Frontiers in Computing and Intelligent Systems 
ISSN: 2832-6024 | Vol. 1, No. 1, 2022 

 

53 

The Digital Economy and the Regional Economic Gap: 
An Inverted U-shaped Relationship 

Ming Zhou* 

School of Economics, Anhui University of Finance and Economics, Bengbu Anhui, China 

* Corresponding author: Ming Zhou (Email: lyzyhy214@163.com) 

 

Abstract: As a new economic form, the digital economy is profoundly changing China, and clarifying the relationship between 

the digital economy and the regional economic gap is of great practical significance for coordinating regional economic 

development. Based on the perspective of new economic geography, the digital economy is incorporated into the core marginal 

model, and the impact of digital economy development on the change of regional economic gap under long-term equilibrium 

conditions is deduced Provincial panel data are used to support this. The study found that with the continuous development of 

the digital economy, the regional economic gap showed an inverted U-shaped trend of first expanding and then narrowing; 

Considering regional heterogeneity, the inverted U-shaped impact of the digital economy in the eastern, central and western 

regions on the regional economic gap is still obvious. Finally, this paper puts forward policy suggestions from the perspective of 

"the digital economy provides a solution to alleviate the imbalance and incongruity of regional development". 

Keywords: Digital economy; Regional economic gap; Inverted U-shape; New economic geography. 

 

1. Introduction 

General Secretary Xi Jinping pointed out: "We must 

soberly realize that the problem of unbalanced and 

insufficient development in China is still relatively prominent, 

and the gap between urban and rural development, regional 

development and income distribution is relatively large. The 

new round of scientific and technological revolution and 

industrial transformation has given a strong impetus to 

economic development, but also has a profound impact on 

employment and income distribution, including some 

negative effects, which need to be effectively addressed and 

addressed [1]. "As a new economic form, the digital economy 

has been written into the government work report for the fifth 

time, which is profoundly changing China, becoming a new 

engine to enhance people's happiness and promote high-

quality economic development, laying out a "digital picture" 

with a thousand meteorological conditions. However, 

whether the development of the digital economy can 

effectively alleviate the current "regional economic 

development imbalance" still needs theoretical explanations 

and empirical tests. Therefore, it is of great practical 

significance to comprehensively clarify the relationship 

between the digital economy and the regional economic gap 

for coordinating regional economic development.  

Digital technologies based on information technologies 

such as artificial intelligence, big data, and cloud computing 

are developing rapidly and play a key role in production, 

circulation and other fields. Based on the perspective of 

spatial economics, the development of the digital economy, 

through the monetary externality of "saving production and 

transaction costs" and the technical externality of "knowledge 

spillover" improve the efficiency of information and 

knowledge exchange, and Continuously promote economic 

convergence [2], affecting regional economic disparities 

fluctuation .Regarding the monetary externalities of the 

development of the digital economy, at the micro level, digital 

technologies have general purpose technologies(GPT). All of 

these characteristics can improve the operational efficiency 

and business coverage radius of the enterprise, and reduce the 

marginal production cost of the enterprise, significantly 

improve the efficiency of product transportation. In the face 

of diversified and fragmented individual needs, the 

production management and marketing model of the 

"industrial era" is slow to respond and high in cost. Under the 

influence of the digital economy, the structure of the 

production system of enterprises has undergone qualitative 

changes, and the production system has been divided into 

universal modules and heterogeneous element modules. In 

other words, the combination of the basic public elements of 

the product together constitutes a universal module, and the 

heterogeneous elements of other products in addition to the 

public production factors are combined to efficiently produce 

differentiated products with different functions or the same 

function but different performance, so as to cope with the 

diversified market demand and complete the benign 

transformation of the production and marketing model to 

precision and refinement [3], so as to improve the production 

efficiency of enterprises, and effectively improve the 

efficiency of meeting the diverse needs of users. On the 

technological spillover effects of the digital economy, how to 

innovate in the context of responding to uncertainties and 

responding to diversified needs has become an inevitable 

problem for enterprises. Compared with the traditional 

closed-source innovation model, the open innovation model 

led by the digital age enables enterprises to maintain the 

acceptance of new ideas and good ideas at any time and place 

through the form of "crowdsourcing of R&D functions" [3], 

so as to continuously develop and innovate, reduce the 

average production cost. 

As the general secretary said, the problem of regional 

economic gap in China has always existed, and from the 

existing literature and works, relevant scholars have explored 

from the perspective of digital economy development, but the 

results of studying the relationship between the two are 

relatively small. Based on the perspective of economic 

agglomeration, some scholars have found that with the 

development of the digital economy, the manufacturing 



 

54 

industry has further concentrated, but at this stage it is not 

conducive to narrowing the regional gap [4]. Based on the 

perspective of labor factor allocation, some scholars have 

found that the digital economy has made the regional 

economic gap show an inverted U-shaped change 

characteristic through technological spillover and economic 

spillover effect [5]. The possible marginal contributions of 

this paper are: (1) to study the regional economic gap from 

the perspective of digital economy, and to explore the 

relationship between digital economy development and 

regional economic gap; (2) Incorporate the digital economy 

into the core-edge model from the perspective of influencing 

"production costs" and "transaction costs", and explore the 

relationship between the digital economy and the regional 

economic gap under long-term equilibrium conditions with 

the help of computer simulation technology. The remaining 

parts of this paper are arranged as follows: The second part, 

based on the CP model that introduces the effect of spatial 

congestion, explores the relationship between the level of 

digital economy and the regional economic gap; Part III, 

Empirical Strategies and Variable Selection; Part IV, reporting 

on estimates based on spatial metrology; Part V further 

discusses the impact of spatial congestion on regional 

economic disparities arising from the digital economy; 

Finally, there are conclusions and policy recommendations.  

2. The Theoretical Analysis 
Framework: Logical Reasoning 
Based on Extended CP Model 

2.1. Model derivation 

2.1.1. Basic assumptions 

Construct a general spatial equilibrium model of two 

regions, two departments, two elements 2× 2× 1. The two 

regions are represented by r=1, 2 respectively; The two 

sectors refer to agricultural sectors A and B, respectively, 

which are capable of trade in products, and the two regions 

meet the conditions of agricultural non-specialization, and 

agricultural production activities meet the general 

equilibrium of Walras, producing homogeneous agricultural 

products, without regard to their transaction costs, that is, the 

price of the product is the same in different regions. With 

regard to the non-agricultural sector, it is discussed under the 

framework of D-S monopolistic competition, in which each 

firm produces only one unit of a certain differentiated product. 

Set the consumer's consumption substitution flexibility for the 

differentiated products they face as σ. One unit of agricultural 

production requires only one unit of agricultural labor, and the 

wage is set to 1, but for each unit of non-agricultural products 

produced, it requires a fixed input of F units and a marginal 

input of MP units, both of which are measured in terms of the 

number of laborers and the yield function 𝐿 = 𝐹 +𝑀𝑃 ×
𝑞𝐵(𝑖). With regard to transaction costs, it is assumed that the 

transport of products in the non-agricultural sector uses 

"iceberg transport costs", in 𝑇𝐵 the sense of ( 𝑇𝐵 > 1). 

Agricultural labour does not choose to move across regions, 

accounting for 1/2 of each, while non-agricultural labour can 

move freely across regions; Because the share of agricultural 

labor force in each region is fixed, only the living space of 

non-agricultural labor force is considered, and with the 

increase of population density in a certain area, resource and 

environmental problems are prominent, traffic congestion, 

high housing prices and land prices and high production and 

living costs Problems gradually emerged. Referring to Pi 

Yabin(2016) modeling ideas, non-farm labor is subject to 

space congestion costs (only land rent and transportation costs 

are considered), taking area 1 as an example, space 

congestion cost C depends on the congestion coefficient η and 

labor share𝑆𝐵, specifically 𝐶 = 𝜂 · 𝑆𝐵  / 2 [6]. The remaining 

settings are basically the same as the CP model, and the 

differences are given as they are analyzed.  

2.1.2. Short-term equilibrium 

2.1.2.1. Consumer Conduct 

The first layer of the consumer utility function, defined as 

the C-D type, is given by the invariant alternative elasticity 

(CES) utility function as follows: 

{
𝑈 = 𝑄𝐴

1−𝜇
𝑄𝐵
𝜇

  𝑌 = 𝑃𝐴𝑄𝐴 + 𝑃𝐵𝑄𝐵 + 𝐶 · 𝐼
            (1) 

Among them, Y is the income of consumers and all comes 

from wages, and there is no saving behavior; C refers to the 

cost of space congestion. 𝐼 =

{
1, Consumers are non − farm labour
 0, Consumers are agricultural labor

；𝜇  is the constant 

(0 < 𝜇 < 1),1 − 𝜇 and𝜇 represents the consumer's share of 

spending on agricultural and non-agricultural products, 

respectively; 𝑄𝐴  and the quantitative indicators that 

represent the combination of 𝑄𝐵  agricultural products and 

non-agricultural products consumed by consumers, 

respectively, and use the constant substitution elastic utility 

function to represent non-agricultural products, that is 𝑄𝐵 =

[∫ 𝑞𝐵(𝑖)
𝜎−1 𝜎⁄ 𝑑𝑖

𝑁𝐵
𝑖=0

]
𝜎 𝜎−1⁄

, 𝑞𝐵(𝑖) the first The consumption of 

i products, 𝜎 which represents the amount between any two 

finished products Substitution elasticity and greater than 1, 

𝑁𝐵 indicating the range of types of products in the region; 

The non-agricultural price index 𝑃𝐵 = [∫ 𝑝𝐵(𝑖)
𝜎−1 𝜎⁄ 𝑑𝑖

𝑁𝐵

𝑖=0
]
𝜎 𝜎−1⁄

 , 

𝑝𝐵(𝑖) which is the price of the ith product.  

According to the level of consumer utility maximization, 

the consumer demand function for agricultural and non-

agricultural products is obtained: 

𝑄𝐴 = (1 − 𝜇)𝑌/𝑃𝐴              (2) 

𝑄𝐵 = 𝜇𝑌/𝑃𝐵                   (3) 

𝑞𝐵(𝑖) = 𝜇𝑌(𝑝𝐵(𝑖)
−𝜎 /𝑃𝐵

1−𝜎)               (4) 

In addition, (2) and (3) are substituted into the consumer 

utility function, and the indirect utility function is obtained, 

and the area i complete price index is defined accordingly 

𝐺𝑖 = (𝑃𝐴
𝑖)
1−𝜇

(𝑃𝐵
𝑖 )
𝜇

.  

2.1.2.2. Producer behavior 

The digital economy will have an impact on the production 

costs of enterprises. Digital procurement is the most effective 

means to solve the pain and difficulties of high procurement 

costs, difficult to find high-quality suppliers, scattered 

procurement needs, and cumbersome manual operations, 

helping enterprises to reduce their comprehensive costs by 12% 

[7]. The advent of the digital economy era has alleviated 

information asymmetry, reduced the cost of labor matching 

for enterprises, and provided a platform that can reduce labor 

sharing and learning costs [8]. In summary, the development 

of the digital economy will reduce the production costs of 

enterprises to a certain extent, drawing on the modeling ideas 

of An Tongliang (2020) to define the production cost function 

𝐶𝑂𝑆𝑇 =
𝑤𝐵

𝜃
(𝐹 + 𝑀𝑃 · 𝑞𝐵(𝑖)) , which 𝑤𝐵 refers to the wage 

level of non-agricultural labor. The meaning of the formula is: 

the level of digital economy development (θ) increase, will 

reduce the original production costs of enterprises 𝑤𝐵(𝐹 +



 

55 

𝑀𝑃 · 𝑞𝐵(𝑖)), and assume that the cost reduction effect of the 

digital economy in enterprises in the two regions is 

comparable.  

According to the given cost function, the profit function of 

the products produced by the representative enterprises of the 

easy to obtain non-agricultural enterprises i: 

𝜋𝑖 = 𝑝𝑖𝑞𝐵(𝑖) −
𝑤𝐵

𝜃
(𝐹 + 𝑀𝑃 · 𝑞𝐵(𝑖))           (5)  

Combined with equation (4), according to the conditions of 

profit maximization, the product price function is obtained, 

and the product price is found to be irrelevant to the type, and 

the product type i can be omitted: 

𝑝𝐵 =
𝜎

𝜎−1

𝑤𝐵

𝜃
 𝑀𝑃                   (6) 

It should be noted that although we are discussing under 

the framework of D-S monopoly competition, enterprises are 

facing a market that belongs to a perfectly competitive market, 

and the pricing strategy of enterprises can be set as a bonus 

pricing method, and the formula (6) is brought into the 

formula (5) to obtain the output of the enterprise𝑞𝐵 =
𝐹(𝜎−1)

𝑀𝑃
. 

It was also found to be independent of the species. Further 

bringing the output of enterprises into the production function, 

the number of enterprises is obtained 𝑁𝐵 =
𝐿𝐵

𝐹𝜎
, and there are 

two regional shares of enterprises engaged in non-agricultural 

production in zone 1, 𝜆 = 𝑆𝐵, 𝑆𝐵referring to zone 1 Share of 

the total non-farm labour force.  

2.1.2.3. Wage level determination 

For the following analysis to be simple, some 

standardization is required before subsequent derivation. First 

determine the market size of each region, the total expenditure 

of easy-to-get areas 1 and 2 is: 

{
𝐸1 =

1

2
(1 − 𝜇) + 𝜆𝜇(𝑤𝐵

1 − 𝜂𝑆𝐵/2) 

𝐸2 =
1

2
(1 − 𝜇) + (1 − 𝜆) 𝜇[𝑤𝐵

2 − 𝜂(1 − 𝑆𝐵)/2]
   (7) 

It should be clear that digital technology can effectively 

reduce the external transaction costs of enterprises, on the one 

hand, the reduction of search costs, the reduction of costs such 

as negotiation and negotiation between enterprises, and the 

flexibility of transaction details can avoid unpredictable costs 

[9]. On the other hand, with the digital economy, the 

emergence of new logistics models, the rapid development of 

the express delivery industry, and the transportation cost 

curve will gradually decline [2]. To this end, the introduction 

of digital technology into the transaction level, set𝛿 to the 

level of digital economy development at the transaction cost 

level, the range of "iceberg transportation costs" is
𝑇𝐵

𝛿
  , 

and1 < 𝛿 < 𝑇𝐵.  

According to formula (6), combined with the premise of 

transportation costs, for the non-agricultural sector, the selling 

prices of enterprises in areas 1 and 2 of zone 1 are: 

respectively 𝑝𝐵
11 =

𝑤𝐵
1

𝜃
 , 𝑝𝐵

12 =
𝑇𝐵

𝛿

𝑤𝐵
1

𝜃
 . Similarly, it can be 

known that the selling prices of products of enterprises in 

Area 1 and 2 are as follows𝑝𝐵
21 =

𝑇𝐵

𝛿

𝑤𝐵
2

𝜃
 , 𝑝𝐵

22 =
𝑤𝐵
2

𝜃
: Where 

𝜃 the level of digital economy development at the level of 

production costs is met θ ≥ 1 ; 𝛿  is the level of digital 

economy development at the level of transaction costs in the 

range. Based on the above analysis, the price function of non-

agricultural sector enterprises is easy to obtain. At the same 

time, according to the definition of the price index, the 

tradable non-agricultural products price index of each region 

is obtained: 

{
 
 

 
 
𝑃𝐵
1 = [𝜆 (

𝑤𝐵
1

𝜃
)
1−𝜎

+ (1 − 𝜆) (
𝑇𝐵𝑤𝐵

2

𝛿𝜃
)
1−𝜎

]

1
1−𝜎⁄

𝑃𝐵
2 = [𝜆 (

𝑇𝐵𝑤𝐵
1

𝛿𝜃
)
1−𝜎

+ (1 − 𝜆) (
𝑤𝐵
2

𝜃
)
1−𝜎

]

1
1−𝜎⁄

   (8) 

In order to meet the market clearance conditions for 

tradable non-agricultural products, all the output of the 

enterprise needs to be sold out in two regions, so the sales 

volume 𝛬𝐵
1 = 𝑝𝐵𝑞𝐵 = 𝑝𝐵

11𝑞𝐵
11 + 𝑝𝐵

12𝑞𝐵
12 , wherein the 

𝑞𝐵
11, 𝑞𝐵

12 non-agricultural enterprises located in Zone 1, are 

located in Zone 1 and Zone 2, respectively. Combined 

formula (4), formula (6) ~ (8), the price of the product in each 

region 𝑝𝐵
11, 𝑝𝐵

12  and standardized conditions, Sales of 

enterprises in Available Zone 1: 

𝛬𝐵
1 = 𝜇 (

𝑤𝐵
1

𝜃
)
1−𝜎

[𝐸1(𝑃𝐵
1)𝜎−1 + (

𝑇𝐵

𝛿
)
1−𝜎

𝐸2(𝑃𝐵
2)𝜎−1]   (9) 

Similarly, sales of Region 2 companies: 

𝛬𝐵
2 = 𝜇 (

𝑤𝐵
2

𝜃
)
1−𝜎

[(
𝑇𝐵

𝛿
)
1−𝜎

𝐸1(𝑃𝐵
1)𝜎−1 + 𝐸2(𝑃𝐵

2)𝜎−1]  (10) 

Suppose that the enterprise produces under the condition of 

zero profit, so the income of the enterprise is the remuneration 

of the labor force, that is 𝛬𝐵 = 𝑤𝐵𝐿𝐵 . In summary, get the 

expression of wage levels for non-farm labor in zones 1 and 

2: 

{
 
 

 
 
𝑤𝐵
1 = (

1

𝜃
)

1−𝜎

𝜎
[𝐸1(𝑃𝐵

1)𝜎−1 + (
𝑇𝐵

𝛿
)
1−𝜎

𝐸2(𝑃𝐵
2)𝜎−1]

1

𝜎

𝑤𝐵
2 = (

1

𝜃
)

1−𝜎

𝜎
[(
𝑇𝐵

𝛿
)
1−𝜎

𝐸1(𝑃𝐵
1)𝜎−1 + 𝐸2(𝑃𝐵

2)𝜎−1]

1

𝜎

   (11) 

2.1.3. Long-term equilibrium 

Through short-term equilibrium analysis, the wage level of 

labor under static conditions can be determined. According to 

the previous deduction, the spatial distribution of satisfying 

labor is equivalent to the spatial distribution of enterprises, so 

under the influence of the digital economy, the spatial 

distribution of economic activities𝜆is the key to identifying 

long-term steady-state equilibrium. Assuming that 

unemployment does not exist within each region, labour 

mobility decisions are based on the level of utility actually 

obtained. Thus, the equation for the spatial mobility of labor 

can be expressed as: 

�̇� = 𝜆(1 − 𝜆)(
𝛺1

𝛺2
− 1)            (12) 

Among 𝛺1 =
𝑤𝐵
1−𝜂𝑆𝐵/2

𝐺1
, 𝛺2 =

𝑤𝐵
2−𝜂(1−𝑆𝐵)/2

𝐺2
  them, the 

actual level of utility obtained by employment labor in the 

non-farm sector in Regions 1 and 2, respectively; 𝐺1 , 𝐺2 

respectively, the two regions face the full price index for 

residents 𝐺𝑖 = (𝑃𝐴
𝑖)
1−𝜇

(𝑃𝐵
𝑖 )
𝜇
  (𝑖 = 1, 2).  

2.1.4. Calculation of regional output per capita 

The model assumes that tradable products have been 

cleared from the market, and that the output of the sector is 

converted into household consumption expenditure, except 

for part of the cost of space congestion. According to Formula 

(7), combined with standardized conditions, the total output 

of residents in area 1 is obtained (total expenditure + cost of 

space congestion): 

𝐺𝐷𝑃1 =
1

2
(1 − 𝜇) + 𝜆𝜇𝑤𝐵

1          (13) 

The total population of region 1 is 
1

2
(1 − 𝜇) + 𝜆𝜇. Based 

on formula (13), the full price index and the total population 

of the region, the real output per capita of the region 1 is 

obtained: 

𝑃𝐶𝐺1 =
1

2
(1−𝜇)+𝜆𝜇𝑤𝐵

1

1

2
(1−𝜇)+𝜆𝜇

1

𝐺1
          (14) 



 

56 

Similarly, the 2 per capita output of the Easy Know region: 

𝑃𝐶𝐺2 =
1

2
(1−𝜇)+(1−𝜆)𝜇𝑤𝐵

2

1

2
(1−𝜇)+(1−𝜆) 𝜇

1

𝐺2
           (15) 

At this point, it is not difficult to find that 

𝑃𝐶𝐺1
𝑃𝐶𝐺2
⁄  regional economic disparities can be measured. 

The indicator is close to 1, which can be considered to be a 

small regional economic gap; 𝑃𝐶𝐺
1

𝑃𝐶𝐺2
⁄  away from 1, the 

greater the regional gap.  

2.2. Numerical simulation 

In view of the difficulty of examining the analytical form 

of the model under the long-term equilibrium situation, the 

numerical simulation is carried out with the help of Matlab 

software. The definition describes the spatial distribution 

curve of economic activity as curve A, and the curve of 

describing regional economic gap as curve B. The vertical 

axis of the image represents the actual utility gap Ω1 Ω2⁄  and 

the regional economic gap 𝑃𝐶𝐺1 𝑃𝐶𝐺2⁄ , and the horizontal 

axis represents the spatial distribution status of enterprises (𝜆). 

The two forces that affect the equilibrium state of enterprise 

site selection are: the agglomeration force based on the 

market proximity effect and the cost of living effect; 

Dispersive forces based on local competitive effects and 

space congestion costs. Digital technology has an impact on 

the system of two forces, and through the endogenous 

mechanism of enterprise transfer (labor mobility), the internal 

two forces are re-balanced, affecting the spatial distribution 

of enterprises. The process of reaching homeostasis again by 

externally influenced systems is more complicated, and we 

only focus on two signals for enterprise transfer to reach 

steady state: (1) The internal equilibrium solution of 

incomplete agglomeration:
Ω1

Ω2
⁄ = 1 satisfied 

𝜕(Ω1/Ω2)

𝜕𝜆
< 0; 

(2) Corner point solution:
Ω1

Ω2
⁄ = 1 satisfied at the place 

𝜕(Ω1/Ω2)

𝜕𝜆
>0.  

First, the common state parameter is set μ=0.4, σ=5.0, 𝜂 =
0.1 , 𝑇𝐵 =1.8; Then, assuming that the digital economies of 

the two regions develop simultaneously, the digital 

technology levels of regions 1 and 2 can be assigned, 

respectively: the lower level (𝛿 ,θ) = (1.1, 1.1); Moderate 

level (𝛿,θ) =(1.4, 1.4); Higher levels (𝛿,θ) =(1.7, 1.7). It will 

be traversed𝜆from 0 to 1 to simulate that under the influence 

of different digital technology levels, the transfer of 

enterprises in area 1 is in a spatially balanced distribution state 

𝜆.  

 
Figure 1. Lower level 

 
Figure 2. Medium level 

 
Figure 3. Higher level 

From the simulation results of Figure 1.1, it can be seen 

that digital technology has just started, and it 
𝜕(Ω1/Ω2)

𝜕𝜆
|
𝜆=𝜆1

<

0 is satisfied that the𝜆1spatial distribution is in a steady state, 

the symmetrical equilibrium is still stable, and the enterprise 

site selection tends to be symmetrical. Observing curve B, it 

is found that the regional economic gap 𝑃𝐶𝐺1 𝑃𝐶𝐺2⁄ tends to 

1, which can be considered that the initial stage of digital 

technology has not had a significant impact on the regional 

gap; As shown in Figure 1.2, with the gradual development of 

digital technology in zone 1 to the medium 

level
𝜕(Ω1/Ω2)

𝜕𝜆
|
𝜆=𝜆2

> 0 , the symmetrical structure has 

obviously become unstable, and there is a tendency for non-

agricultural enterprises to concentrate in a certain area. 

Looking at curve B again, the 𝜆2disturbance of the shift of 

enterprises to the north is applied to the point, and it is found 

that 𝑃𝐶𝐺1 𝑃𝐶𝐺2⁄  gradually moving away from 1, it can be 

considered that the regional gap may be widened in the 

medium term of the development of digital technology; 

Figure 1.3 illustrates the influence of higher digital 

technology levels,
𝜕(Ω1/Ω2)

𝜕𝜆
|
𝜆=𝜆3

< 0symmetrical equilibrium 

from unstable to stable, and 𝑃𝐶𝐺1 𝑃𝐶𝐺2⁄ tends to 1, it can be 

said that with the high development of digital technology, 

regional disparities show a narrowing trend.  

Proposition: In the real world, the relationship between the 

digital economy and the regional economic gap is non-linear, 

that is, with the continuous development of the digital 

economy, the regional economic gap shows an inverted U-

shaped trend of first expanding and then narrowing.  

λ

1 



 

57 

3. Empirical Strategies and Variable 
Selection 

3.1. Empirical strategies 

Numerical simulation results show that the impact of 

digital economy on regional economic gap may show an 

inverted U-shaped relationship, that is, in the early stage of 

digital technology development, with the development of 

digital technology, the regional economic gap is significantly 

widened. However, once the level of digital technology 

crosses a certain threshold, the economic gap between regions 

will tend to narrow. In order to study such non-monotonicity, 

Xu Lifang's (2017) approach was drawn to add the square 

term of the digital economic development level to the 

regression model [10]. Therefore, consider the double fixed 

effect traditional panel regression model as follows. When 

𝛼1 > 0 , 𝛼2 < 0 , it can be preliminarily determined that the 

gap between the digital economy and the regional economy is 

inverted U-shaped.  

GAPi;t = 𝛼0 + 𝛼1DEIi;t + 𝛼2DEIi;t
2 + α3Controli;t + μ

i
+ 𝜅t + εi;t    (16) 

Among them, GAPi;t  the level of regional economic gap 

representing the t-time region i; DEIi;t represents the level of 

digital economy development of the t-time region i; DEIi;t
2  

represents its square term; Controli;t is a set of control 

variables; μi refers to the fixed effect of each provincial 

region; 𝜅t represents the fixed effect of the year; εi;𝑡 is a 

random perturbation term.  

Regional economic gap is a dynamic concept, not only with 

the region's own conditions change, but also affected by 

external conditions, in addition, the economic and social 

relations between neighboring regions are objectively 

existing, the impact on the local regional economic gap 

cannot be ignored, the introduction of spatial effects into the 

empirical model can maximize the persuasiveness of model 

conclusions. To this end, the spatial panel model is established 

as follows to verify the inverted U-shaped relationship 

between the digital economy and the regional economic gap.  

Moran' I index and local Moran scatter plot are used to test 

whether the spatial factors introduced into the panel model are 

reasonable. Further, according to the LM test, Robust-LM test, 

LR test and Robust-LR test, a spatial panel model suitable for 

the object of study in this paper can be selected. According to 

the impact mode of the space term, the model is as follows: 
GAPi;t = β0 + rho ∙ W

G ∙ GAPi;t + β1DEIi;t + β2DEIi;t
2

+ β3Controli;t + μi + 𝜅t + εi;t 

εi;t = λW
Gεi;t + ξi;t            (17) 

Where, WG for the spatial geographic weights matrix; 

rho is the spatial hysteresis coefficient; λ is the number of 

autoregressive systems for spatial error. When𝜆 = 0 , 𝑟ℎ𝑜 ≠
0  the spatial panel model is a spatial lag model (SAR); When 

𝜆 ≠ 0 , 𝑟ℎ𝑜 = 0 , the spatial panel model is a spatial error 

model (SEM).  

For the spatial econometric model constructed above, since 

the regional economic gap is introduced as an explanatory 

variable as a first-order lag term, problems related to the 

random error term may arise, and it is particularly important 

to choose an appropriate estimation method. In this paper, the 

maximum likelihood estimation method MLE suitable for 

spatial panel models is used to estimate the parameters of the 

model. It is worth noting that some factors and variable 

measurement errors missing from the model may seriously 

affect the ability of the digital economy level to interpret the 

interpreted variables, for this reason, all metrology models in 

this paper use robust standard errors to eliminate the influence 

of heteroscedasticity on coefficient estimation.  

3.2. Variable selection 

3.2.1. Interpreted variables 

Regional Economic Gap (GAP). How to describe regional 

economic disparities is a thorny issue, and some studies have 

used economic indicators such as per capita GDP and per 

capita GDP growth rate to measure inter-regional economic 

disparities, but I believe that this approach does not reflect the 

disparities in the model. This paper synthesizes the 

measurement method of Lu Hongyou (2012) [11] and 

BianYuanchao (2018) [12] to measure regional gap, and 

selects the logarithm of per capita GDP in each provincial-

level region the difference from the logarithm of gdp per 

capita in the country. The time span of per capita GDP is from 

2011 to 2019, and the per capita GDP in all years is based on 

2010, according to the per capita of each provincial region 

The GDP Reduction Index is smoothed out to calculate 

comparable variable values for each year.  

3.2.2. Core explanatory variables 

Table 1. Measurement index system for the level of 

digital economy development 

Level 1 

indicators 

Secondary 

indicators 

Metric 
measurement 

method 

Attribute 

Digital 
economy 

Development 

Index 

Internet 

penetration 

Number of 
Internet broadband 

access users per 100 

people 

+ 

Number of 

Internet-

related 
practitioners 

Computer service 

and software 

employees account 

for the proportion of 
employed personnel 

in urban units (%) 

+ 

Internet-
related 

outputs 

Total 
telecommunications 

services per capita 

+ 

Number of 

mobile 
Internet 

users 

Number of 

mobile phone 
subscribers per 100 

people (%) 

+ 

Inclusive 
development 

of digital 

finance 

China Digital 

Financial Inclusion 
Index 

+ 

Digital Economy Development Index (DEI). For the sake 

of data availability, drawing on the practice of Zhao Tao (2020) 

[13], the five indicator sets of the digital economy 

development level of different provinces (2011-2019) are 

used Take measurements. The index is based on Internet 

penetration, the number of Internet-related practitioners, 

Internet-related output, the number of mobile Internet users, 

and the inclusive development of digital finance, and the 

specific measurement methods are shown in Table 1. For the 

dimensionality reduction treatment of the data of 5 indicators, 

we use the method of principal component analysis. The 

indicator covers a wide range of socio-economic content, 

captures the industry penetration of digital technology, and 

more accurately conforms to the definition of the digital 

economy in the economic community.  

3.2.3. Control variables 

In order to prevent the influence of other factors on the 

empirical results of this paper, with reference to the relevant 

domestic literature, the following control variables are 

selected, and the time span is 2011-2019, including: (1) the 



 

58 

development level of the tertiary industry (TIS). Measured by 

tertiary industry output value/regional GDP of each region; (2) 

Human capital level (SH). Human capital is a decisive factor 

in the region's long-term economic growth, plays an 

important role in a sustainable economy, and is also a non-

negligible part of the regional economic gap. There are many 

ways to measure the stock of human capital, and this paper 

proposes to use the average number of years of education to 

measure, that is, SH=6 a1 +9 a2 +12 a3 +16 a4 , 

a1, a2, a3, a4 which indicates that the number of primary 

school culture, primary Chinese, high Chinese and college 

education in each provincial-level area accounts for the 

proportion of the permanent population; (3) Level of public 

expenditure (GY). The level of government public 

expenditure measures the government's contribution to the 

people's livelihood such as education and health, and is 

measured in terms of regional government general public 

budget expenditure/regional GDP; (4) Level of opening to the 

outside world (FE). The intensity of foreign trade dependence 

is an important source of economic growth, which is 

expressed by the import and export trade volume (RMB) of 

each region as the regional GDP. (5) Foreign Investment 

Level (TI). The year-end registered foreign investment 

amount (RMB) / regional GDP measurement of each region 

is selected.  

3.3. Data sources and descriptive statistics 

Table 2. Descriptive statistics for variables 

Variable 
Symbo

l 

Observa

tions 

Me

an 
SD 

Mi

n 

Ma

x 

Regionaleconomi

cdisparities 
GAP 270 

-

0.016 

0.3

91 

-

0.808 

0.9

11 

Digital economy DEI 270 
0.2

90 

0.1

66 

0.0

16 

0.9

11 

Tertiary industry 

development 
TIS 270 

0.4

86 

0.0

90 

0.3

27 

0.8

37 

Human capital SH 270 
0.1

88 

0.0

99 

0.0

80 

0.6

22 

Level of public 

spending 
GY 270 

1.2

53 

0.5

84 

0.4

53 

3.4

68 

The level of 

opening up to the 

outside world 

FE 270 
0.2

79 

0.2

97 

0.0

13 

1.4

64 

Levelofforeignin

vestment 
TI 270 

0.3

84 

0.3

65 

0.0

48 

1.7

35 

This paper uses annual data modeling to collect panel data 

from 2011 to 2019 from 30 provincial-level administrative 

units in China (due to the lack of some data, not considering 

Tibet, Hong Kong, Macao and Taiwan). Statistical Yearbook", 

"Statistical Yearbook of Chinese and Employment", "Peking 

University Digital Inclusive Finance Index" and "China Trade 

and Foreign Economic Statistics Yearbook". According to the 

search data, it is possible to explore the role of digital 

economy, economy and other factors on regional economic 

disparities. The descriptive statistics of variables are shown in 

Table 2.  

4. Empirical Analysis: Verification 
based on spatial metrology 

4.1. Spatial autocorrelation test 

To determine whether spatial factors are introduced, a 

spatial autocorrelation test was performed on the level of 

economic difference in each region using the global Moran' I 

test based on the cross-sectional data perspective, and the 

results are reported in Table 3. The results show that we are 

99% sure that the level of regional economic gap is 

contiguous and agglomeration at the spatial level, which 

verifies that there is a significant spatial dependence of 

regional economic gap, and it is necessary to adopt spatial 

panel regression methods for the subsequent inverted U-

shaped impact of the digital economy on the regional 

economic gap.  

Table 3. Global Moran test results 

 Geographical distance  

year Moran 

value 

Z 

value 

year Moran 

value 

Z 

value 

2011 0.377*** 4.336 2016 0.362*** 4.160 

2012 0.376*** 4.318 2017 0.361*** 4.153 

2013 0.371*** 4.262 2018 0.361*** 4.158 

2014 0.365*** 4.202 2019 0.363*** 4.187 

2015 0.362*** 4.162    

Note: "***, **, *" indicates that it has passed the significance test at the 

levels of 1%, 5%, and 10%, respectively. 

4.2. Benchmark regression 

The two types of non-observed effects of province and year 

effects are controlled differently, and the traditional panel 

model of double fixed effects is used, and the regression 

results are shown in the columns of table 4 (1), (2). Further, 

regional economic disparities exist spatially dependent, while 

L M tests, Robust LM tests also show the rationality of using 

spatial panel models, and therefore, in tables Columns (3)~(6) 

of 4 list the estimated results of the spatial lag model and the 

spatial error model. 

The empirical results show that compared with the 

traditional panel model under the same fixed effect, the level 

coefficient of digital economy development is at least 10% 

significant, and the square term estimation coefficient is 

significant at the 1% level, which confirms the proposition 

"nonlinear inverted U of the gap between the digital economy 

and the regional economy." type relationship". Compared 

with the spatial panel model under the same fixed effect, the 

coefficients of the primary and secondary terms of the digital 

economy development level are at least at least 10%, and the 

coefficient symbol satisfies the inverted U-shape condition. It 

can be considered that the inverted U-shaped relationship 

between the digital economy and the regional economic gap 

is still established under the consideration of spatial factors. 

The reason is that the gradual improvement of the level of 

digital economy development in a certain region may lead to 

the improvement of the level of enterprise industrial 

agglomeration, entering a virtuous circle of "agglomeration-

optimization-re-agglomeration", which is in stark contrast to 

the vicious cycle of "loss-deterioration-re-loss" in peripheral 

areas, causing regional economic gaps; When digital 

technology is further developed, the agglomeration power of 

enterprises in the region becomes weaker, and the effect of 

spatial congestion gradually appears, which, together with the 

competition effect of local market in the region, jointly affects 

the location choice of enterprises and causes regional gaps to 

converge. The spatial term coefficient rho value of the SAR 

model is significantly positive at least 1%, indicating that 

there is a positive spillover effect on the regional economic 

gap in China. The spatial term coefficient λ of the SEM model 

is significantly positive at least at least 1%, indicating that 

there is an omission in the selection of control variables in this 

paper, and the spatial autocorrelation may be achieved by the 

error term, although the digital economy is inverted to the 

regional gap. The type relationship is still significant, 

revealing the correctness of the proposition.  

 

 



 

59 

Table 4. Benchmark regression results 

variable 
(1) (2) (3) (4) (5) (6) 

FE SAR SEM 

Main       

DEI 

0.528**

* 

0.773**

* 

0.324**

* 
0.452** 

0.468**

* 
0.554** 

(0.116) (0.250) (0.102) (0.217) (0.139) (0.238) 

DEI^2 

-

0.496**

* 

-

0.442**

* 

-

0.316**

* 

-0.328** 

-

0.443**

* 

-

0.449**

* 

(0.108) (0.155) (0.099) (0.151) (0.136) (0.159) 

TIS 
-0.424** -0.407* -0.386** -0.429** 

-

0.422**

* 

-

0.416**

* 

(0.194) (0.215) (0.162) (0.169) (0.149) (0.150) 

SH 
-0.387** 

-

0.519**

* 

-0.299** 

-

0.388**

* 

-0.319** -0.340** 

(0.151) (0.174) (0.123) (0.142) (0.142) (0.143) 

GY 

0.087**

* 

0.081**

* 

0.079**

* 

0.075**

* 

0.084**

* 

0.085**

* 

(0.018) (0.026) (0.018) (0.024) (0.016) (0.021) 

FE 
0.090 0.132** 0.109** 0.116* 0.111* 0.120** 

(0.054) (0.058) (0.055) (0.060) (0.057) (0.058) 

TI 

-

0.066**

* 

-0.054** 

-

0.051**

* 

-0.052** -0.048** -0.046** 

(0.024) (0.025) (0.020) (0.021) (0.020) (0.020) 

Spatial       

rho 
  

0.588**

* 

0.541**

* 
  

  (0.079) (0.090)   

λ 
    

0.702**

* 

0.637**

* 

    (0.074) (0.089) 

Sample 

size 
270 270 270 270 270 270 

The 

provinc

e is 

fixed 

Yes Yes Yes Yes Yes Yes 

The 

year is 

fixed 

No Yes No Yes No Yes 

Log-L — — 672.9 681.9 670.2 143.4 

R2 0.604 0.656 0.558 0.513 0.601 0.557 

Note: "***, **, *" indicates that it has passed the significance test at the 
level of 1%, 5% and 10% respectively; Robust standard errors are in 

parentheses. The same applies below. 

Regarding the control variables, under the regional fixed 

and double fixed models, increasing the proportion of the 

tertiary industry in regional GDP and increasing the level of 

regional human capital can significantly narrow the regional 

economic gap, because the industrial structure and human 

capital are important links between economic activities, and 

play an important role in optimizing the allocation of factor 

resources, improving technological efficiency, and enhancing 

the ability to match modern industries, thereby curbing the 

trend of widening regional economic gaps; Improving the 

level of foreign investment has to a certain extent made up for 

the shortage of funds in development, brought a large number 

of jobs and taxes, improved the level of regional economic 

development, and narrowed the regional economic gap. 

Government public spending and opening up will 

significantly widen regional economic disparities and will be 

analyzed in detail when the sample returns later. 

4.3. Robustness testing 

In order to verify the reliability of the inverted U-shape, 

this paper replaces the measure of regional economic gap 

from the perspective of variables, and selects the difference 

between the logarithm of per capita disposable income of 

residents in each provincial-level region and the national level. 

Columns 5 (13) to (15) of Table 5 report the estimation results 

of replacing the explanatory variables, and the primary term 

coefficients of the digital economy are significantly positive 

and the secondary term coefficients are significantly negative, 

which satisfies the conditions of the inverted U-shaped 

relationship, which shows that the estimation results are 

basically consistent with the above conclusions. Next, the 

economic distance weight matrix is replaced, and the inverted 

U-shape is further tested using the spatial lag double fixed 

effect model, columns (16), (17). Estimates have been 

reported and remain consistent with the conclusions above. It 

is worth noting that the coefficient of spatial lag terms under 

the economic weight’s matrix does not correspond to the 

preceding paragraph, but does not affect the conclusion of this 

article, because the rho value does not accurately refer to the 

spatial spillover effect of the digital economy on economic 

disparities.  

Table 5. Robustness test results 

variabl

e 

Replace the interpreted variable 
Replace the spatial 

weight matrix 

(13) (14) (15) (16) (17) 

FE SAR SAR WE WE 

Main      

DEI 

0.717*

** 

0.388*

** 

0.406*

** 

0.927*

** 

0.704*

** 

(0.140) (0.150) (0.115) (0.293) (0.205) 

DEI^2 

-

0.292*** 

-

0.181** 

-

0.168** 

-

0.600*** 

-

0.452*** 

(0.059

5) 

(0.091

5) 

(0.071

9) 
(0.151) (0.141) 

Control 

variables 

Contro

lled 

Contro

lled 

Contro

lled 

Contro

lled 

Contro

lled 

Spatial      

rho 

 
0.749*

** 

0.677*

** 

-

0.492*** 

-

0.666*** 

 
(0.084

7) 
(0.115) (0.133) (0.120) 

Sample 
size 

270 270 270 270 270 

Fixed 

province/

year 

Y/Y Y/Y Y/Y Y/Y Y/Y 

Log-L  801.9 849.1 604.3 678.3 

R2 0.538 0.013 0.025 0.325 0.600 

4.4. Regional heterogeneity analysis 

The previous article clarified that the regional economic 

gap shows a non-linear trend of first expanding and then 

narrowing with the development of the digital economy. In 

order to ensure the reliability of the estimated results and 

consider the influence of regional heterogeneity on the 

invertedU-shaped relationship, this paper will "re-test" the 

inverted U-shaped trendfrom the sample segmentation 

method. Table 6 is the result of the spatial SAR estimation of 

the three major regions of eastern, central and western China.  

The results show that with regard to the eastern sample, 

whether it is a fixed effect in the controlled area, an annual 

fixed effect, or a double fixed effect, the inverted U-shaped 

effect of the digital economy on the regional economic gap is 

at least significant at the level of 10%. In the central sample, 

under the fixed effect and double fixed effect of the control 

area, the primary and secondary term coefficients of the 

digital economy development level were at least significant at 

the level of 5%, while after the fixed effect of the control time, 

the secondary term of the digital economy level was not 

significant; Regarding the western sample, from a statistical 

point of view, the inverted U-shaped relationship between the 

fixed effect of the controlled area and the fixed effect of time 

is more significant, and the coefficient symbol is consistent 

with the regression result of the whole sample, which verifies 

the inverted U-shaped relationship between the digital 

economy and the regional economic gap. The inverted U-

shaped relationship under the double fixed effect is not 



 

60 

significant. The reason for the above estimates is that, on the 

one hand, the development of the digital economy in various 

regions is unbalanced and uncoordinated, and there is a 

"digital divide" between regions; On the other hand, some 

underdeveloped areas may be constrained by multiple factors 

such as geographical conditions, factor endowments, 

industrial supporting capabilities, and technical levels, and it 

is not easy to form a strong market proximity effect and cost 

of living effect, resulting in difficulties in cultivating local 

leading industries, weak endogenous development 

capabilities, and gradual economic development gaps Be 

pulled. Regarding control variables, changes in the level of 

government public expenditure in different regions will have 

a heterogeneous effect. The increase in the level of 

government public expenditure in the eastern and central 

regions will widen the regional economic gap; The increase 

in the level of expenditure in the western region will 

significantly narrow the regional economic gap. Therefore, as 

a fiscal policy that coordinates horizontal imbalances between 

regions, central transfer payments play an important role in 

narrowing the gap in regional economic development. 

Therefore, it is necessary to continue to deepen the 

cooperation and joint construction between the eastern and 

central and western regions, promote the development of 

open and high-quality development along the border, promote 

the acceleration of the formation of an open pattern of land 

and sea linkage, and mutual assistance between the east and 

the west, so as to alleviate the regional economic gap.  

 

Table 6. Fixed effect regression results of sub-sample 

space panel based on SAR model 

variable 
Eastern region Central Region Western region 

(7) (8) (9) (10) (11) (12) 

Main       

DEI 

0.346**

* 
0.551** 0.479** 

1.710**

* 

0.919**

* 
0.558 

(0.119) (0.238) (0.225) (0.504) (0.241) (0.458) 

DEI^2 

-

0.306**

* 

-

0.377**

* 

-

0.568**

* 

-

1.664**

* 

-

0.957**

* 

-0.276 

(0.0865) (0.120) (0.218) (0.620) (0.230) (0.487) 

Control 

variables 

Controll

ed 

Controll

ed 

Controll

ed 

Controll

ed 

Controll

ed 

Controll

ed 

Spatial       

rho 

-

0.359**

* 

-

0.461**

* 

0.456**

* 
0.348** 0.326** 0.146 

(0.0753) (0.0729) (0.124) (0.157) (0.161) (0.217) 

Sample 

size 
99 99 81 81 90 90 

Fixed 

province/y

ear 

Y/N Y/Y Y/N Y/Y Y/N Y/Y 

Log-L 262.0 286.5 212.3 238.3 220.0 229.9 

R2 0.427 0.257 0.686 0.471 0.744 0.002 

The results show that with regard to the eastern sample, 

whether it is a fixed effect in the controlled area, an annual 

fixed effect, or a double fixed effect, the inverted U-shaped 

effect of the digital economy on the regional economic gap is 

at least significant at the level of 10%. In the central sample, 

under the fixed effect and double fixed effect of the control 

area, the primary and secondary term coefficients of the 

digital economy development level were at least significant at 

the level of 5%, while after the fixed effect of the control time, 

the secondary term of the digital economy level was not 

significant; Regarding the western sample, from a statistical 

point of view, the inverted U-shaped relationship between the 

fixed effect of the controlled area and the fixed effect of time 

is more significant, and the coefficient symbol is consistent 

with the regression result of the whole sample, which verifies 

the inverted U-shaped relationship between the digital 

economy and the regional economic gap. The inverted U-

shaped relationship under the double fixed effect is not 

significant. The reason for the above estimates is that, on the 

one hand, the development of the digital economy in various 

regions is unbalanced and uncoordinated, and there is a 

"digital divide" between regions; On the other hand, some 

underdeveloped areas may be constrained by multiple factors 

such as geographical conditions, factor endowments, 

industrial supporting capabilities, and technical levels, and it 

is not easy to form a strong market proximity effect and cost 

of living effect, resulting in difficulties in cultivating local 

leading industries,  weak endogenous development 

capabilities, and gradual economic development gaps Be 

pulled. Regarding control variables, changes in the level of 

government public expenditure in different regions will have 

a heterogeneous effect. The increase in the level of 

government public expenditure in the eastern and central 

regions will widen the regional economic gap; The increase 

in the level of expenditure in the western region will 

significantly narrow the regional economic gap. Therefore, as 

a fiscal policy that coordinates horizontal imbalances between 

regions, central transfer payments play an important role in 

narrowing the gap in regional economic development. 

Therefore, it is necessary to continue to deepen the 

cooperation and joint construction between the eastern and 

central and western regions, promote the development of 

open and high-quality development along the border, promote 

the acceleration of the formation of an open pattern of land 

and sea linkage, and mutual assistance between the east and 

the west, so as to alleviate the regional economic gap.  

5. Conclusions and recommendations 

5.1. Conclusions of the study 

This paper attempts to explore the trend of the digital 

economy affecting the changes in regional economic gaps 

from the perspective of new economic geography. The main 

conclusion is that :(1) The digital economy profoundly affects 

the behavior of consumers and producers. Overall, with the 

continuous development of the digital economy, the regional 

economic gap shows an inverted U-shaped trend of first 

expanding and then narrowing. From the perspective of 

heterogeneous regions, the inverted U-shaped relationship 

between the level of digital economy development in the 

eastern, central and western regions and the regional 

economic gap is still significant. (2) In the low-level stage of 

the digital economy, the effect of the digital economy on the 

regional economic gap is a positive effect, and a higher degree 

of congestion will promote this positive effect. At the high 

level of the digital economy, the digital economy rises to a 

negative effect on regional economic disparities, and space 

congestion will strengthen this negative effect. (3) Extremely 

developed areas such as Beijing and Shanghai benefit from 

the digital economy, and higher levels of agglomeration may 

occur, opening up the economic gap with other regions.  

5.2. Policy recommendations 

This article does not negate agglomeration, but agrees with 

the fact that agglomeration of economic factors improves 

production efficiency. However, the author is more concerned 

about the subsequent development of areas with weak 

agglomeration in the context of the digital economy, or what 



 

61 

impact the development of the digital economy will have on 

the changes in the regional economic gap. Therefore, while 

promoting higher levels of agglomeration in developed areas, 

it is even more necessary to find a way out for underdeveloped 

areas, especially in declining areas. To this end, this paper 

proposes: (1) The digital economy provides a solution to 

alleviate the imbalance of regional development. 

Continuously become bigger, stronger and better china's 

digital economy, promote the deep integration of digital 

technology and enterprises, and give play to the "picking 

beams" function of the digital economy in stable growth and 

development. Taking this important opportunity to increase 

investment in underdeveloped areas in the form of "new 

infrastructure", so that underdeveloped areas can catch the 

"digital economy +" free ride Explore unique advantageous 

endowments from the local area, such as geographical 

comparative advantages, natural resource endowments, 

cultural conditions, good market competition environment 

and high-quality local government services, etc., seize the 

historic opportunity of industrial transfer in the eastern region, 

cultivate regional leading industries, expand local market 

potential, cross the inflection point of the inverting U,, and 

narrow the regional economic gap. (2) Highly developed 

areas should rely on digital technology to digitize and 

network the spatial behavior of production and life, connect 

factor resources, improve the business environment of 

enterprises, optimize the allocation of resources, land and 

other elements, alleviate environmental and traffic pressures, 

and reduce production and living costs Promote the 

construction of smart communities and digital health services, 

improve the quality of residents after birth, properly cope with 

a series of negative impacts caused by excessive 

agglomeration of factors, lead to higher levels of 

agglomeration, and promote high-quality economic 

development.  

References 

[1] Xi Jinping. Solidly promoting common prosperity [J]. Qiushi, 
2021(20): 4-8. 

[2] Wang Ruyu, Liang Qi, Li Guangqian. Virtual Agglomeration: 
A New Form of Spatial Organization with Deep Integration of 

New Generation Information Technology and Real Economy 
[J].Management World, 2018, 34(02):13-21. 

[3] Qi Yudong, Xiao Xu. Enterprise Management Reform in the 
Era of Digital Economy [J].Management World, 
2020,36(06):135-152+250. 

[4] Duan Bo, Shao Chuanlin, Duan Bo. Has the digital economy 
exacerbated regional disparities? Empirical evidence from 284 
prefecture-level cities in China [J].World Geographic Research, 
2020,29(04):728-737. 

[5] Huang Jinfang. The Impact of Internet Development on 
Regional Economic Gap under the Background of Digital 
Economy: Based on the Perspective of Labor Factor Allocation 
[J].Business Economic Research, 2021(21):167-171. 

[6] Pi Yabin. Study on the heterogeneous effect of regional 
integration on social welfare improvement [J].Modern Finance 
and Economics (Journal of Tianjin University of Finance and 
Economics), 2016,36(08):3-14. 

[7] Xi Yue. Under the tide of digital procurement, Shang Yue is 
rising with the trend [J].China Logistics and Procurement, 
2021(09):24-25. 

[8] An Tongliang, Yang Chen. The Internet Reshapes China's 
Economic Geography: Micro Mechanisms and Macro 
Effects[J].Economic Research,2020,55(02):4-19. 

[9] Yuan Chun, Xiao Tusheng, Geng Chunxiao, Sheng Yu. Digital 
Transformation and Enterprise Division of Labor: 
Specialization or Vertical Integration [J].China Industrial 
Economics, 2021, (09):137-155. 

[10] Xu Lifang, Xu Zhiwei,WangPengfei. Financial Development 
and National Savings Rate: An Inverted U-Shaped 
Relationship[J].Economic Research,2017,52(02):111-124. 

[11] Lu Hongyou, Zheng Fachuan, Jia Sha. Frontier Technological 
Progress, Technological Efficiency and Regional Economic 
Gap [J].Chinese, Resources and Environment, 2012, 
22(05):120-125. 

[12] Bian Yuanchao,Wu Lihua, Bai Junhong. The opening of high-
speed rail, factor flow and regional economic gap [J].Finance 
and Trade Economics, 2018, 39(06):147-161. 

[13] Zhao Tao, Zhang Zhi, Liang Shangkun. Digital Economy, 
Entrepreneurial Activity and High-quality Development: 
Empirical Evidence from Chinese Cities [J].Management 
World, 2020, 36(10):65-76.