Frontiers in Business, Economics and Management 
ISSN: 2766-824X | Vol. 5, No. 1, 2022 

 

21 

Research	on	the	Evaluation	Index	System	of	Urban	
Competitiveness	in	Nanjing	Metropolitan	Circle	
Juan Pan1, a, *, Guangzhong Li1, b 

1Strategy and Development Research Center of Jiangsu Province, Nanjing, 210036, China 
ajoypanda0908@163.com, blgznj@sina.com 

 

Abstract: The analysis of the level of urban competitiveness in the Nanjing metropolitan area can effectively promote the 
healthy, sustainable and higher-quality development of the metropolitan area. Based on the basic connotation of urban 
competitiveness, this paper constructs the urban competitiveness index system of Nanjing metropolitan area under the new 
development concept from five dimensions of innovation, coordination, greenness, openness and sharing. Combined with the 
data of the member cities of Nanjing Metropolitan Circle, use the AHP method to conduct empirical analysis. The study found 
that Nanjing has the highest level of urban competitiveness in the Nanjing metropolitan area, followed by Yangzhou and 
Zhenjiang，the level of competitiveness of the two cities is comparable.While the overall competitiveness of the four cities in 
Anhui Province is relatively low, of which Xuancheng has the lowest overall competitiveness. 

Keywords: City competitiveness, Evaluation index, AHP method. 
 

1. Introduction 
As an important regional unit for the development of new 

urbanization, metropolitan area is becoming an important 
space carrier for accelerating regional integration 
construction. With the coordinated development of regions, 
the process of regional integration and development is 
gradually accelerating, and the research on the development 
competitiveness of metropolitan circles is becoming the focus 
of attention. There are still development gaps among the cities 
within the Nanjing metropolitan area, and the problem of 
unbalanced and insufficient urban development still exists. 

The comprehensive competitiveness of metropolitan areas is 
an important focus of regional economic development and an 
important measure of the level of regional economic 
development. The planning scope of the Nanjing 
Metropolitan Area has been expanded to include Nanjing, 
Zhenjiang, Yangzhou, Huai'an, Wuhu, Ma'anshan, Chuzhou, 
Xuancheng 8 cities, as well as Jintan District of Changzhou 
and Liyang City, with a total area of 66,000 square kilometers 
and a permanent population of about 35 million by the end of 
2020. , the GDP of the metropolitan area is 4175.078 billion 
yuan, accounting for 4.1% of the national proportion. The 
economic data of each member city and the data of the three 
industries are shown in Table 1 and Figure 1. 

 
Table 1. Main economic data of Nanjing metropolitan area in 2020 

City 
Permanent 

population 
(10,000 people) 

GDP (100 
million yuan) 

Per capita 
production Gross 

value (yuan) 

Urbanization 
rate (%) 

Proportion of GDP 
in total metropolitan 

area (%) 
 

Nanjing 931.97 14 818 159322 86.8 35.5 
Yangzhou 456.10 6 048 132784 71.0 14.5 
Zhenjiang 321.10 4220 131580 79.5 10.1 

Huaian 455.92 4025 87507 65.7 9.6 
Chuzhou 398.71 3032 76048 61.8 7.3 

Ma'anshan 216.50 2187 101011 54.1 5.2 
Xuancheng 250.00 1608 64301 57.0 3.9 

Wuhu 364.58 3753 102964 63.8 9.0 
Jintan District 58.52 973 166294 65.8 2.3 

Liyang 78.55 1072 137457 62.9 2.6 
 

Wu Fuxiang[1] and others believe that cities at all levels in 
the Yangtze River Delta should promote complementary 
advantages and coordinated development according to their 
own development characteristics, and realize the dynamic 
allocation of industrial resources in the urban agglomeration 
under the condition of free flow of factors.There are not many 
domestic studies on the evaluation index of the five major 
development concepts. Yi Changliang (2016) [2] compiled a 
development index report with five core concepts as the 
evaluation system on the basis of benchmarking 

analysis.Yang Xinhong[3] constructed a statistical evaluation 
index system from five parts of the new development concept, 
and only analyzed the economic and social development of 
Shenzhen in the past five years. Based on the perspective of 
competitiveness, Ni Pengfei and Xu Haidong (2019) [4] 
predicted the future trend and pattern of national central cities 
from the perspectives of economic competitiveness, 
sustainable competitiveness, business competitiveness, and 
livable competitiveness. Li Jiaqi (2020) [5]conducts research 
on the coordinated development competitiveness of five 



 

22 

development concepts based on Delphi and AHP methods. 
Based on the previous research results, this study intends to 
use the AHP method to discuss the evaluation index system 
of urban competitiveness in Nanjing metropolitan area, in 
order to put forward relevant suggestions for the high-quality 
development of Nanjing metropolitan area. 

2. Methods and Materials 
The AHP method [6], also known as the Analytic Hierarchy 

Process, was proposed by the American operations researcher 
T.L.Saaty in the 1970s. The principle is to classify the 
research problems into different levels according to different 
principles, and then construct a judgment matrix according to 
the logical relationship of each level, and calculate the single-
ranking structure of the factors of a certain level relative to 
the factors of the previous level and relative to the upper level. 

One-level total ranking weight, this method is a decision 
analysis method that combines qualitative and quantitative 
analysis. 

Step1: Build a hierarchical model, that is, the classification 
of each influencing factor at the next level.  

Step2: Construct judgment matrix. The matrix is 
constructed according to the consistency matrix check 
method, but all matrix elements are compared pairwise, and 
finally the consistency check is used to determine whether the 
matrix construction is correct. Construct the judgment matrix 
A (αij is the element of the matrix, i represents the row, j 
represents the column), the value of each element in the 
judgment matrix reflects the importance judgment of one 
element to another element, and the value method is 
innovative, αij is not mandatory Select the data in Table 2, and 
can use any score between the above data to judge, each value 
represents a different relationship between them, see Table 1. 

 
Table 1. AHP Assignment Meaning Table 

assignment explanation  
 

1 Both are equally important 
3 i is slightly more important than j 
5 i is important than j 
7 i is more important than j 
9 i is definitely more important than j 

2,4,6,8 
The median value of two adjacent judgments Further subdivision is required between two 

adjacent judgments 

reciprocal If the ratio of the importance of i to j is 𝑎 , then the ratio of the importance of j to i is  

 
Step3: Weight vector and consistency indicator  
The judgment matrix A obtained by pairwise comparison is 

not necessarily consistent. 
 

             A
𝑎 ⋯ 𝑎

⋮ ⋱ ⋮
𝑎 ⋯ 𝑎

          (1) 

 
Assuming W w , w ⋯ w  is the sorting weight 

vector of the n-order judgment matrix A, when A is the 
consistency matrix, there are obviously: 

 

A

𝑤 𝑤⁄ 𝑤 𝑤⁄ ⋯ 𝑤 𝑤⁄
𝑤 𝑤⁄ 1 ⋯ 𝑤 𝑤⁄

⋯ ⋯ ⋯ ⋯
𝑤 𝑤⁄ 𝑤 𝑤⁄ ⋯ 𝑤 𝑤⁄

𝑤
𝑤
⋮

𝑤

 ⋯         (2) 

 

This shows that W （𝑤 ,𝑤 , ⋯ 𝑤 ） is the eigenvector 
of A, and the eigenvalue is n, That is to say, for a consistent 
judgment matrix, the sorting vector W is the eigenvector of A. 
Conversely, if A is a consistent inverse matrix, then: 

 
When 𝑎 1, 𝑎 𝑎 , 𝑎 ∙ 𝑎 𝑎  , A 𝑎

⎣
⎢
⎢
⎡
𝑎
𝑎

⋮
𝑎 ⎦

⎥
⎥
⎤

𝑎  𝑎  ⋯ 𝑎        (3) 

 

So,A

⎣
⎢
⎢
⎡
𝑎
𝑎

⋮
𝑎 ⎦

⎥
⎥
⎤

⎣
⎢
⎢
⎡
𝑎
𝑎

⋮
𝑎 ⎦

⎥
⎥
⎤

𝑎  𝑎  ⋯ 𝑎

⎣
⎢
⎢
⎡
𝑎
𝑎

⋮
𝑎 ⎦

⎥
⎥
⎤

⎣
⎢
⎢
⎡
𝑎
𝑎

⋮
𝑎 ⎦

⎥
⎥
⎤

𝑛

𝑛

⎣
⎢
⎢
⎡
𝑎
𝑎

⋮
𝑎 ⎦

⎥
⎥
⎤
                  (4) 

 

So this shows that W （𝑎 , 𝑎 , ⋯ 𝑎 ） is the feature 
vector of A, and since A is a relative vector of judgment W 
about the target Z matrix, then W is an ordering of the values 
of specific indicators under the criteria at all levels.  

Regarding the positive reciprocal matrix A, we have the 
following conclusions: Assuming A be an n-order the positive 
reciprocal matrix, and 𝜆  is the eigenroot with the largest 
modulus (norm) of A, then:  
①  λ_max must be a positive eigenroot, and its 

corresponding eigenvector is a positive vector;  
②Any other characteristic root λ of A always has: |𝜆|＜

𝜆 ;  
③𝜆 𝑛 is the single characteristic root of A. 
Accordingly, if 𝜆 ＞n , the judgment matrix is 

inconsistent, and the feature vector W at this time cannot truly 
reflect the proportion of 𝑦 , 𝑦 , ⋯ 𝑦 in the target matrix Z. 
The quantitative index to measure the degree of inconsistency 
is called the consistency index, which is defined as: 

 

                CI                   (5) 
 
In fact, CI is equivalent to the average of n-1 eigenvalues 

(except the largest). Obviously, for the consistent positive 



 

23 

reciprocal matrix, CI=0. However, it is not enough to use CI 
as the criterion for judging whether the matrix A has 
satisfactory consistency. For this reason, the average 
randomness consistency index RI is introduced: 

 

          RI                  (6) 
 
Step4: Calculation of Analytic Hierarchy Process  
How to judge the largest eigenroot of a matrix and its 

corresponding eigenvector is the fundamental problem of 
AHP calculation. The calculation method is as follows: 
①Normalize each column of the judgment matrix: 
 

   𝑎
∑

（𝑖. 𝑗 1,2, ⋯ 𝑛）        (7) 

 
② The normalized matrix is added row by row: 
 

     𝑤 ∑ 𝑎 𝑖 1,2, ⋯ 𝑛            (8) 
 

③Normalize the vector, that is, 𝑤
∑

is the desired 

eigenvector.  
④  Calculate the maximum eigenroot of the judgment 

matrix: 
 

       𝜆 ∑
⃗

              (9) 

 

In the above formula A𝑤 i represents the i-th element of 
the vector.  

Step5: Weight vector and combinatorial consistency check  
The total ranking of the hierarchy is to calculate the ranking 

weight of the relative importance of all factors at the same 
level to the highest level (total target), which is carried out 
layer by layer from the highest level to the lowest level.  
① Calculate the combined weight vector  
Assume the sorting weight vector of 𝑛  elements on the 

k-1 layer to the total target (the highest layer) be: 
 

   𝑊 𝑤 , 𝑤 , ⋯ 𝑤       (10) 
 
The weight vector of the 𝑛  elements on the kth layer to 

the jth element on the previous (k-1) 
 

𝑃 𝑝 , 𝑝 , ⋯ 𝑝 ，j 1,2, ⋯ 𝑛    (11) 

 
Then the matrix 𝑃 𝑃 , 𝑃 , ⋯ , 𝑃  is  𝑛  ×𝑛  

matrix, represents the sorting weight vector of the elements 
on the kth layer to the elements of the k-1th layer. Then the 
total sorting weight vector of the elements on the kth layer to 
the target layer (the highest layer) is:  

 
   𝑊 𝑃 ∙ 𝑊 𝑃 , 𝑃 , ⋯ , 𝑝 ∙

𝑊 𝑤 , 𝑤 , ⋯ , 𝑤       (12) 
 
or  

𝑤 ∑ 𝑝 𝑤 , 𝑖 1,2, ⋯ , 𝑛       (13) 
 
For any k>2 there is a general formula 𝑊 𝑃 ∙

𝑃 ⋯ ⋯ 𝑃 ∙ 𝑊 k 2  , and 𝑊   is the total sorting 
weight vector of each element on the second layer to the target 
layer. 
②Combination consistency test 
Assume the consistency index of layer k is 

𝐶𝐼  ,  𝐶𝐼  ,… 𝐶𝐼  , The random consistency index 

is:  𝑅𝐼  ,  𝑅𝐼  ,…  𝑅𝐼  Then the combined consistency 
index of the kth layer to the target layer (the highest layer) is:  

 
     𝐶𝐼 𝐶𝐼 , 𝐶𝐼 , ⋯ , 𝐶𝐼 ∙ 𝑊     (14) 
 
The combined random consistency index is:  
 

𝑅𝐼 𝑅𝐼 , 𝑅𝐼 ⋯ , 𝑅𝐼 ∙ 𝑊    (15) 
 
The combined consistency ratio indicator is:  
 

      𝐶𝑅 𝐶𝑅 𝑘 3           (16) 
 
When 𝐶𝑅 ＜0.10, the comparison judgment matrix of the 

whole level is considered to pass the consistency test. The 
corresponding RI numbers see Table 2. 

 
Table 2. Average random consistency index RI 

n 1 2 3 4 5 6 7 8 9 

R
I 

0 0 0
.52 

0
.89 

1
.12 

1
.26 

1
.36 

1
.41 

1
.46 

 
Data normalization: In order to facilitate the comparison 

and analysis of the data, standardization processing is 
required. This paper adopts the range standardization method. 
The calculation formula of the factor with positive correlation 
in the final score is (16), and the calculation formula of the 
factor with negative correlation is (17) 

 

         𝑋              (17) 

 

         𝑋              (18) 

 
Step6: Adjust and recalculate the situation that does not 

meet the consistency check.  
When using the AHP method to construct the matrix, the 

Delphi method is used for comprehensive decision-making. 
Experts evaluate and score the importance of the indicators, 
and multiple experts adjust the influencing factors. On the 
basis of the AHP method, the original matrix is normalized to 
obtain the results of each judgment matrix, and the 
consistency test (CI/RI) is carried out to confirm that the 
matrix is valid, and the final index weight result is obtained[7].  

Data sources:There are 30 indicators at the indicator level 
in this article. The data are all obtained from the statistical 
yearbooks officially released by the cities in 2021 and the 
statistical bulletins of national economic and social 
development, and are obtained directly or indirectly. Due to 
the limitation of data samples, the analysis of Nanjing 
metropolitan area in this paper is limited to prefecture-level 
cities, and Liyang and Jintan in Changzhou are not included 
in the scope of data analysis. 



 

24 

3. Results and Discussion 

3.1. Evaluation System 
The connotation of urban competitiveness is relatively rich. 

Before evaluating urban competitiveness, it is necessary to 
build a relatively complete evaluation index system for urban 

competitiveness, and try to select all indicators that can reflect 
the connotation of urban competitiveness. There are 3 levels 
in the evaluation index system of city competitiveness which 
belongs to the index weight determination, namely the target 
level (A), the criterion level (B) and the index level (C),see 
Table 3. Criterion layer weights and Index layer weights are 
shown in Table 4 and Table 5. 

 
Table 3. Urban Competitiveness Index System 

City Competitiveness A 

Economic Coordination 
Competitiveness B1 

GDP per capita C1 
GDP growth rate C2 

Total GDP C3 
Urbanization rate C4 

Income ratio of urban and rural residents C5 
Total investment in fixed assets C6 

The output value of the secondary industry accounts for the 
proportion of GDP C7 

The output value of the tertiary industry accounts for the 
proportion of GDP C8 

Ratio of consumption expenditure of urban and rural residents 
C9 

Innovation Ability 
Competitivenes 

B2 

R&D expenditure C10 
R&D spending as a percentage of GDP C11 

Number of patent applicationsC12 
Number of patents granted C13 

The number of ordinary colleges and universities C14 

Openness Competitiveness 
B3 

Total Import and Export C15 
The proportion of total imports and exports to GDP C16 

Total actual utilization of foreign capital C17 
Tourism Income C18 

Total freight volume C19 
Passenger volume C20 

Green Development 
Competitiveness 

B4 

Park green space per capita C21 

Green coverage in built-up areas C22 

Shared level 
competitiveness 

B5 

Population density C23 
Urban registered unemployment rate C24 

Local fiscal revenue per capita C25 
Local fiscal expenditure per capita C26 

Urban road area per capita C27 
Rail transit mileage C28 

Number of beds in hospitals C29 
Number of public libraries C30 

 
Table 4. Criterion layer weights of urban competitiveness evaluation index system 

Index B1 B2 B3 B4 B5 
Weights 0.3521 0.2358 0.0946 0.1082 0.2093 

 
Table 5. Index layer weights of urban competitiveness evaluation index system 

Index Weights Index Weights Index Weights Index Weights Index Weights 
C1 0.2091 C7 0.0391 C13 0.3042 C19 0.0824 C25 0.0445 
C2 0.1538 C8 0.0734 C14 0.4627 C20 0.1123 C26 0.3327 
C3 0.1223 C9 0.0937 C15 0.1242 C21 0.0901 C27 0.0482 
C4 0.0751 C10 0.0157 C16 0.1089 C22 0.5351 C28 0.1021 
C5 0.086 C11 0.0457 C17 0.3498 C23 0.4649 C29 0.2065 
C6 0.0553 C12 0.0308 C18 0.3654 C24 0.0537 C30 0.2123 

 

3.2. Indicator Description  
By measuring the spatial distribution of the industrial 

structure in the region by the location entropy, it is found that 
in the Nanjing metropolitan area, only Nanjing is dominated 

by the tertiary industry, and other cities are mainly 
concentrated in the primary or secondary industry.see Figure1 
and Figure2. The formula for calculating location entropy is: 

𝐿𝑄
⁄

⁄
,Among them, 𝐺  represents the output value of 



 

25 

industry j in city i, 𝐺  represents the output value of industry 
j in the metropolitan area, 𝐺 represents the total output value 
of city i, G represents the total output value of metropolitan 

area, and 𝐿𝑄  represents the location entropy. The evaluation 
results are shown in Table 5 and Figure 3. 

 
 

 
Figure 1. Location entropy of three industries in Nanjing metropolitan area in 2020 

 
 

 
Figure 2. Three industries in Nanjing metropolitan area 

 
 

Table 5. Criterion layer index score and competitiveness index of city competitiveness in Nanjing metropolitan area 
City Nanjing Zhenjiang Yangzhou Huaian Chuzhou Ma'anshan Wuhu Xuancheng 
B1 0.2749 0.1853 0.1899 0.1693 0.1092 0.1185 0.1349 0.0862 
B2 0.2963 0.1734 0.1921 0.1576 0.0548 0.0461 0.0812 0.0554 
B3 0.1325 0.0921 0.1254 0.1164 0.0564 0.0794 0.1187 0.0441 
B4 0.0654 0.0215 0.0403 0.0321 0.0167 0.0221 0.0376 0.0067 
B5 0.0745 0.0436 0.0564 0.0322 0.0195 0.0396 0.0412 0.0074 

Total 
score 

0.8436 0.5159 0.6041 0.5076 0.2566 0.3057 0.4136 0.1998 

Sort 1 3 2 4 6 7 5 8 
 

0

0.5

1

1.5

2

2.5

Nanjing Yangzhou Zhenjiang Huaian Chuzhou Ma'anshan Xuancheng Wuhu

LQ1 LQ2 LQ3

0.00
1000.00
2000.00
3000.00
4000.00
5000.00
6000.00
7000.00
8000.00
9000.00

10000.00

Primary industry secondary industry tertiary industry



 

26 

 
Figure 3. Evaluation results of urban competitiveness in Nanjing metropolitan area 

 
 

The indicator system includes innovation, coordination, 
green, openness and sharing 5 parts, and a total of 30 
indicators are set up, the coordinated development and shared 
development parts set 9 and 8 indicators respectively.Taking 
into account the particularity of the development orientation 
of the Nanjing metropolitan area. In order to build a national 
model area for the development of the same city, 5 indicators 
are set for the development of innovation capability, 6 
indicators are set for the openness level, and 2 indicators are 
set for the green development part.  

4. Conclusions 
Based on the AHP method, this paper obtains the 

competitiveness ranking of cities in the Nanjing metropolitan 
area, and puts forward specific ideas and innovative measures 
for the development of the metropolitan areas under the new 
requirements of high-quality and integrated development, so 
as to enhance the overall strength and competitiveness of the 
metropolitan area. 

(1) As the core city in the metropolitan area, Nanjing has 
leading advantages in terms of industrial development, 
innovation level, coordinated development and public service 
sharing. Compared with other cities, Nanjing has the highest 
level of urban competitiveness. In terms of economic strength, 
Nanjing has an absolute advantage, and its total scale is 9 
times that of the lowest Xuancheng city. Therefore, it is 
necessary to give full play to the leading and leading role of 
Nanjing's central city. The advantages of Nanjing's innovation 
primacy should be fully amplified, the integration of 
innovation resources should be deepened, the R&D platform 
should be jointly built in the metropolitan area, and the 
technological breakthroughs of major projects should be 
coordinated to form a cross-regional industry-university-
research innovation complex. 

(2) Overall, Nanjing, Yangzhou, Zhenjiang and Huai'an of 
the 4 cities in Jiangsu Province are more competitive than the 
adjacent 4 cities in Anhui Province. Among them, Yangzhou 

ranks second in city competitiveness, and Zhenjiang and 
Huai'an are comparable in competitiveness, ranked third and 
fourth respectively.The overall competitiveness of Chuzhou 
and Xuancheng is relatively low. Because Xuancheng is 
relatively far away from Nanjing, it is limited in accepting the 
radiation power of core cities, resulting in the lowest level of 
urban competitiveness.In the future, it is even more necessary 
to break down the administrative barriers between regions, 
and at the same time, further strengthen collaborative 
innovation and industrial cooperation. 

Acknowledgment 
This research was financially supported by:(1) Social 

Science Applied Research Excellent Engineering Project of 
Jiangsu Province (No:22SYA-005); (2)Social Science 
Fundation of Jiangsu Province(21GLB007). 

References 
[1] F.X.Wu,H.P.Shen:New urbanization,spatial spillover of 

infrastructure and upgrading of regional industrial structure t—
An empirical analysis based on 16 core cities in the Yangtze 
River Delta urban agglomeration, financial science, 
(2013),No.7,p.89-98. 

[2] C.L.Yi:2015 China Development Index Report—New Concept 
and New Development of Innovation, Coordination, Green 
Opening and Sharing(Economic Science Press, China, 
2016)p.283-290.  

[3] X.H.Yang:Construction of Five Development Concepts 
Statistical Evaluation Index System—Taking Shenzhen as an 
Example,Research World,(2017)No.7,p.3-7. 

[4] P.F.Ni, H.D.Xu:The New Pattern of Competitiveness of 
National Central Cities,China Real Estate.,(2019),No.23,p.14-
15.  

[5] J.Q.Li:Research on the Competitiveness of Coordinated 
Development of Five Development Concepts t—Comparative 
Analysis of National Central Cities and Sub-provincial Cities 
Based on Delphi and AHP Methods,Journal of Party School of 

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

Nanjing Zhenjiang Yangzhou Huaian Chuzhou Ma'anshan Wuhu Xuancheng

Coordination level Innovation level Open level Green development Shared level



 

27 

Sichuan Provincial Committee Committee of CPC, 
Vol.6(2020)No.2, P.99-109.  

[6] H.Y.Xing:Research on the Design of Evaluation Index System 
of Online Courses based on AHP Method and Delphi Method, 
China Electrochemical Education,(2006)No.9,p. 78-81.  

[7] W.Chen,T.J.Fang,X.D.Jiang: Research and Application of 
Group Decision-making based on Delphi Method and AHP 
method, Computer Engineering, Vol. 29(2003)No.5,p.18-20.