Journal of Innovation & Knowledge 8 (2023) 100287

Journal of Innovation
& Knowledge
ht t p s: // w w w . j our na ls .e l se vi e r .c om /j ou r na l -o f - in no va t i on -a n d- kn owl e dg e

Evaluating high-tech industries’ technological innovation capability and

spatial pattern evolution characteristics: Evidence from China
Wenjuan Tua, Lixiang Zhangb, Dongying Suna,*, Wenqing Maoc
a
School of Management, Jiangsu University, Zhenjiang, Jiangsu 212013, China
b
School of Management, Shanghai University, Shanghai 200444, China
c
Information Office, University of Shanghai for Science and Technology, Shanghai 200093, China

A R T I C L E I N F O A B S T R A C T

Article History: The development of high-tech industries in an open economy has become an inevitable requirement for all
Received 25 March 2022 countries to enhance their competitive advantages and overall national strength. Technological innovation is
Accepted 21 December 2022 a direct driving force for promoting high-tech industries’ development. The technological innovation capabil-
Available online 3 January 2023
ity of China’s high-tech industries over 2010−2019 was measured using factor analysis. Moreover, the spatial
pattern and evolution characteristics of the technological innovation capability of high-tech industries were
Keywords:
analyzed using the Moran index. The empirical results demonstrate that: (1) during the research period, the
Technological innovation
technological innovation capability of China’s high-tech industries rose, but regional development was unco-
Capability of high-tech industries
Factor analysis
ordinated. Input-output and transformation guarantee capacities were the main factors influencing techno-
Moran index logical innovation. (2) Improvements in the technological innovation capability of China’s high-tech
Spatial pattern industries were spread over the southeast coast to the central and southwest regions. The technological inno-
vation capabilities of high-tech industries in different regions exhibited an evident spatial correlation. The
spatial agglomeration distribution is relatively stable, mainly characterized by “Low-Low” (L-L) agglomeration.
This study offers managerial enlightenment for improving the technological innovation capability of high-tech
industries to accelerate industrial development to the middle and high end of the global value chain.
© 2022 The Authors. Published by Elsevier España, S.L.U. on behalf of Journal of Innovation & Knowledge. This
is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)

Introduction industries into six categories: manufacturing medicines, aircraft and

spacecraft and related equipment, electronic and communication equip-
High technology, as an emerging cutting-edge technology, signifi- ment, computers and office equipment, medical equipment and meters,
cantly influences changes in the economic landscape of countries. and electronic chemicals. The contents of the first to fifth categories in
High-tech industries significantly contribute in industrial restructur- the classification table are linked to the international classification to
ing and economic transformation. Their development is fundamental accommodate international comparison. Technological innovation, as a
for overcoming the middle-income trap and building a modern and prerequisite for technological development, is an engine for developing
powerful country. Their participation in the global value chain (GVC) high-tech industries and plays a catalytic role in upgrading the GVC.
plays a pivotal role in the country’s international status. Accelerating This study examines the provincial innovation capacity of high-
industrial development toward the middle and high ends of the GVC tech industries in China during 2010−2019. The remainder of this
has become a strategic choice for numerous regions. paper is structured as follows: Section 2 provides a literature review.
In 2020, the profits of China’s high-tech industries increased by 16.4% Section 3 elucidates the research methods, indicator systems, and
year on year, and the profits achieved by high-tech industries were evaluation results. Section 4 introduces the evolutionary characteris-
17.8%. These industries maintained relatively rapid profit growth and, as tics of the spatial pattern. Section 5 concludes.
the industrial segment with the fastest profit growth, strongly promoted
the continuous optimization of the industrial profit structure. Owing to
Literature review
its considerable economic benefits and optimizing effect on the existing
industrial structure, high-tech industries are important for international
In “The Theory of Economic Development”, Schumpeter opined
economic and technological competition. China classifies high-tech
that innovation creates a new production function, introducing new
factors and conditions into the production system that never existed
* Corresponding author. before (Schumpeter, 2003). The Organization for Economic Coopera-
E-mail address: [email protected] (D. Sun). tion and Development (OECD) considers technological innovation to

https://doi.org/10.1016/j.jik.2022.100287
2444-569X/© 2022 The Authors. Published by Elsevier España, S.L.U. on behalf of Journal of Innovation & Knowledge. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/4.0/)
W. Tu, L. Zhang, D. Sun et al. Journal of Innovation & Knowledge 8 (2023) 100287

include the commercialization of new products and the application of and error models to empirically test the key factors affecting the spa-
new processes. Innovation encompasses a range of scientific, techno- tial spillover of a country’s high-tech industry. Their results showed
logical, organizational, financial, and commercial activities. Freeman that R&D investment and international trade contribute positively to
and Soet (1997) stated that technological innovation is the first com- the spatial spillover of high-tech industries in different regions. Jo et
mercial transformation of new products, processes, systems, and al. (2020) demonstrated that industrial agglomeration in the Korean
services. The existing literature possesses rich research on the influ- region enhances the innovation efficiency and capacity of high-tech
encing factors, index systems, and evaluation methods of high-tech firms, leading to sustained innovation performance in the region.
industries’ technological innovation capability. Seddighi and Mathew (2020) presented a theoretical/empirical
The extent to which influencing factors affect different industries framework for the promotion of innovation via enhancement of a
varies significantly (Zhu et al., 2019). As a key to developing high- firm’s core competence, and improvement in its output/product char-
technology industries, there has been a long-standing debate on the acteristics.To meet their respective needs for the results of evaluation
influence of different firm sizes on technological innovation. Scherer studies, scholars use evaluation methods appropriate for the subject
(1965) argued for an inverted U-shaped relationship, whereas Kraft and direction of the study. For example, the Moran index measures
(1989) posited that there is no significant relationship. Braga and the degree of spatiotemporal autocorrelation. Moran scatter plots
Willmore (1991) demonstrated a positive relationship; however Lee effectively reveal the characteristics of spatiotemporal clustering pat-
et al. (2010) suggested that this relationship is negative. Zhou et al. terns and changes in patterns (Shen et al., 2016).
(2017) found that the contribution of research and development This study contributes to the literature in two respects. First, it
(R&D) capability to the innovation incubation capacity of high-tech enriches the research on the technological innovation capability of
industries is more important, and Lopez-Garcia et al. (2012) found a high-tech industries by combining the main influencing factors with
positive connection between human resources and industrial-tech- the degree of spatio-temporal autocorrelation. Existing studies have
nological innovation capability. Menaker and Ozoliņa (2018) found focused on spatial variation, which has practical implications for
that government assistance was important in Latvia. The lack of tech- national planning studies; however, the analysis in the temporal and
nical experts and complexity of the real estate space were important spatial dimensions must be supplemented. Second, this study is moti-
limiting factors. Chen et al. (2020a) found that government support, vated by the development of a composite indicator system for tech-
R&D investment intensity, industrial agglomeration, outward eco- nological innovation capabilities. This study adopts the factor
nomic orientation, and the development of modern service industries analysis method to effectively screen the evaluation indicators to
exhibit different degrees of influence on innovation efficiency. Guo et eliminate indicators that are not explanatory. Therefore, factor analy-
al. (2022) conceptualized the novel concept of generative capability, sis and the Moran index were chosen as the main research methods
a unique capability by which is positively associated with firm inno- to evaluate the innovation capability, spatial and temporal distribu-
vation performance.Thus, numerous factors influence innovation. tion characteristics, and spatial autocorrelation of high-tech indus-
The evaluation index system was constructed using various tries in China. This study’s conclusions provide a reference for other
methods. Yu et al. (2018) studied the key factors in the ecological developing countries’’ high-tech industries.
development of high-tech industries in the Hubei Province.
Among the 14 evaluation indicators, the top five were product
innovation, cost and exchange rate, knowledge dissemination,
industrial structure, and experience accumulation. Sun and Sun Evaluation of technological innovation capability in china’s high-
(2018) used the social network analysis method to extract indica- tech industries
tors step by step and determine the regional science and technol-
ogy innovation capability evaluation index in China. Primary Data sources
indicators include the foundation, input, output, and benefits of
science and technology innovation. Yao and Ma (2017) selected The data were sourced from the China Statistics Yearbook on High
13 input and output indicators and identified 12 indicators after Technology Industry 2011−2020 and the China Statistical Yearbook
the correlation analysis to construct the technical efficiency of 2011−2020. Given that the 2018 data from the China High Technol-
the high-tech industry’s evaluation model in Jilin Province. Sumrit ogy Industry Statistical Yearbook were not included, some of the data
and Anuntavoranich (2012) designed an evaluation index system relating to 2018 were selected from the China Statistical Yearbook on
based on three dimensions—management, input, and innovation Science and Technology 2019, and some of the missing data were
—and found that management capability was the most important filled using the mean value method. As the breakdown in the various
influencing factor. Wang et al. (2020) divided technological inno- statistical yearbooks predominantly pertains to the breakdown of
vation activities into R&D and commercialization phases, includ- manufacturing industries, high-tech industries in this study refer
ing shared, intermediate, and free intermediate outputs. Their specifically to high-tech manufacturing industries. As the original
experimental results suggested reduced R&D in high-tech indus- data for Tibet and Qinghai were incomplete, the analysis did not
tries in China, however, there is potential for breakthroughs. involve these two provinces, nor did it include Hong Kong, Macao,
There is no single application for evaluation methods and per- and Taiwan.
spectives. Zhu et al. (2019) constructed a semiparametric model to
comprehensively compare the factors influencing the technological
innovation performance of different high-technology enterprises.
Chen et al. (2020a) applied data envelopment analysis (DEA) and a Design of the evaluation indicator system
spatial econometric model to measure the innovation efficiency and
influencing factors of high-tech enterprises in China. They deter- This study considers the decomposition of the technological inno-
mined significant differences among provinces. Using factor analysis, vation capability system and its constituent elements while consider-
distinguishing the impacts of the four main factors on regional inno- ing combining total and relative indicators and qualitative and
vation capacity is possible (Mikel Buesa et al., 2006). This method can quantitative indicators, referring to the research results of related lit-
also measure a region’s innovative capacity (Martínez Pellitero et al., erature and features of the high-tech industry. The indicator system
2008). He et al. (2018) used factor analysis to compare the competi- is divided into four dimensions: innovation input, transformation,
tiveness of high-tech industries in China and discussed their spatial environmental protection, and output capacities. Then, it is decom-
distribution characteristics. Chen et al. (2020b) established spatial lag posed layer-by-layer (Table 1).
2
W. Tu, L. Zhang, D. Sun et al. Journal of Innovation & Knowledge 8 (2023) 100287

Table 1
Evaluation index system of technological innovation capability of high-tech industries.

Target level First-level index Secondary index Tertiary index

Technological Innovation input capacity B1 Financial input C1 Intramural expenditure on R&D (10000yuan) D1
innovation capability Proportion of equipment in intramural expenditure on R&D (%) D2
Expenditure on new products development (10000yuan) D3
Human resource input C2 Full-time equivalent of R&D personnel (man-year) D4
Number of researchers among R&D personnel (person) D5
Personnel in R&D institutions(person) D6
Innovation transformation B2 Technology conversion C3 Expenditure on technology conversion (10000yuan) D7
capacity Rate of Projects Completed and Put into Use (%) D8
Environmental safeguard B3 Innovative places C4 Number of R&D Institutions in Enterprises (unit) D9
capacity Government support C5 Proportion of science and technology expenditure in financial expenditure (%) D10
Proportion of government funds in intramural expenditure on R&D (%) D11
Labor productivity C6 The average number of GDP generated by all employees in society per person D12
(yuan/person)
Innovation output capacity B4 Patent output C7 Growth in the number of valid patents (piece) D13
Number of patents filed per 10,000 R&D personnel (piece/10,000 people) D14
New product output C8 Profits (100 million yuan) D15
Proportion of sales revenue of new products in revenue (%) D16

Table 2
KMO and Baetlett’s test values.

Test method Test value Test condition Test result

KMO test 0.837 Kaiser test criteria Suitable for factor analysis
Bartlett’s spherical test Approximate cardinality 7373.713 Less than the significance level of 0. 05 Suitable for factor analysis
df 120
Sig. 0.000

Factor analysis both ends. The rotated component matrices are listed in Table S1 in
the Supporting Information.
Basic ideas of factor analysis The indicators were grouped into two categories, and the factors
The basic idea of factor analysis is transforming multiple indica- were named based on indicator rankings; the results are shown in
tors into a small number of composite indicators by creating a linear Table 4. The Kaiser standardized orthogonal rotation method was
mapping from a high-dimensional to a low-dimensional space using applied to obtain the factor score coefficient matrix. Factor scores
dimensionality reduction. These unmeasured composite indicators were calculated based on their function, the product of the factor
are often referred to as common factors. The original variables are score coefficients, and standardized values of each original variable.
grouped based on the magnitude of the correlation such that the vari- The results are presented in Tables S2 and S3 in the Supporting Infor-
ables within the group are more correlated and those between the mation.
groups are less correlated. Each group of variables represents a com- Each factor’s combined score was calculated by taking the vari-
mon factor that reflects one aspect of the problem. The variance con- ance of each factor in the rotated cumulative variance contribution
tribution of several common factors is then used as weights to table as the weight of each factor, and the formula was F = 0.868*F1
construct a comprehensive evaluation function, simplifying the +0.132*F2. The same SPSS calculation can be used to obtain the com-
numerous original variables and effectively dealing with repetitive bined scores of the technological innovation capability of each region
information between indicators. and its rankings. For observation, the ranking is sorted in ascending
order, as shown in Table S4 in the Supporting Information.
Process of factor analysis
SPSS26.0 was used to conduct the factor analysis to assess the
technological innovation capability of 29 provinces in China’s high-
tech industries. In the first screening, variables with an information Table 3
deficit value greater than 60% were excluded according to the vari- Eigenvalues and variance contribution rates corresponding to each public factor.
able commonality scale (D2, D11, D12, D14, and D16).
Category F1 F2
Factor analysis requires a strong correlation between original var-
iables. Therefore, correlation analysis and testing of variables were Eigenvalue 8.378 1.234
required before the factor analysis. The results are presented in Variance contribution rate% 76.161 11.220
Cumulative variance contribution % 76.161 87.381
Table 2.
Following the criterion of eigenvalues greater than 1, two public
factors were selected with a cumulative contribution of 87.381% to Table 4
the variance. They represent most of the information and fully reflect Main factor analysis table.
high-tech industries’ comprehensive development level in China. The
Factor Explained indicators Factor naming
corresponding eigenvalues and variance contribution rates for each
public factor are listed in Table 3. F1 D6,D1,D3,D4,D15,D9,D13,D7,D5 Technology innovation input-output
To facilitate the classification of the factors, factor loadings on dif- capacity
F2 D8,D10 Technology innovation transforma-
ferent original variables allowed significant differences. Therefore,
tion guarantees capacity
the initial factor-loading matrix must be rotated to concentrate it at
3
W. Tu, L. Zhang, D. Sun et al. Journal of Innovation & Knowledge 8 (2023) 100287

Table 5 regions. Therefore, the central and western regions need to focus on the
Regional divisions. input and output capacities. However, the technological innovation
Category Region input and output capacities of the eastern region have grown steadily
to enable China’s overall technological innovation input and output
Eastern Region Beijing, Tianjin, Hebei, Liaoning, Shanghai, Jiangsu, Zhejiang,
capacity to achieve effective improvement.
Fujian, Shandong, Guangdong, Hainan
Central Region Shanxi, Jilin, Heilongjiang, Anhui, Jiangxi, Henan, Hubei, Fig. 2 shows that during 2010−2019, the high-tech industry’s
Hunan technological innovation capacity generally showed a “long-tailed U-
Western Region Chongqing, Sichuan, Guizhou, Yunnan, Shaanxi, Gansu, Ning- shaped” trend, declining before rising, with 2015 being an inflection
xia, Xinjiang, Guangxi, Inner Mongolia
point. Although changes in the East have fluctuated, the region has
consistently been above above the national average. The central
region has been on an upward trend since reaching its lowest point
For achieving an enhanced overall analysis, the 29 provinces and in 2015. Similarly, the distribution of technological innovation trans-
municipalities studied were divided based on the National Bureau of formation security capacity within the region is uneven, as shown in
Statistics division criteria, as shown in Table 5. Table S3, with seven of the top 15 in eastern regions, two in the cen-
To compare the means of the factor and composite scores of the tral region, and six in the western region. Although the eastern region
eastern, central, and western regions with the national average each still ranks high, the trend in the national average is mainly consistent
year, as shown in Figs. 1−3. with changes in the central region. Thus, efforts should be made to
coordinate the strategic arrangements between the eastern and
western regions in terms of the transformation guarantee capacity
Results and analysis and raising the level of conversion security capacity in the eastern
and western regions.
Fig. 1 shows that the technological innovation input and output Fig. 3 illustrates that the technological innovation capacity of
capabilities of high-tech industries in China generally showed an insig- high-tech industries in China generally showed an upward trend dur-
nificant upward trend during 2010−2019. The eastern, central, and ing 2010−2019. The change in trend is almost the same as the tech-
western regions showed clear characteristics of regional heterogeneity, nological innovation input-output capacity, which shows that the
and the gap between the regions expanded. The rising trend is faster in input-output capacity is the dominant influencing factor of the tech-
the eastern region and slower in the central and western regions. nological innovation capacity of high-tech industries in China. Thus,
Table S2 shows that the top 15 regions have eight in the east-central, the uneven distribution within regions is similar to the differences in
five in the central, and two in the western regions. The national average the input-output capabilities. Table S4 shows that eight of the top 15
trend is mainly consistent with the changes in the central and western regions were in the central-eastern region, five in the central region,
and two in the western region. The technological innovation capacity
in the central and western regions did not reach the national average
during the study period, which indicates a long journey ahead.
We provide the following summary by examining the results.

(1) From the perspective of factor extraction, the main factors influenc-
ing the technological innovation capability of high-tech industries
are input-output and transformation guarantee capabilities. The
input and output are the main forces, the transformation guarantee
is the auxiliary force, and the two are complementary. Therefore,
enhancing the technological innovation capability of high-tech
industries in China by increasing resource input, improving output
efficiency, and strengthening environmental safeguard capability,
especially government policy and financial support, is necessary to
Fig. 1. Trend of technological innovation input-output capacity in high-tech industries improve technological transformation capability.
of China from 2010−2019. (2) The scores of the technological innovation capabilities in high-
tech industries indicate uneven development between regions in

Fig. 2. Trend of technological innovation transformation guarantees capacity in high- Fig. 3. Trend of technological innovation capacity in high-tech industries of China from
tech industries of China from 2010−2019. 2010−2019.

4
W. Tu, L. Zhang, D. Sun et al. Journal of Innovation & Knowledge 8 (2023) 100287

Fig. 4. Distribution of composite scores of technological innovation capability in high-tech industries by region in 2010, 2013, 2016, and 2019.

China. The eastern region leads in technological innovation capa- According to the distribution characteristics of the values, the total
bility, in which regard the central and western regions are weaker. scores for technological innovation capability over 2010−2019 were
They did not achieve the national average during the study divided into five intervals, namely -0.501−0.000, 0.001−1.000, 1.001
period; further, there is uneven development within regions. −2.000, 2.001−4.000, 4.001−7.000. The distribution graphs of the
Additionally, there is a slight difference in the concentrations of three-year intervals for 2010, 2013, 2016, and 2019 were drawn
technological innovation input-output and transformation guar- using ArcGIS 10.2 (Fig. 4) to analyze the distribution characteristics of
antee capacities, with weaker input-output capacity being stron- the composite score of technological innovation capability in differ-
ger in technology transformation and environmental safeguards. ent provinces, excluding Tibet, Qinghai and Hong Kong, Macao, and
The Chinese government is aware of this imbalance and has Taiwan.
ensured environmental safeguards in capacity coordination to
upgrade regions with weaker input and output capacities. Results of the analysis of spatial and temporal distribution
Although the state has focused on promoting high-tech industries characteristics
in the central and western regions, providing policy and financial The five regions divided by the total score value were divided into
support, the western regions are clearly weaker than the central the first, second, third, fourth, and fifth echelons according to the size
regions in transformation guarantee capacity. Considering that of the value—from the smallest to the largest. The more regions in
most western regions are more geographically isolated and less the fifth echelon, the darker the color of the regions in the graph, and
economically developed, they are limited in the transformation the higher the level of technological innovation in China’s high-tech-
capacity level. They cannot fully integrate resources, including nology industry. Based on Fig. 4, the specific analysis is as follows.
equipment and workforce, with national policy support to facili- Regarding changes in the number of echelons, most regions
tate the technological innovation transformation. Meanwhile, the remained in the first echelon during the study period. In contrast,
eastern region was the first province and city to implement a regions in the second echelon increased from five in 2010 to 11 in
coastal opening policy. Thus, it has a higher level of economic 2019. Regions in the third echelon and above increased from one
development and is the first to enjoy national support and devel- region in 2010 to three in 2019, indicating a steady increase in the
opment due to its location (Banwo et al., 2017). technological innovation capability in high-tech industries in China.
However, the gradients are insignificant, and there are insufficient
high-level areas. In 2019, only one region was in the 3rd, 4th and 5th
Analysis of the characteristics of the evolution of the spatial echelons, Zhejiang, Jiangsu, and Guangdong, which were also ranked
pattern of technological innovation capability in high-tech in terms of input and output capacity. However, a wide gap exists in
industries of China transformation guarantee capacity, with Zhejiang scoring 1.480
(ranked 3rd), Jiangsu scoring 1.563 (ranked 1st), and Guangdong
Analysis of the spatial and temporal distribution characteristics of scoring -0.220 (ranked 22nd), thereby demonstrating the importance
technological innovation capability in high-tech industries of China of input-output efficiency for technological innovation capacity
enhancement.
Process of the analysis of spatial and temporal distribution Regarding changes in the geographical distribution of each eche-
characteristics lon, the spatial echelons of technological innovation capabilities in
To better demonstrate the spatial and temporal distribution char- high-tech industries in China changed significantly during the study
acteristics, we observe the total scores of each region in Table S4. period, with higher levels in the southeastern coastal regions and
5
W. Tu, L. Zhang, D. Sun et al. Journal of Innovation & Knowledge 8 (2023) 100287

lower levels in the northeast and other inland regions. Over time, Table 6
with Jiangsu, Zhejiang, and Guangdong as the centers, the technologi- Moran index of technological innovation capability of China’s high-tech industries.

cal innovation capabilities of the surrounding regions spread and Year Moran’s I P Value Year Moran’s I P Value
penetrated the inland regions. This indicates that national policies
2010 0.1122* 0.081 2015 0.1030* 0.074
are steadily advancing, with the strategy for the rise of the central
2011 0.1263* 0.067 2016 0.0947* 0.087
region and for the evolution of the western region gaining in terms of 2012 0.1063* 0.093 2017 0.0790* 0.097
technological innovation in high-tech industries, promoting the coor- 2013 0.1223* 0.068 2018 0.0589* 0.091
dinated development of all regions. 2014 0.1308* 0.057 2019 0.0602* 0.089
Note. * represents the 10% significance level.
Spatial autocorrelation analysis of technological innovation capabilities
in high-tech industries of China A Local Moran’s I greater than 0 indicates the existence of H-H
agglomeration or L-L agglomeration in the area. When it is less than
To further examine whether the technological innovation capabil- zero, it implies an H-L agglomeration or an L-H agglomeration in the
ity of high-tech industries in China is agglomerative, the global area.
Moran index was used to show the degree of difference in globally
relevant industries. Simultaneously, to reflect the spatial dependence
Process of spatial autocorrelation analysis
between a region and its neighbors in more detail, the local Moran
(1)Analysis of global spatial autocorrelation
index was used to scale the degree of spatial agglomeration of the
This study uses the 0-1 spatial weight matrix constructed by
technological innovation capability of high-tech industries.
GeoDa to measure the spatial global Moran index during 2010−2019,
using stata16.0 software to analyze the spatial correlation of the tech-
Moran index
nological innovation capability of high-tech industries in China.
Spatial correlation analysis was used to characterize the degree of
Table 6 presents the results.
clustering and dispersion of the distribution of the same variable in
(2)Analysis of local spatial autocorrelation
different regions. The stronger the spatial correlation, the more con-
This study uses GeoDa software to draw LISA scatter plots of the
centrated the distribution.
local Moran’s I for 2010, 2013, 2016, and 2019, with the four distribu-
tion patterns to reflect further the spatial agglomeration distribution
characteristics of the technological innovation capability of high-tech
(1) Global spatial autocorrelation
industries in China (shown in Fig. 5). The analysis does not include
Hainan Province, as no neighboring provinces exist.
The Global Moran’s I is generally used to represent the correlation
characteristics of the spatial distribution and is calculated by the fol- Results of spatial autocorrelation analysis
lowing Eq. (1). (1)Outcome of the global spatial autocorrelation analysis
P
m P
m Table 6 shows that the global Moran index values for 2010−2019
Wij ðxi  xÞðxj  xÞ
i¼1 j¼1
are all greater than 0, and the p-values are all significant at the 10%
Global Moran0 s I ¼ ð1Þ confidence level. This means that the technological innovation capa-
P
m P
m
s2 Wij bility of high-tech industries in China is not randomly distributed.
i¼1 j¼1
There is also a positive spatial correlation between them. Considering
Xi is the area observation, and Wij is the spatial weight matrix, where the global Moran Index value, although it fluctuated during 2010
x and s2 are the mean and variance of the area observation, respec- −2015, it consistently remained above 0.1. However, since 2016, the
tively. The spatial weight matrix is a 0-1 matrix generated by the value has been less than 0.1 and has declined significantly over 2016
Geoda software. The value of the spatial weight matrix corresponding −2018, falling to the lowest point of 0.0589, implying the relevance
to two adjacent regions is 1, while that corresponding to two non- of the technological innovation capability of the high-tech industry of
adjacent regions is 0. The Global Moran’s I takes the value interval China as a whole gradually weakens. Accordingly, it can be inferred
[-1,1]. When the value is in the interval [0,1], the attribute shows a that the spatial spillover effect of China’s high-tech industries is not
spatial clustering phenomenon, and the larger the value, the stronger strong, and the regional driving effect is not evident.
the spatial correlation. When the value is in the interval [-1, 0], the (2)Results of local spatial autocorrelation analysis
attribute tends to be discrete; the smaller the value, the more pro- According to the LISA scatter diagram in Fig. 5, each province can
nounced the dispersion. When the value is 0, the attribute is ran- be divided into four different spatial correlation states. The first is the
domly distributed in space, and there is no clustering or dispersion. “H-H” agglomeration, showing that the technological innovation
capacity of the region as well as that of the neighboring regions is
high. This is the spatial correlation of high-level regions. The second
(1) Local spatial autocorrelation is the “L-L” agglomeration, showing that the technological innovation
capacity of the region as well as that of the neighboring regions is
low, a spatial correlation of low-level regions. Thirdly, the “L-H”
The Global Moran’s I was decomposed into individual cells to test agglomeration indicates that the region’s technological innovation
for the presence of localized agglomeration. The Local Moran’s I capacity is low. Nevertheless, the technological innovation capacity
reflects the agglomeration characteristics of the different areas. The of the neighboring regions is high and is catching up with develop-
LISA distribution was plotted using GeoDa software, containing four ment; thus, the spatial correlation is expressed as the development
distribution patterns: “High-High” (H-H) clustering, “Low-Low” (L- stage. The fourth is the “H-L” agglomeration, which indicates a high
L) clustering, “High-Low” (H-L) clustering, and “Low-High” (L-H) level of technological innovation capability in the region. However,
clustering. The formula for calculating the Local Moran’s I is shown the level of technological innovation capability in the neighboring
in Eq. (2). regions is low. Therefore, the increase in technological innovation
ðxi  xÞ Xm capability in the region has a countereffect on the level of technologi-
Local Moran0 s I ¼ 2
Wij ðxj  xÞ ð2Þ cal innovation capability in the neighboring regions; hence, the spa-
s i¼1
tial correlation indicates the inverse development stage.
6
W. Tu, L. Zhang, D. Sun et al. Journal of Innovation & Knowledge 8 (2023) 100287

Fig. 5. LISA scatter plot of technological innovation capacity in high-tech industries by region in 2010, 2013, 2016 and 2019.

Fig. 5 shows only three spatial correlation states: “H-H”, “L-L” and the region. Additionally, input-output and transformation guaran-
“L-H” agglomerations, and the main focus is on the “L-L” agglomera- tee capabilities are the main factors influencing technological
tion. The “L-L” agglomeration is predominant in the northeast and innovation capability.
southwest regions; the “H-H” agglomeration and the “L-H” agglomer- (2) Regarding spatial changes, the increase in high-tech industries’
ation are in the southeast region; and the agglomeration status of technological innovation capacity shows a trend of spreading
each region is relatively stable, with no evident changes. Fujian, from the southeast coast to the central and southwest regions.
located in southeastern China, is the only “H-H” agglomeration, while The high-level areas are mainly centralized in Zhejiang, Jiangsu,
Jiangxi, a neighboring province, is the only “L-H” agglomeration. This and Guangdong. The development differences between these
suggests that the southeast coastal region has a strong industrial three regions are mainly concentrated in the gap between techno-
base, transportation network coverage, and development of basic logical innovation input-output capacities. The transformation
industries. However, the desirable locational advantage is bound to guarantee capacity is more auxiliary to the effective realization of
provide a better external environment to support the development input and output. This shows that the improvement in technologi-
of high-tech industries’ technological innovation capacity. This spa- cal innovation capacity in high-tech industries lies in the effi-
tial agglomeration is characterized by the radiation of high-tech ciency of inputs and outputs, rather than relying solely on
industrial centers, such as the Shanghai High-Tech Zone and Shenz- external financial support to improve. Most regions have incom-
hen High-Tech Zone. The Northeast and Southwest regions are the plete echelon crossing and remain in the lower echelons. During
main distribution areas of the “L-L” agglomeration. Due to the large the study period, the technological innovation capability of high-
gap in economic development and industrial structure, these regions tech industries in China exhibits a relatively significant spatial
have not benefitted from the spillover and driving effects of the correlation, with the global Moran index being greater than 0.
regions with high technological innovation capacity, resulting in However, this correlation tends to weaken. The agglomeration
insufficient innovation power. status of each region is relatively stable, indicating that the devel-
opment of high-tech industries is related to the geographical loca-
Conclusions and recommendations tion and economic development level of each region.

Conclusions
Recommendations
This study evaluates the technological innovation capabilities of
high-tech industries in China’s 29 provinces and analyzes the evolu- Based on the above findings, the following countermeasures and
tion characteristics of their spatial patterns between 2010 and 2019. suggestions are proposed to promote the technological innovation
capability of high-tech industries and accelerate industrial develop-
(1) Overall, the development of technological innovation capability in ment toward the middle and high ends of the GVC.
high-tech industries in China for 2010 −2019 exhibited an
upward trend. Notably, the development was unbalanced in the (1) Strengthen effective investment in human and financial resources
eastern, central, and western regions. The eastern region exceeds and improve the mechanism for transforming technological
the national average, whereas the western region is developing achievements. The selected indicators are well represented in
slowly. However, there is an imbalance in the development within terms of human and financial resources, among the main
7
W. Tu, L. Zhang, D. Sun et al. Journal of Innovation & Knowledge 8 (2023) 100287

influencing factors. Therefore, we should strengthen human Supplementary materials

resources and financial investments to improve technology out-
put capacity and improve the mechanism for transforming scien- Supplementary material associated with this article can be found
tific and technological achievements to enhance the input-output in the online version at doi:10.1016/j.jik.2022.100287.
capacity. Local governments should provide sufficient human and
financial resources for innovation in high-tech industries. China
needs time to build science-based innovation capacity, even now
with large R&D investments. However, guiding high-tech enter-
References
prises to effectively connect with the capital market and attract
R&D investments from non-government sectors is necessary. Banwo, A. O., Du, J., & Onokala, U. (2017). The determinants of location specific choice:
Building a government-enterprise-market multi-party integration Small and medium-sized enterprises in developing countries. Journal of Global
platform to help high-tech enterprises quickly find resources and Entrepreneurship Research, 7(1), 1–17.
Braga, H., & Willmore, L. (1991). Technological imports and technological effort: An
improve the efficiency of technological innovation is crucial. It analysis of their determinants in Brazilian firms. The Journal of Industrial Economics,
cannot be ignored to consider the impact of standards on the effi- 39(6), 421–432.
ciency of techno-economic activity in high-tech industries (Tas- Chen, H., Lin, H., & Zou, W. (2020). Research on the regional differences and influencing
factors of the innovation efficiency of China’s high-tech industries: Based on a
sey, 2018). Further, it is vital to consider the U-shaped
shared inputs two-stage network DEA. Sustainability, 12, 3284.
relationship between intellectual property protection and techno- Chen, Y., Shi, H., Ma, J., & Shi, V. (2020). The spatial spillover effect in hi-tech industries:
logical innovation in high-tech industries (Liu et al., 2018). Empirical evidence from China. Sustainability, 12, 1551.
Freeman, C., & Soet, L. (1997). The economics of industrial innovation. Routledge.
(2) Strengthen industrial collaboration and technical exchanges to
Guo, J., Cui, L., Sun, S. L., & Zou, B. (2022). How to innovate continuously? Conceptualiz-
narrow inter-provincial differences. The spatial transfer of high- ing generative capability. Journal of Innovation & Knowledge, 7,(1) 100177.
tech industries can be implemented orderly to improve the tech- He, W., Lin, Z., & Hao, R. (2018). Analysis of regional competitiveness in the high-tech
nological innovation capacity in the central and western regions. industry. Current Science, 114(4), 854–860.
Jo, Y., Chung, W. Y., & Lee, D. (2020). The capability-enhancing role of government-
This requires the western region to provide complete supporting driven industrial districts for new technology-based firms in South Korea. Asia &
policies for the transferred high-tech enterprises. It is necessary to the Pacific Policy Studies, 7(3), 306–321.
use the industrial value chain linkage to promote scientific and Kraft, K. (1989). Market structure, firm characteristics and innovative activity. The Jour-
nal of Industrial Economics, 37(3), 329–336.
technological innovation resource flow between regions and form Lee, S., Park, G., Yoon, B., & Park, J. (2010). Open innovation in SMEs—An intermediated
an orderly and win-win spatial distribution pattern. Knowledge- network model. Research Policy, 39, 290–300.
sharing is the core of enhancing the technological innovation Liu, Z., Mu, R., Hu, S., Wang, L., & Wang, S. (2018). Intellectual property protection, tech-
nological innovation and enterprise value—An empirical study on panel data of 80
capability of enterprises (Saenz et al., 2009). Additionally, break-
advanced manufacturing SMEs. Cognitive Systems Research, 52, 741–746.
ing down local protectionism by establishing a regional develop- Lopez-Garcia, P., & Montero, J. M. (2012). Spillovers and absorptive capacity in the deci-
ment coordination mechanism to avoid duplication and the waste sion to innovate of Spanish firms: the role of human capital. Economics of Innova-
tion and New Technology, 21(7), 589–612.
of limited human and financial resources is necessary. Neverthe- Menaker, J., & Ozoliņa, V. (2018). Latvian high-tech industry: Trends and develop-
less, the relevant departments of the central government should ments. Economics and Business, 32, 160–171.
coordinate the layout of high-tech industries and provide prefer- Pellitero, M. M., Blanco, M. B., Heijs, J. J., & Baumert, T. (2008). A Novel way of mea-
suring regional systems of innovatios: factor analysis as a methodological
ential policies for the western region to attract foreign direct lisis Industrial y Financiero,
approach. Documentos de trabajo del Instituto de Ana
investment and domestic industrial transfer. 60, 1–47.Chicago.
(3) Promote high-quality development of high-tech industrial Saenz, J., Aramburu, N., & Rivera, O. (2009). Knowledge sharing and innovation perfor-
mance: A comparison between high-tech and low-tech companies. Journal of Intel-
agglomerations and enhance the complementary and spillover
lectual Capital, 10(1), 22–36.
effects of technological innovation. The construction of high-tech Scherer, F. M. (1965). Size of firm, oligopoly, and research: A comment. Canadian Jour-
industrial agglomerations should consider the geographical, his- nal of Economics and Political Science, 31, 256–266.
torical, and cultural differences between regions and the current Schumpeter, J. A. (2003). The theory of economic development. (pp. 61−116). Kluwer
Academic Publishers.
state of industrial development to maximize the spatial comple- Sumrit, D., & Anuntavoranich, P. (2012). Using DEMATEL method to analyze the causal
mentarity and spillover effects, enhancing the technological inno- relations on technological innovation capability evaluation factors in Thai technol-
vation capacity of China’s high-tech industry. However, we should ogy-based firms. International Transaction Journal of Engineering Management &
Applied Sciences & Technologies, 4, 81–103.
be wary of the inhibitory effect of industrial diffusion on coordi- Sun, X., & Sun, Q. (2018). Construction of the evaluation index system for innovation
nated regional economic development (Zheng et al., 2019). For ability of regional science and technology in China. Journal of Economics and Sus-
example, the Chengdu-Chongqing Economic Zone and Beibu Gulf tainable Development, 9(4), 1–10.
Shen, C., Li, C., & Si, Y. (2016). Spatio-temporal autocorrelation measures for nonsta-
Economic Zone represent high-tech industrial development in the tionary series: A new temporally detrended spatio-temporal Moran’s index. Phys-
western region. Therefore, these economic zones should be ics Letters A, 380(1-2), 106–116.
encouraged to use their unique resource, location, and policy Seddighi, H. R., & Mathew, S. (2020). Innovation and regional development via the
firm’s core competence: some recent evidence from North East England. Journal of
advantages to create an ecosystem of high-tech industries within
Innovation & Knowledge, 5(4), 219–227.
the economic zones. Tassey, G. (2018). Standards and expansion paths in high-tech industries. Handbook of
Innovation & Standards.
Wang, Y., Pan, J., Pei, R., Yi, B., & Yang, G. (2020). Assessing the technological innovation
Acknowledgments
efficiency of China’s high-tech industries with a two-stage network DEA approach.
Socio Economic Planning Sciences, 71, 100810.
The study is supported by the MOE (Ministry of Education in Yao, J., & Ma, H. (2017). Empirical study on the measurement of the technical efficiency
China) Project of Humanities and Social Sciences (18YJA630105), the of Jilin Province’s five high-tech industries. MATEC Web of Conferences, 108, 06006.
Yu, K., Gong, R., Hu, S., & Luo, Y. (2018). Based on analytic hierarchy process to discuss
National Social Science Fund of China (20BGL113; 21AZD067), and key factors in high-tech industrial ecology development. Ekoloji, 27(106), 449–454.
Key Research Base of Universities in Jiangsu Province for Philosophy Zheng, Y., Cheng, Y., & Li, L. (2019). Factors affecting regional economic synergy in
and Social Science ''Research Center for Green Development and China−Based on research on high-tech industry. IEEE Access, 8, 14123–14133.
Zhou, J., Wang, G., Lan, S., & Yang, C. (2017). Study on the innovation incubation ability
Environmental Governance'' is highly appreciated by authors of this Evaluation of high technology industry in China from the perspective of value-
study. Moreover, we gratefully acknowledge the editors and anony- chain−An empirical analysis based on 31 provinces. Procedia Manufacturing, 10,
mous referees for their helpful suggestions and corrections on our 1066–1076.
Zhu, J., Wang, Y., & Wang, C. (2019). A comparative study of the effects of different fac-
manuscript’s earlier version, according to which we improved the tors on firm technological innovation performance in different high-tech indus-
content. tries. Chinese Management Studies, 13(1), 2–25.