors (Lr and Mr)

By referring to Bai et al. [61], the panel regression model of variable intercept and the variable slope was used to estimate the output elasticity of land and marine factors, based on the least squares virtual variable method (LSDV). The binary economic comparison coefficient was used to estimate the factor mismatch degree compared with the benchmark economic system. The calculation process is as follows.

Calculation of factor output elasticity. The panel regression model of variable intercept and the variable slope was used to estimate the factor output elasticity of each province (β_ki and β_li) using the residual method; the C-D function assumed that the scale remains unchanged, and the least squares virtual variable method (LSDV). The calculation model was as follows:

$$Y_{it}={AK}_{it}^{{\beta }_{ki}}{L}_{it}^{1-{\beta }_{ki}}$$ (11)

The natural logarithm was taken on both sides of Eq. (11), and the individual effect μ_it and the time effect ə_it were added. Equation (12) was set as follows:

$$\mathit{Ln}\left(\frac{Y_{it}}{L_{it}}\right)=LnA+{\beta }_{ki}\mathit{Ln}\left(\frac{{\mathrm{K}}_{it}}{L_{it}}\right)+{\mu }_{\text{it}}+{\partial }_{it}+{\epsilon }_{it}$$ (12)

where Y_it represented the GDP of binary economies, L_it was the number of employments in binary economies, and K_it was the capital stock of binary economies.

The distortion of land and marine capital and labor factors. After calculating the output elasticity of factors in binary economies, the binary economic comparison coefficient was used to represent the factor mismatch. (13), (14) were used:

$$L{r}_{it}=\left|\left(\frac{{\beta }_{oil}}{{\beta }_{oik}}\times \frac{K_{oit}}{L_{oit}}\right)/\left(\frac{{\beta }_{1il}}{{\beta }_{1ik}}\times \frac{K_{1it}}{L_{1it}}\right)\right|$$ (13)

$$M{r}_{it}=\left|\left(\frac{{\beta }_{oil}}{{\beta }_{oik}}\times \frac{K_{oit}}{L_{oit}}\right)/\left(\frac{{\beta }_{2il}}{{\beta }_{2ik}}\times \frac{K_{2it}}{L_{2it}}\right)\right|$$ (14)

where 0 was the overall economic system in coastal areas (the benchmark economic system); 1 and 2 were the land and marine economies, respectively; β_iK and β_iL were the output elasticity of capital and labor of i province (city or district), respectively; K_it and L_it were the capital and labor inputs of c economic system in year t of the province (city or district) i; Lr_it and Mr_it were the factor mismatch degree of land and sea in year t of the province (city or district) i, i.e., the relative factor mismatch coefficient between systems, may indicate “too much labor and too little capital” or “too much capital and too little labor”, compared to the benchmark economic system. When both L_r and M_r were less than 1, it indicated “too much capital and too little labor”. When both L_r and M_r were greater than 1, it indicated “too little capital and too much labor”. When both L_r and M_r were equal to 1, it indicated no factor mismatch, similar to the benchmark economic system. The closer the L_r and M_r coefficients were to 1, the more reasonable factor allocation was in the binary economies, whereas the higher factor mismatch degree, the greater the deviation from the ideal state.

3.2.4. Control variables

To control other variables that may affect GTFP in coastal areas, the following control variables were selected: market index (Mar) [49] (measured by adopting cross-year comparable data processed by technical connection in China inter-provincial market index database), industrial concentration level (Indc) [50] (measured by the ratio of the number of employed personnel to the area of the administrative division), foreign investment level (Fdi) [51] (measured by the proportion of total inter-provincial foreign investment in GDP), level of government fiscal expenditure (Gov) [52] (measured by the ratio of local financial general budget expenditure to regional GDP), human capital stock (Edu) [53] (measured using the following formula: average years of education = (illiterate number * 1 + number of primary school education * 6 + number of junior high school education * 9 + number of high school and technical secondary school education * 12 + number of college degree or above * 16)/total population over 6 years old).

3.2.5. Data source and processing

The study area comprised 11 coastal provinces (city or district): Tianjin, Hebei, Liaoning, Shanghai, Jiangsu, Zhejiang, Fujian, Shandong, Guangdong, Guangxi, and Hainan (excluding Hong Kong, Macao Special Administrative Region, and Taiwan). The data were obtained from the China Statistical Yearbook, statistical yearbook of provinces (city or district) in coastal areas, and Wind and Guotai'an databases. The regional GDP and gross ocean product were reduced using the regional GDP index and 2000 as the base period to reduce the influence of price fluctuations. The expedition period was between the year 2009–2018 owing to the availability of data. Variables description statistics are shown in Table 3.

Table 3.

Descriptive statistics of variables.

Variable	Objects	Mean	Standard deviation	Min.	Max.
LGtfp	110	1.0359	0.0697	0.9069	1.2723
MGtfp	110	0.9584	0.1758	0.5503	1.3649
Lr	110	1.9933	1.5936	0.5328	6.4647
Mr	110	0.1933	0.4755	0.0532	1.9442
Dige	110	0.0909	0.0531	0.0024	0.2317
Mar	110	9.0049	1.5038	5.5500	11.3800
Gov	110	0.1776	0.0602	0.0964	0.3500
Indc	110	0.2083	0.2383	0.0490	1.0000
Fdi	110	0.0500	0.0540	0.0000	0.2710
Edu	110	9.3695	0.7848	7.9030	11.4560

Since the total fixed asset investment value of the whole population has not been published in the China Statistical Yearbook since 2019, the data were estimated using the growth rate of fixed asset investment (excluding rural households) over the previous year. The perpetual inventory method was used to estimate the capital stock of provinces (cities, regions) by taking 2000 as the base period [63]. Moreover, the faster the regional economic or science and technology development, the faster the replacement of fixed assets; therefore, the capital depreciation rate was represented by the inter-provincial depreciation rate [64]. Based on the system theory, the input and output values of the land economy were obtained by deducting those of the marine economies from the total economic system.

4. Empirical results and analysis

4.1. Description of the mismatch of land-marine factors

Table 4 shows the factor mismatch between binary economies in coastal areas from 2009 to 2018. The factor mismatch coefficient of the binary economies in most coastal provinces generally deviated from the benchmark state. The factor mismatch coefficient of the land economy was closer to 1 compared to that of the marine economy except for Guangdong and Guangxi, indicating that the factor allocation in the land economy was more reasonable. The factor mismatch coefficient of the land economy was greater than 1 except for Fujian and Hainan, showing that the factor allocation in the land economy was “too much labor and too little capital”. Factor mismatch in the marine economy was less than 1 except for Tianjin, indicating that factor allocation in the marine economy was “too little labor force and too much capital”. The results of factor mismatch were deducted by 1 to facilitate the analysis, making the final factor mismatch results deviate from 1.

Table 4.

The degree of factor mismatch in binary economies.

		2009	2010	2011	2012	2013	2014	2015	2016	2017	2018
Lr	Tianjin	1.7317	1.7583	1.7718	1.7908	1.8107	1.8217	1.8274	1.8259	1.8189	1.8129
	Hebei	1.3624	1.3622	1.3623	1.3629	1.3632	1.3630	1.3628	1.3626	1.3621	1.3617
	Liaoning	1.1371	1.1360	1.1359	1.1378	1.1422	1.1432	1.1307	1.1203	1.1172	1.1185
	Shanghai	1.6878	1.6880	1.6847	1.6834	1.6858	1.7469	1.7427	1.7406	1.7392	1.7387
	Jiangsu	1.6919	1.6906	1.6893	1.6883	1.6874	1.6866	1.6859	1.6851	1.6844	1.6838
	Zhejiang	1.3301	1.3281	1.3263	1.3248	1.3233	1.3217	1.3209	1.3204	1.3209	1.3207
	Fujian	0.5328	0.5338	0.5422	0.5455	0.5435	0.5462	0.5498	0.5499	0.5495	0.5486
	Shandong	1.6411	1.6401	1.6389	1.6385	1.6372	1.6362	1.6353	1.6343	1.6315	1.6220
	Guangdong	3.4059	3.4098	3.4160	3.4197	3.4238	3.4325	3.4387	3.4457	3.4554	3.4457
	Guangxi	6.4647	6.4634	6.4608	6.4416	6.4393	6.4375	6.4374	6.4368	6.4341	6.4337
	Hainan	0.6327	0.6356	0.6413	0.6510	0.6627	0.6725	0.6755	0.6744	0.6825	0.6728
Mr	Tianjin	1.9442	1.8515	1.8045	1.7385	1.6692	1.6308	1.6111	1.6161	1.6405	1.6614
	Hebei	0.0752	0.0756	0.0754	0.0741	0.0733	0.0738	0.0744	0.0748	0.0759	0.0768
	Liaoning	0.3404	0.3426	0.3428	0.3391	0.3306	0.3287	0.3528	0.3730	0.3791	0.3765
	Shanghai	0.1582	0.1581	0.1595	0.1600	0.1591	0.1340	0.1357	0.1366	0.1371	0.1373
	Jiangsu	0.1232	0.1255	0.1281	0.1298	0.1314	0.1329	0.1343	0.1356	0.1370	0.1380
	Zhejiang	0.1849	0.1871	0.1891	0.1908	0.1924	0.1943	0.1952	0.1956	0.1952	0.1954
	Fujian	0.1460	0.1447	0.1346	0.1307	0.1331	0.1299	0.1255	0.1254	0.1259	0.1269
	Shandong	0.0848	0.0854	0.0861	0.0864	0.0872	0.0878	0.0883	0.0889	0.0906	0.0964
	Guangdong	0.0720	0.0715	0.0707	0.0702	0.0696	0.0685	0.0677	0.0667	0.0655	0.0667
	Guangxi	0.0532	0.0535	0.0540	0.0581	0.0586	0.0589	0.0590	0.0591	0.0597	0.0597
	Hainan	0.7690	0.7604	0.7440	0.7157	0.6817	0.6531	0.6443	0.6476	0.6240	0.6524

4.2. Analysis of the estimation results of the empirical model

Considering the spatial heterogeneity between digital economy and land-marine factor mismatch, a fixed effects model was used for empirical evidence.

4.2.1. Benchmark regression results and their analysis

Table 5 shows the benchmark regression results of models (1) and (2). The results demonstrated that digital economy positively affects the GTFP of the binary economies, indicating that digital economy development is conducive to GTFP improvement in both binary economies in the coastal areas. This confirms hypothesis 1.

Table 5.

Benchmark regression results of digital economy affecting GTFP of binary economy.

	(1)	(2)
	LGtfp	MGtfp
Dige	0.105***	0.243***
	[0.0242]	[0.0475]
Mar	−0.381***	0.079
	[0.1282]	[0.2060]
Gov	−0.87	−1.312*
	[0.5398]	[0.6960]
Indc	−0.0383	0.0774
	[0.1158]	[0.2552]
Fdi	0.0011	6.96E-04
	[0.0017]	[0.0047]
Edu	−0.206	−0.147
	[0.2887]	[0.5948]
cons	2.888***	1.989
	[0.6792]	[1.4805]
Individual fixation effect	Yes	Yes
Time fixed effect	Yes	Yes
N	110	110
Adj.R - sq	0.7044	0.8168
AIC	−385.7	−234.6
BIC	−315.5	−164.4

Note: *p < 0.1, * *p < 0.05, * * *p < 0.01, indicated in parentheses, the same below.

4.2.2. Intermediary effect results analysis and significance test

The stepwise test method was used to evaluate the mechanism of digital economy on GTFP of binary economies in the coastal areas. The results of intermediary effects are shown in Table 6, where columns (1)–(3) contain the results of models (1), (3), and (5), whereas columns (4)–(6) contain the results of models (2), (4), and (6).

Table 6.

Results of the intermediary effect of digital economy on GTFP of land-marine economies.

	(1)	(2)	(3)	(4)	(5)	(6)
	LGtfp	Lr	LGtfp	MGtfp	Mr	MGtfp
Dige	0.105***	−0.023**	0.090***	0.243***	0.065**	0.278***
	[0.0242]	[0.0097]	[0.0231]	[0.0475]	[0.0294]	[0.0426]
Lr			−1.497***			−0.877
			[0.4415]			[0.9318]
Mr			−0.291*			−0.860***
			[0.1507]			[0.2576]
X	Yes	Yes	Yes	Yes	Yes	Yes
cons	2.888***	1.218***	5.525***	1.989	2.801***	5.467**
	[0.6792]	[0.2097]	[1.0878]	[1.4805]	[0.6075]	[2.2109]
Individual fixation effect	Yes	Yes	Yes	Yes	Yes	Yes
Time fixed effect	Yes	Yes	Yes	Yes	Yes	Yes
N	110	110	110	110	110	110
Adj.R - sq	0.7044	1.0000	0.7399	0.8168	0.9948	0.8317
AIC	−385.7	−624.8	−398.4	−234.6	−400.8	−242.6
BIC	−315.5	−554.6	−322.8	−164.4	−330.6	−167.0

4.2.2.1. The impact of digital economy on factor mismatch

The results in column (2) show that the coefficient of Lr and Dige was −0.0228, indicating that digital economy reduced factor mismatch in the land economy. The results of column (4) show that the coefficient of Mr and Dige was 0.0647, showing that digital economy improved factor mismatch in ME. As mentioned above, the land economy was in the “too much labor and too little capital” stage (i.e., its factor mismatch coefficient was greater than 1), while the marine economy is in the “too little labor and too much capital” stage (i.e., its factor mismatch coefficient is less than 1). Therefore, the results indicated that digital economy was conducive to promoting the convergence of the factor mismatch of binary economies to 1, respectively, improving the factor mismatch of land factors.

4.2.2.2. The intermediary role of factor mismatch in the digital economy's impact on sustainable economic development

The coefficient of Dige and LGtfp in column (3) was 0.0897 and passed the significance test, showing that digital economy has a significant positive effect on GTFP in the land economy. While the coefficient of Lr and Mr with LGtfp were −1497 and −0.291, respectively, showing that factor mismatch significantly negatively affected the GTFP of binary economies.

The coefficient of Dige and MGtfp in column (6) was 0.278 and passed the significance test, indicating that digital economy had a significant positive effect on GTFP in marine economy. The coefficients of Lr and Mr with LGtfp were −0.877 and −0.860, respectively, but only the former was statistically significant. It showed that the digital economy alleviated factor mismatch in marine economy and improved its GTFP. This confirms Hypothesis 2 in this paper.

4.2.3. Results of the regulating effect of marketization level

Table 7 shows the results of the marketization's regulating effect on the digital economy's impact on the GTFP of binary economies.

Table 7.

Results of the regulatory effect of marketization on the digital economy's impact on GTFP of binary economies.

	(1)	(2)	(3)	(4)
	Lr	Mr	LGtfp	MGtfp
Dige	−0.036***	0.097***	0.062**	0.227***
	[0.0082]	[0.0286]	[0.0267]	[0.0585]
Mar	0.067**	−0.111	−0.293**	−0.066
	[0.0295]	[0.0965]	[0.1347]	[0.1981]
Dige*Mar	−0.014***	0.034***	−0.018*	−0.034
	[0.0029]	[0.0099]	[0.0097]	[0.0242]
Lr			−1.788***	−1.430
			[0.4906]	[1.0379]
Mr			−0.236	−0.757***
			[0.1565]	[0.2672]
X	Yes	Yes	Yes	Yes
_cons	1.106***	3.077***	5.581***	5.572**
	[0.1927]	[0.5570]	[1.1283]	[2.2903]
Individual fixation effect	Yes	Yes	Yes	Yes
Time fixed effect	Yes	Yes	Yes	Yes
N	110	110	110	110
A dj.R - sq	1.0000	0.9960	0.7488	0.8364
AIC	−662.3	−428.6	−401.6	−245.1
BIC	−589.4	−355.7	−323.3	−166.8

4.2.3.1. Land economy results

Column (1) shows that after adding the cross term of digital economy and market index (Dige*Mar), the Dige and Lr coefficient was −0.0360 and passed the significance test, showing that the digital economy had a significant negative effect on the factor mismatch in the land economy. The Dige*Mar and Lr coefficient was −0.0139 and passed the significance test, indicating that the higher the degree of marketization, the stronger the guiding role of the “invisible hand” on the price mechanism, and then strengthened the digital economy's impact to reduce the factor mismatch in the land economy, that is, marketization construction strengthened the positive effect of digital economy on the factor mismatch in the land economy.

Column (3) shows that after adding the cross term of the digital economy and market index (Dige*Mar), the Dige and LGtfp coefficient was 0.062 and passed the significance test, indicating that the digital economy had a significant positive effect on the GTFP of the land economy. The Dige*Mar and LGtfp coefficient was −0.0181 and passed the significance test, showing that market construction reduced the impact of digital economy on the GTFP of the land economy.

4.2.3.2. Results of the ME

Column (2) shows that after adding the cross term of digital economy and market index (Dige*Mar), the Dige and Mr coefficient was 0.0970 and passed the significance test, indicating that the digital economy had a significant positive effect on the factor mismatch of ME. The Dige*Mar and Mr coefficient was −0.0340 and passed the significance test, indicating that the higher the degree of marketization, the stronger the guiding role of the price mechanism and, consequently, the stronger the impact of digital economy on improving the factor mismatch of marine economy.

Column (4) shows that the Dige and MGtfp coefficient is −0.0343, which failed the significance test, indicating that market construction had no effect on the mechanism of digital economy on the GTFP of ME.

5. Robustness test

5.1. Change of the core explanatory variable

The panel entropy method was used to remeasure the digital economy development level index to avoid the bias of the estimation results caused by calculation method selection, and the factor mismatch's intermediary effect was retested. The results are shown in Table 8, consistent with the previous research, demonstrating the validity of those findings.

Table 8.

Robustness test of replacing DE.

	(1)	(2)	(3)	(4)	(5)	(6)
	LGtfp	Lr	LGtfp	MGtfp	Mr	MGtfp
Dige	0.068***	−0.011*	0.060***	0.125**	0.032**	0.137***
	[0.0183]	[0.0066]	[0.0135]	[0.0475]	[0.0147]	[0.0474]
Lr			−1.532***			−1.105
			[0.4365]			[0.9282]
Mr			−0.281*			−0.777***
			[0.1443]			[0.2731]
X	Yes	Yes	Yes	Yes	Yes	Yes
cons	2.475***	1.283***	5.174***	1.246	2.614***	4.695**
	[0.6645]	[0.1967]	[1.0697]	[1.5098]	[0.5700]	[2.1757]
Individual fixation effect	Yes	Yes	Yes	Yes	Y es	Y es
Time fixed effect	Yes	Yes	Yes	Yes	Yes	Yes
N	110	110	110	110	110	110
Adj.R - sq	0.7098	1.0000	0.7490	0.7996	0.9946	0.8090
AIC	−387.7	−620.8	−402.3	−224.8	−396.8	−228.7
BIC	−317.5	−550.6	−326.7	−154.6	−326.6	−153.1

5.2. Change of the explanatory variable

The GTFP in coastal areas was measured using the global DEA-Malmquist index method, and then the intermediary effect of factor mismatch was retested. The results are shown in Table 9, consistent with the previous study, demonstrating the robustness of those findings.

Table 9.

Robustness test of replacing GTFP.

	(1)	(2)	(3)	(4)	(5)	(6)
	LGtfp	Lr	LGtfp	MGtfp	Mr	MGtfp
Dige	0.237***	−0.023**	0.272***	0.066**	0.065**	0.118***
	[0.0447]	[0.0097]	[0.0402]	[0.0289]	[0.0294]	[0.0406]
Lr			−0.397*			0.096
			[0.8287]			[0.3570]
Mr			−0.676***			−0.767***
			[0.2318]			[0.1421]
X	Yes	Yes	Yes	Yes	Yes	Yes
cons	2.354*	1.218***	4.729**	0.447	2.801***	2.477**
	[1.3440]	[0.2097]	[2.0078]	[0.9818]	[0.6075]	[0.9858]
Individual fixation effect	Yes	Yes	Yes	Yes	Y es	Y es
Time fixed effect	Yes	Yes	Yes	Yes	Yes	Yes
N	110	110	110	110	110	110
Adj.R - sq	0.8320	1.0000	0.8436	0.879	0.9948	0.9081
AIC	−254.7	−624.8	−261.3	−303.8	−400.8	−332.7
BIC	−184.5	−554.6	−185.6	−233.6	−330.6	−257.1

5.3. Change of the parameter estimation

The results of fixed effect estimation might have an endogenous bias due to endogenous problems of reverse causality between digital economy and GTFP, and insufficient consideration has been given to the influencing factors of GTFP. The following parameters were used separately before testing: The number of fixed phones per 100 people and the number of post offices per million people were constructed with the national software product revenue as the tool variable of the digital economy (IV 1 and IV 2) [65]. The results are shown in Table 10, which are consistent with the previous research.

Table 10.

Robustness test of substitution parameter estimation method (IV).

	(1)	(2)	(3)	(4)	(5)	(6)
	LGtfp	Lr	LGtfp	MGtfp	Mr	MGtfp
Dige	0.189***	−0.033**	0.180***	0.346***	0.098***	0.419***
	[0.0462]	[0.0129]	[0.0465]	[0.0540]	[0.0352]	[0.0719]
Lr			−1.314***			−0.592
			[0.4288]			[0.8861]
Mr			−0.366***			−0.977***
			[0.1391]			[0.2300]
X	Yes	Yes	Yes	Yes	Yes	Yes
cons	2.863***	1.221***	5.487***	1.958	2.791***	5.407***
	[0.7396]	[0.2055]	[1.0645]	[1.2864]	[0.5983]	[2.0275]
Individual fixation effect	Yes	Yes	Yes	Yes	Yes	Yes
Time fixed effect	Yes	Yes	Yes	Yes	Yes	Yes
N	110	110	110	110	110	110
Adj.R - sq	0.6634	1.0000	0.6957	0.8070	0.9947	0.8147

Second, the GTFP values are merged data ranging from 0 to 2, so the Tobit regression model regression test might be more appropriate. The results are shown in Table 11, which are consistent with the previous research.

Table 11.

Robustness test of substitution parameter estimation method (Tobit).

	(1)	(2)	(3)	(4)	(5)	(6)
	LGtfp	Lr	LGtfp	MGtfp	Mr	MGtfp
Dige	0.105***	−0.023***	0.090***	0.243***	0.065**	0.278***
	[0.0213]	[0.0085]	[0.0201]	[0.0417]	[0.0258]	[0.0370]
Lr			−1.497***			−0.877
			[0.3829]			[0.8082]
Mr			−0.291**			−0.860***
			[0.1307]			[0.2234]
X	Yes	Yes	Yes	Yes	Yes	Yes
cons	2.888***	1.218***	5.525***	1.989	2.801***	5.467***
	[0.5963]	[0.1841]	[0.9435]	[1.2997]	[0.5333]	[1.9177]
Individual fixation effect	Yes	Yes	Yes	Yes	Yes	Yes
Time fixed effect	Yes	Yes	Yes	Yes	Yes	Yes
N	110	110	110	110	110	110
AIC	−383.7	−622.8	−396.4	−232.6	−398.8	−240.6
BIC	−310.8	−549.9	−318.1	−159.7	−325.9	−162.3

6. Discussion

Economic sustainable development has been the subject of hot debate, and digital economy and factor allocation are important variables for sustainable economic development. Therefore, examining how the digital economy and factor allocation affect sustainable development in coastal areas is important.

First, the digital economy's positive effect on the GTFP of binary economies is consistent with Zhao et al. [66] and Han et al. [67]. These studies discovered that the digital economy integrated labor flow, logistics, capital flow, and data factors, promoted intelligence and digitization, and improved the GTFP level of binary economies.

The second reason for digital economy's effect of improving factor mismatch of the land economy might be the popularization and application of digital technology, which facilitated factor allocation in the industrial and spatial dimensions of the land economy. The other factor was the popularization and application of digital technology, which facilitated free factor flow between binary economies, and factors flew into marine economy due to the excessive factor density in the land economy, slowing down the factor mismatch degree [68].

Third, digital economy alleviated the factor mismatch in marine economy and improved its GTFP, consistent with Yu et al. [69] and Zhang et al. [70]. It could be because digital economy reduced the technical potential energy differences between the binary economies, improving their coordinated development. Although performance in marine economy was different, evidence showed that digital economy development improved GTFP in coastal areas by improving the factor mismatch of ME. Although both binary economies took the coastal zone as the space carrier, the port as the guide for economic development, science and technology as the efficiency support, and material and energy gradients flow as the link [71], the reason for the different performance in the binary economies could be that the improved land factor mismatch had a weak feedback effect on the GTFP of marine economy due to the entry threshold and allocation cost transferring from land to ME, thus failed to promote its sustainable development.

Fourth, the reasons for marketization's heterogeneous regulatory effect on digital economy's impact on factor mismatch in binary economies (which strengthened the former but weakened the latter) could be: First, digital economy development and marketization construction had lags, and the information factor had two-way feedback lags that were insufficient to meet the requirements of economic changes; therefore, marketization construction weakened the positive effect of digital economy on GTFP in the short term [72,73]. Second, the weak recipient ability and insufficient talent [74] were not conducive to market-oriented digital economy construction, and the market's factor allocation function had not been fully played [75].

Finally, the failure of market construction to affect the mechanism of digital economy on the GTFP of marine economy may be due to the low conversion rate in the marine economy [76] and insufficient investment in marine science and technology R&D [77]. The scale advantage of marine scientific research resources had not been effectively transformed into market value and industrial advantages, resulting in a weak marketization construction effect.

7. The conclusions and recommendations

In this study, we investigated how digital economy affected the factor mismatch of binary economies, reducing the loss of GTFP and promoting sustainable economic development. After analyzing digital economy's impact on GTFP of binary economies in coastal areas, the factor mismatch's intermediary and marketization's regulating effects from a theoretical perspective, the intermediary effect and regulating effect models were used for empirical analysis using panel data from 11 provinces (city or district) in China's coastal area between 2009 and 2018, and the following conclusions were made.

• (1)Factor mismatch existed to a greater extent in the land economy in China's coastal areas than in ME, with “too much labor and too little capital” in the land economy but “too little labor and too much capital” in ME.
• (2)Digital economy had direct and intermediary effects on the GTFP in coastal areas. The digital economy directly improved the GTFP significantly and indirectly affected GTFP through a significant negative impact on factor mismatch in the land economy and a positive impact on the factor mismatch of ME.
• (3)The marketization level was conducive to alleviating the factor mismatch between binary economies. The marketization degree increased digital economy's impact on factor mismatch in the land economy, with no significant effect in ME.

Therefore, the following recommendations are made: (1) Apply digital technology in environment governance and business activities. A large number of new technologies such as digital technology, intelligent technology, and network technology shall be applied in the governance of atmospheric, water, energy, community, etc. to improve the efficiency of environmental governance and reduce the long-term cost of environmental protection for enterprises. (2) Accelerate market digitization construction. The construction of the Internet, Internet of Things, and digital infrastructures shall be strengthened in the factor markets, thus laying the foundation for the deep integration of digital and traditional factors. (3) Improve regional cooperation platforms, strengthening the cooperations between digital and traditional enterprises, and the cooperations between the marine and land companies. (4) Lower the threshold for factor flow. For example, in the human resources market, more attention should be paid to optimizing the access training for human resources in the marine industry, providing low-cost and easily accessible professional skills training for workers, and lowering the industry entry threshold.

This study has some limitations: First, it did not utilize inter-city data for empirical research because of data availability, and the accuracy of the results was limited. The official statistics of marine economy are expected to improve in future. Second, considering the strong spatial spillover effects of the digital economy, the spatial effects of the main variables need further testing.

Data availability statement

The data is available from https://doi.org/10.5061/dryad.69p8cz97r.

CRediT authorship contribution statement

Shujuan Wu: Writing – review and editing, Writing – original draft, Project administration, Conceptualization. Jianhua Tang: Writing – review and editing, Methodology, Data curation. Minmin Li: Writing – review and editing, Formal analysis. Jianhua Xiao: Project administration, Funding acquisition.

Declaration of competing Interest

Wu ShuJuan reports financial support was provided by Wuyi University. Tang JianHua reports a relationship with Wuyi University that includes: board membership.

Abbreviations

TFP

total factor productivity

GTFP

green total factor productivity, binary economies, land and marine economies