The Effect of Digitization on Enterprise Innovation

Full text

The Effect of Digitization on Enterprise Innovation
Xiaolong Han*
Central University of Finance and Economics Beijing, 102206, China
* Corresponding author: 2019312258@email.cufe.edu.cn
Abstract
This paper mainly studies the impact of enterprise digital transformation on technological innovation. Based on the
whole industry data of A-share listed enterprises, the paper analyzes the text of enterprise annual report and extracts
digital keywords to construct core explanatory variables, and takes cost expansion multiple as an intermediary
variable to explain the mechanism between digitalization and enterprise technological innovation. The regression
results show that: first, the digital transformation has a significant inverted U-shaped impact on the quantity and
quality of technological innovation input and innovation output. Second, before the inflection point, the marginal
benefit of digitalization is greater than the marginal cost, and the technological innovation of enterprises is promoted
under the effect of economies of scale. After the inflection point, the marginal benefit of digitalization is less than the
marginal cost, and the effect of diseconomies of scale makes the marginal input and output of technological
innovation of enterprises decrease. Thirdly, a large number of samples are located before the inflection point where
the marginal utility of the digitalization level is zero, which indicates that the vast majority of Enterprises in China are
still in an uphill phase in the promotion of digital transformation. Fourthly, the empirical results show that the impact
of enterprise digitization level on technological innovation has a positive decreasing effect over time, which can be
regarded as a long-term impact on enterprise fixed assets in the depreciation process.
Keywords-Digital transformation; Enterprise technology innovation; Cost expansion multiple
1.I
NTRODUCTION
The digital economy has become a key driving force
for the national economy these years[1]. As the previous
research shows, enterprise digital transformation is of
great importance in the context of the digital economy.
Besides that, it also has an important impact on
enterprise technological innovation[2]. Based on the
above research background, we can know the importance
of digital transformation for enterprises, which is a
common feature of the market economy in the future,
and enterprise innovation is highly necessary for a
national strategy and market competition. Although the
two are considered to be highly correlated in reality, the
specific impact of digital transformation on enterprise
innovation has not been comprehensively and deeply
studied in the academic circle. Therefore, this paper will
comprehensively and deeply explore the impact of digital
transformation on enterprise innovation, dig out the
specific logic and mechanism of digital transformation
on enterprise innovation, and provide theoretical, policy,
and practical suggestions for future policies and practices.
Moreover, the current academic circle analyzes digital
transformation and enterprise innovation either in a
specific enterprise or industry or at the municipal and
provincial level[3]. Therefore, it is limited by the
research field and scope, and cannot carry out follow-up
and expansion research well. In this paper, a-share listed
companies in the whole industry will be taken as the
research object, so there is A large space for
development in follow-up research and expansion, and
the research conclusions are more universally applicable.
2.
R
ESEARCH DESIGN AND PROGRESS
2.1. Research Design
2.1.1. Research hypothesis
Before proceeding with the empirical design, I
extensively read the literature and made the following
hypothesis:
Hypothesis 1: Digital transformation of enterprises
has a positive impact on technological innovation
Digital transformation brings the reduction of
transaction cost, communication cost and information
search cost of enterprises, improves the operation
© The Author(s) 2023
G. Vilas Bhau et al. (Eds.): MSEA 2022, ACSR 101, pp. 175–, 2023.
https://doi.org/10.2991/978-94-6463-042-8_27

efficiency of enterprises, promotes the quality and
efficiency of the whole process, and thus promotes the
technological innovation of enterprises[4].
Hypothesis 2: Digital transformation of enterprises
has an inverted U-shaped impact on technological
innovation
When digital transformation reaches a certain scale,
the marginal cost of digitalization is greater than the
marginal benefit, which increases the cost of enterprise
experience management, introduction and maintenance,
adaptation and training, etc., reduces the profit space of
enterprises, reduces innovation resources, and inhibits
enterprise technological innovation.
In order to verify the hypothesis, I adopted the
one-by-one regression method. First, the impact of the
primary term at the level of digitization on the
enterprise's technological innovation was not significant,
and then the quadratic term was added to obtain
significant regression results. Under this, model
construction and mechanism testing were carried out.
2.1.2. Econometric model
In order to explore the impact of digitalization level
on innovation, we set the model as follows:
ln󰇛Index󰇜Represents the innovation level,
which is measured from three aspects: R&D input, patent
application and invention patent application, respectively
representing innovation input, innovation output quantity
and innovation output quality. Represents the
digitalization level. Different indicators of enterprise
digitalization level are obtained by the text analysis
method, including the total word frequency of all
keywords, the total word frequency of frequent keywords
and the binary dummy variable constructed by the total
word frequency of keywords/the total word number of
the annual report. XRepresents the set of control
variables, including some enterprise microdata that
affects enterprise technological innovation but is not
completely correlated with core explanatory variables;
εIs the error term, which contains some disturbance that
cannot be observed.
2.1.3. Data source and pretreatment
The explained variables in this paper come from the
data required for modeling and empirical analysis from
The Tidal Information network, WIND database and
CSMAR database. The tide of information network is the
China Securities Regulatory Commission designated
website information disclosure of listed companies, is
China's first comprehensive deep Shanghai more than
2500 listed companies disclosure announcement of large
securities professional website information and market
data, this article on the website through web crawler for
a-share listed company's annual report data, is used to
construct the enterprise digital level; Obtain data of
patent applications and invention patent applications in
CSMAR database; Obtain enterprise R&D investment
and control variable data from WIND database. In the
process of data cleaning, ST, ST* and financial industry
enterprises were eliminated, and missing values and
incomplete information enterprises were eliminated.
Finally, 2,500 a-share listed companies' data were
obtained.
2.1.4. Variable construction
2.1.4.1. Explained variable: The logarithmic form of
enterprise innovation variable
index:𝑙𝑛󰇛𝐼𝑛𝑑𝑒𝑥

󰇜
In order to evaluate the impact of digitization on
innovation, this paper uses the data of the third period to
obtain the innovation indicators of enterprises from the
perspectives of R&D investment, patent application
number and invention patent application number by
referring to Liu Chunlin and Tian Ling (2021)[5],
explores the lag effect of digitization on innovation, and
uses the index to de-expand the data.
󰇛lnRD󰇜󰇛lnApplycation󰇜󰇛lninvent󰇜
2.1.4.2. Explanatory variable: Enterprise digitization
level:𝐼𝑛𝑑𝑒𝑥

In order to quantitatively evaluate the digitalization
level of enterprises, this paper conducts text analysis on
the annual reports of A-share listed companies,
constructs keywords to measure the digitalization level
and counts word frequency, so as to reflect the degree of
digitalization transformation of enterprises. The specific
steps are as follows:
A total of 159 possible keywords were given by
referring to relevant works of literature.
Search the annual reports of a-share listed companies
in the industry from 2014 to 2019 on Ju Chao
Information website;
Using keywords segmentation, text analysis statistics
of the word frequency of segmentation in each
company's annual report, statistics of the total word
frequency of keywords, defined as dig-level;
According to the association rules of word
segmentation, the frequent item set of keywords was
screened out and the keyword range was narrowed to
obtain 18 frequent keywords. The total word frequency
of frequent keywords was counted and defined as
SCR_diglevel.
Control variables set X contains a series of factors
that may affect the company's innovation level, including
X. Han
176

a fixed number of years of the company's business (Age),
the total assets of the company takes logarithm (TA), the
company take logarithm (FA) net value of fixed assets,
enterprise investment return and the ratio of total (ROA),
the enterprise asset ratio and the ratio of total assets
(DAR), Cash ratio (CR) derived from (monetary capital
+ marketable securities)/current liabilities, StateOwned,
Size.
Figure 1. Variable Definition
2.1.5. Endogenous treatment
Although the empirical results have shown that there
is a significant correlation between digital transformation
and enterprise innovation, further discussion is needed to
determine the causal effect of the two. It is worth noting
that although this paper has controlled many variables
that may affect enterprise innovation, such as enterprise
size, enterprise operating years, and total net asset value
of a company, there are still many other unobtainable
factors affecting enterprise innovation in reality, so there
may be a problem of missing variables in this study.
Measure error may arise at the same time, the data
obtained in this paper, enterprise level lead to false or
exaggerated the digitized transformation errors, the data
of the digital transformation of information disclosure is
insufficient or excessive make enterprise annual report
cannot accurately reflect the extent to which the
company digital transformation, keywords cannot reflect
comprehensively the digital transformation of enterprise,
etc. , can cause endogenous problems. Secondly, in
theory, enterprises' promotion of digital transformation,
The Effect of Digitization on Enterprise Innovation 177

such as digital management and digital deployment, can
promote the optimization and upgrading of corporate
governance structure, thus promoting enterprise
innovation, and the improvement of enterprise
innovation, in turn, will further promote enterprise digital
transformation, and there may be a reverse causality
relationship between the two.
In order to solve the endogeneity problem, this paper
adopts instrumental variable method. In the field of
economics, a good instrumental variable needs to satisfy
the conditions of externality and correlation. In terms of
correlation, the level of digitalization by province affects
enterprise innovation. In regions with higher
digitalization levels, enterprises can make better use of
digital information to promote digital transformation,
optimize corporate governance structure, promote
efficient allocation of resources, and improve enterprise
innovation. In exogenesis, the digitalization level of the
region is independent of other random disturbance terms.
Based on this, this paper uses the digitization level of
provinces as an instrumental variable. In terms of
specific indicators, a comprehensive index of
digitalization level is obtained by entropy weight method
using mobile phone penetration rate, Internet penetration
rate, number of Internet broadband access ports and
number of mobile Internet users. Data source: China
Economic Network database - provincial annual
database.
This paper will use two-stage least square regression
to test the rationality of instrumental variables in the
subsequent research. In the first-stage model, we will use
endogenous explanatory variables to perform OLS
regression on instrumental variable IV and other control
variables. The second-stage model was substituted into
the original OLS model with the fitting results of the first
step, so as to obtain the new fitting results and compare
them with the fitting results of the original OLS model.
3.
A
NALYSIS
3.1. Mechanism analysis
Based on the regression results, this paper analyzes
the causal effect of digitization level on technological
innovation, and explores the realistic logic and specific
mechanism of digitalization influencing innovation
behavior at the micro level of enterprises. Preliminary
regression analysis results show that there is an inverted
U-shaped relationship between digitization level and
technological innovation (including R&D investment
and invention patent application). Through literature
analysis and logical deduction, this paper further
explains the inverted U-shaped nonlinear relationship
with cost expansion multiple as a mediator variable. In
the regression results above, we use cost expansion
multiple as a mediator variable to test the inverted
U-shaped relationship between enterprise digitization
level and technological innovation. Combined with the
economic significance, it can be explained as follows:
3.1.1. There is a U-shaped relationship between
enterprise digitization level and cost expansion
multiple.
Before the inflection point, the digitalization level of
enterprises can reduce the cost expansion multiple, and
the marginal benefit of digitalization is greater than the
marginal cost of digitalization, thus producing economies
of scale effect and promoting enterprise innovation. First,
the transformation and upgrading of digital infrastructure
will improve the quality and efficiency of the whole
process of enterprises and reduce efficiency costs.
According to the resource-based theory and the theory of
IT capability, enterprise through the digital
transformation, can make full use of big data such as the
advantages of digital technology, play an important role
as a factor of production data, through close enterprise
internal and external resource sharing, information
exchange between the main body, mutual collaboration,
the whole process to improve the work efficiency and
reduce unnecessary costs. Second, digital transformation
sends positive signals to the outside world, increases the
availability of enterprise resources and reduces enterprise
communication costs. Based on the signal theory,
enterprises pay attention to promoting digital
transformation and related publicity, which is in line with
the government's policy orientation and easier to obtain
the government's orientation in policies, resources and
other aspects. At the same time, it sends a signal to the
market that the development prospect is sustainable and
the enterprise is trustworthy and capable, and it is easier
to win the trust of the market in the investment and
financing activities, thus it is more likely to provide more
support for its own innovation behavior. Digitization
improves the availability of government and market
resources, which can greatly reduce the communication
costs of enterprises. Third, increase the quality supply of
enterprises, improve user demand, reduce the cost of
supply and demand matching. On the supply side, the
key to the digital transformation of enterprises lies in the
formulation of enterprise development plans from the
perspective of the whole industrial chain, abandoning the
strong control and solidification mode, building digital
systems in an enabling, open and ecological
collaborative way, and truly focusing on market demand
to improve efficiency and product and service value from
a global perspective. Increasingly advanced digital
equipment and technology support the transformation of
business models between upstream and downstream
enterprises and between enterprises and channels, from a
simple contractual relationship to mutual benefit, joint
exploration of profit model, through the construction of 2
more efficient enterprise infrastructure and collaborative
model to promote quality supply; On the demand side,
X. Han
178

the new data capture the new demand of the market,
creating the possibility of enterprise innovation. With the
popularization of the Internet and the development of the
digital economy, the increasingly complex and
changeable market environment and the improvement of
information transparency at the consumer end have put
forward higher requirements for enterprises to efficiently
respond to consumer needs and provide quality services.
Digital technology includes big data analysis and
research, application of digital scene interaction,
real-time dynamic new data monitoring and rapid
capture of new market demand for enterprises.
Digitization brings efficiency to both supply and demand,
thus reducing the marketing cost for enterprises to
expand the market and attract users. Fourth, cultivate the
cultural innovation and personal efficiency of
high-quality employees, reduce the demand for
manpower and reduce the cost of enterprise manpower.
The construction of the core explanatory variables in this
paper, to a certain extent, can reflect the importance and
publicity of the enterprise to digitalization, and reflect
the level of digital culture construction of the enterprise.
Dare to innovate, demand-oriented as the core of digital
adapted to the digital construction of enterprise culture,
encourages employees to take the initiative to change,
grasp opportunities, trial and error in attempting to
promote service quality, and the effect product, helps to
cultivate the ability to accept new things, learn new
knowledge, inspire their exploring spirit, and, in turn,
promote the realization of technological innovation. At
the same time, the improvement of the personal ability
and efficiency of high-quality employees and the digital
replacement of part of the labor force can effectively
reduce the human cost of enterprises.
After the inflection point, enterprises'
over-investment in digitalization will lead to
diseconomies of scale, which will increase their cost
expansion multiple, reduce the efficiency of enterprises'
production and operation, and thus weaken their
innovation ability. First, the cost of technology
introduction is rising. Similar to the logic of technology
introduction between countries and regions, when an
enterprise's digitalization level is in a backward position
in the industry, it can learn from the experience and
lessons of other enterprises in the same industry in the
process of digitalization transformation through path
imitation, and realize the improvement of digitalization
level in a shorter time with higher efficiency and lower
cost. With the continuous acceleration of technology
iteration and update, enterprises based on higher level
digital technology have improved their digitalization
level, and the introduction of core digital technology will
face the explicit cost increase brought by stricter
technology patent protection. At the same time, the
reduction of reference cases in the industry makes
enterprises face greater risks and uncertainties, and
increases the potential cost of technology introduction.
Second, the equipment maintenance costs, as companies
introduced in the process of digital equipment,
technology is the nature of the fixed assets investment or
corresponding to a large number of the form a complete
set of investment in fixed assets, when digitized already
meet the demand of enterprise production and business
innovation, further investment and cannot ascend the
marginal output of enterprise innovation, it means idle
resources; Third, the cost of operation and management
is rising. The digitization of enterprises is highly
dependent on digital infrastructure equipment and digital
platform, while digital equipment and platform are
public goods. Due to the widespread governance
problems of public goods, with the increasing degree of
digitization of enterprises and the continuous growth of
digital equipment, the difficulty and cost of governance
will increase. The fourth is to adapt to the rising cost of
learning. Enterprise for the investment, the introduction
of digital hardware and software and services need to
learn to adapt to, high levels of the digital input increased
with employee training and difficulty of enterprise
management system integration, even if the cost of
adaptability than the digital equipment technology is
applied to improve the production and business operation
process to save the cost, improve enterprise digital
degree will gain weight instead burden, Therefore, the
continuous improvement of digitization level after the
inflection point has a negative impact on enterprise
efficiency.
3.1.2. Cost expansion multiple is negatively
correlated with technological innovation.
The larger the cost expansion multiple is, the weaker
the efficiency of the enterprise in each process of
production and operation. The efficiency capacity of an
enterprise is positively correlated with its innovation
capacity, and the specific mechanism is as follows: A
higher efficiency capacity of an enterprise means that it
is better at rationally allocating various production
factors including data and capital to improve the quality
and efficiency of the whole process and provide
economic support for its innovation behavior; It means to
perceive changes in the external environment and market
demand in a more timely manner and respond quickly to
create space and lead the direction for technological
innovation; It means more efficient realization of
resource sharing, knowledge exchange, contact and
cooperation among internal and external departments to
establish the organizational foundation for realizing
technological innovation, so as to promote enterprise
technological innovation.
In addition, the increase of cost expansion multiple
means that corporate profit margin is reduced. In quite
common cases, the overall growth rate of operating costs
of enterprises is faster than the overall growth rate of
turnover. To some extent, the overall growth of operating
The Effect of Digitization on Enterprise Innovation 179

costs will squeeze the profit margin, leading to the
decline of profits in the short term. Faced with a
short-term profit decrease, business operators are likely
to make corresponding adjustments, such as changing the
existing cost structure, which will reduce the existing
cost in many aspects, including innovation input (an
important part of the cost structure). In other words, the
reduction of profit margin will squeeze out innovation
input.
3.1.3. The relationship between enterprise
digitization level and technological innovation is
inverted U-shaped.
There is an optimal level of enterprise digitization
that is suitable for its industry attributes, scale and
self-development needs, and there are problems of
insufficient development and excessive investment in
enterprise digitization. Before reaching the inflection
point of optimal level, the improvement of digitization
level promotes the growth of enterprise technological
innovation, and after the inflection point, the continued
investment in digitization will weaken the enterprise's
technological innovation ability. There is an inverted
U-shaped relationship between digitization level and cost
expansion multiple, and the cost expansion multiple is
negatively correlated with enterprise technological
innovation. Enterprise efficiency capacity (cost
expansion multiple) plays a mediating role as a
mechanism variable. The relationship among the three is
shown in Figure 2.
Figure 2. Enterprise digital level, cost expansion
multiple and technological innovation relationship
This paper will further explore the operation
mechanism of the causality effect between enterprise
digitization and technological innovation by means of
sub-sample regression, searching for intermediary
variables and indirect transmission mechanisms.
3.2. Preliminary regression results analysis
The research selected the cross-section data of 2015
for preliminary research and empirical results to verify
the feasibility and value of the study.
3.2.1. Descriptive statistics
Descriptive statistics of the main variables are shown
in Figure 3. As can be seen from the table, the average
digitalization degree of all enterprises except financial
enterprises in 2015 is 33. 59, the maximum value is 792,
and the minimum value is 1. The mean value of the
square of digitalization degree in 2015 is 5247, the
maximum value is 627264, and the minimum value is 1.
In 2015, 2016 and 2017, the logarithmic average of r&d
investment of all enterprises except financial enterprises
was 17. 83, 17. 97 and 18. 17, the maximum value was
23. 68, 23. 59 and 23. 65, and the minimum value was 8.
795, 8. 454 and 7. 721, respectively. In 2015, 2016 and
2017, the logarithmic average value of patent
applications by all enterprises except financial
enterprises was 3. 487, 3. 612 and 3. 547, the maximum
value was 9. 779, 9. 909 and 9. 823, and the minimum
value was 0. 693, 0. 760 and 0. 715, respectively. In
2015, 2016 and 2017, the average value of inventions of
all enterprises except financial enterprises was 2. 964, 2.
987 and 2. 991, respectively. The maximum value was 8.
874, 9. 095 and 9. 108, and the minimum value was 0.
693, 0. 0052 and 0. 693, respectively.
Figure 3. Descriptive Statistics
3.2.2. Correlation coefficient matrix and scatter
diagram
The correlation coefficient matrix of independent and
dependent variables is shown in Figure 4. In Figure 4, it
can be seen that the correlation coefficient between the
digitalization degree index extracted after correlation
analysis and the untreated index is about 0. 95. At the
same time, the correlation coefficients between the
digitalization degree index and the logarithm of R&D
investment, the logarithm of the patent application and
the logarithm of invention patent are relatively small, but
these coefficients are significant when P >0. 05. This
indicates that although the coefficient is relatively small,
the value of the coefficient is significantly not 0.
X. Han
180

Figure 4. Correlation coefficient matrix
As shown in Figure 5, it can be found that the
correlation between R&D investment and digitalization
degree conforms to the inverted U-shaped curve, but the
correlation between patent application and invention
number and digitalization degree is not obvious.
Figure 5. Scatter Diagram
3.2.3. Fundamental regression
The return of the preliminary results show that the
explanation variable diglevel coefficient is positive, and
extremely significant, the squared coefficient is negative,
and is also very significant, we can according to this
result preliminary judgment, the digital level and
technology innovation is an inverted u-shaped
relationship, the inverted u-shaped on innovation input
and output quantity is significant, but the quality of
innovation output was not significant. The influence of
digitalization degree on r&d investment is positively
correlated before reaching the peak point. This is because
digital transformation is closely related to the
technological innovation of enterprises. The
improvement of the degree of digitalization will lead to
the increase in the efficiency of innovation research and
development of enterprises and promote technological
innovation. However, after the peak point, the
improvement of digitization degree is negatively
correlated with the r&d investment of enterprises. This is
because, in the process of digitization improvement,
enterprises improve the digitalization level on a large
scale, resulting in an excessive investment of enterprises,
resulting in an excessive cost of enterprises, reduced
R&D efficiency, and inhibiting enterprise innovation.
Figure 6. Preliminary Regression Result
3.3. Mediating effect model
Previous research discussed the influence of internal
micro management digital transformation on
input-output efficiency [6], and deeply analyzed the
nonlinear relationship between digital input and
efficiency. Their research on the nonlinear relationship
of variables provides the thinking support for this paper
to analyze the inverted U-shaped relationship between
enterprise digitization level and technological innovation.
The studies on the mediating mechanism in the existing
literature also provide useful inspiration for us to find the
key mediating changes and further expand the
mechanism analysis.
We construct the intermediary variable Cost_expand
expansion ratio = (based on the research and
development costs of enterprise operating cost plus a
divided by r&d and a logarithmic), it's meaning total
operating costs for the enterprise to eliminate r&d
section, in the sectors of production and management
processes can be on behalf of the enterprise efficiency,
cost expansion ratio, the smaller the stronger the ability
on behalf of the enterprise efficiency. The preliminary
regression results show that there is an inverted
U-shaped relationship between enterprise digitization
The Effect of Digitization on Enterprise Innovation 181

level and technological innovation, and the cost
expansion factor (enterprise efficiency capacity) plays a
mediating role.
In the regression of Figure 7, it can be seen that the
diglevel coefficient of the explanatory variable is
significantly negative, and the quadratic coefficient is
significantly negative at the level of 5%, and the
quadratic relationship is obvious, confirming that there is
a U-shaped relationship between enterprise digitization
level and cost expansion multiple.
In Figure 8, the coefficients before Cost_expand are
all negative, that is, cost expansion multiple is negatively
correlated with technological innovation. The smaller the
cost expansion multiple is, the stronger the efficiency
capability of the enterprise is, which means that the
enterprise is better at making full and effective use of
various production factors including information
resources, realizing resource sharing and close
cooperation between internal and external departments of
the enterprise, thus promoting the technological
innovation of the enterprise.
Figure 7. Intermediary effect: digitalization and cost
expansion
Figure 8. Intermediary Effect；Cost expansion and
Technology Innovation
3.4. Robustness test
Given that the influence of digital input lags in 2016
and 2017 corporate R&D, patent applications, and
invention patent applications at the enterprise digital
level, the results shown in Figure 9 indicate that the
enterprise's digital innovation input, innovation quantity,
and quality have at least three periods of sustainable and
significant influence, though the degree of influence
decreases as the number of years increases. The
digitalization level of an enterprise can be regarded as a
kind of fixed assets, which will have an impact on the
technological innovation of the enterprise in the
subsequent years after the investment. However, due to
the depreciation of fixed assets, the updating and
iteration of technical knowledge and other factors, the
degree of such impact will decrease.
Figure 9. Robustness: Digitalization and Sustainable
Impact
Considering the dispersion and marginality of 159
original digital keywords, this paper uses association
analysis to select the frequent item set of enterprise
digital transformation keywords, including 18 frequent
keywords such as digitalization, intelligent terminal,
software, information system and e-commerce, and
defines variable SCR_diglevel as the total number of
frequent keywords. Enterprise R&D investment, patent
application and invention patent application in period T
were used to regression the total number of frequent
keywords respectively, and the results were still
significant, as shown in Figure 10.
Figure 10. Robustness: Frequent Keyword regression
X. Han
182

Considering the different enterprise the length of the
annual report to the total number of words in the text
analysis, this paper use keywords to the total number of
words and keywords to the total number of words
frequently divided by the annual report to the total
number of words, and will be located in the binary digits
above 1, for the digital high hydration enterprise, located
in the binary digits and below 0, for low digital level of
enterprises, Get the virtual variables Diglevel12,
scr_digLevel12. The two dummy variables are replaced
by the core explanatory variables, and the results
obtained are shown in Figure 11, indicating that at the
significance level of 1%, enterprises with high
digitalization level will have higher innovation input and
high innovation output with both quantity and quality
compared with those with low digitalization level.
Figure 11. Double-Quartile Virtual Variable
Regression
4.
C
ONCLUSION
As the general trend of enterprise development,
digitization will be the necessary condition for enterprise
survival [7]. Under the innovation-driven strategy,
enterprise innovation will be the necessary
competitiveness for enterprise survival. In this topic, we
study the influence of digital transformation on
enterprise technological innovation and the deeper
mechanism behind it. Therefore, we chose to take
enterprise technological innovation as the explained
variable and enterprise digitization level as the core
explanatory variable. In the preliminary regression, 2015
was taken as the base period and after controlling the
control variables of 2015, the explained variables of
three periods (2015, 2016 and 2017) were regressed to
the explanatory variables of 2015.
The regression results show that: first, the digital
transformation of enterprises has a significant inverted
U-shaped impact on the quantity and quality of
technological innovation input and innovation output,
which is of great value for enterprises to carry out digital
transformation. Second, based on the empirical results of
regression and intermediate mechanism tests obtained
directly before the inflection point, digital transformation
can improve production efficiency, optimize resource
allocation, reduce enterprise cost expansion, thus
improving the input and output of technological
innovation. The inflection point, after the digital
transition, may lead to the cost of maintenance, the
introduction of high increase, and reduced efficiency.
With the rapid expansion of cost, the marginal revenue
brought by digitalization is less than the marginal cost,
which makes the input and output marginal of enterprise
technological innovation decrease. Thirdly, a large
number of samples are located before the inflection point
where the marginal utility of digitalization level reaches
zero, which indicates that the vast majority of Enterprises
in China are still in an uphill phase in the promotion of
digital transformation, and there is still a great space for
promotion in theoretical research, policy formulation and
implementation, enterprise practice and innovation.
Fourthly, the empirical results show that the impact of
enterprise digitization level on technological innovation
has a positive decreasing effect over time, which can be
regarded as a long-term impact on enterprise fixed assets
in the depreciation process.
R
EFERENCES
[1] Understanding the digital economy: data, tools, and
research[M]. MIT press, 2002.
[2] Zhu K, Dong S, Xu S X, et al. Innovation diffusion
in global contexts: determinants of post-adoption
digital transformation of European companies[J].
European journal of information systems, 2006,
15(6): 601-616.
[3] Buhaichuk O. Strategies of information and
innovation activity development at enterprise in
digital conditions[J]. Humanities Studies, 2019 (1
(78)): 75-85.
[4] Gobble M A M. Digitalization, digitization, and
innovation[J]. Research-Technology Management,
2018, 61(4): 56-59.
[5] Liu Chunlin, TIAN Ling. "Endorsement" can
promote enterprise innovation talent policy [J].
China industrial economy, 2021 (03): 156-173. The
DOI: 10. 19581 / j. carol carroll nki ciejournal.
2021. 03. 009.
[6] Tabrizi B, Lam E, Girard K, et al. Digital
transformation is not about technology[J]. Harvard
business review, 2019, 13(March): 1-6.
[7] Guo H, Yang Z, Huang R, et al. The digitalization
and public crisis responses of small and medium
enterprises: Implications from a COVID-19
survey[J]. Frontiers of Business Research in China,
2020, 14(1): 1-25.
The Effect of Digitization on Enterprise Innovation 183

Open Access This chapter is licensed under the terms of the Creative Commons Attribution-NonCommercial 4.0 International License (http://
creativecommons.org/licenses/by-nc/4.0/), which permits any noncommercial use, sharing, adaptation, distribution and reproduction in any
medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons
license and indicate if changes were made.
The images or other third party material in this chapter are included in the chapter’s Creative Commons license, unless indicated otherwise
in a credit line to the material. If material is not included in the chapter’s Creative Commons license and your intended use is not permitted
by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder.
X. Han
184