Government subsidies and green innovation in chinese enterprises based on the synergy of executive incentives

Asian Journal of Technology Innovation

ISSN: (Print) (Online) Journal homepage: www.tandfonline.com/journals/rajt20

Government subsidies and green innovation in
Chinese enterprises-based on the synergy of
executive incentives

Hu Liu, Xiaoxuan Yu & Yijun Peng

To cite this article: Hu Liu, Xiaoxuan Yu & Yijun Peng (2023) Government subsidies and green
innovation in Chinese enterprises-based on the synergy of executive incentives, Asian Journal
of Technology Innovation, 31:3, 534-555, DOI: 10.1080/19761597.2022.2131586
To link to this article: />
Published online: 10 Oct 2022.
Submit your article to this journal
Article views: 254
View related articles
View Crossmark data
Citing articles: 2 View citing articles

Full Terms & Conditions of access and use can be found at
/>
ASIAN JOURNAL OF TECHNOLOGY INNOVATION
2023, VOL. 31, NO. 3, 534–555
/>
Government subsidies and green innovation in Chinese
enterprises-based on the synergy of executive incentives

Hu Liu, Xiaoxuan Yu and Yijun Peng

International Business School, Shaanxi Normal University, Xi’an, People’s Republic of China

ABSTRACT KEYWORDS
Government subsidies; green
Based on the micro data of China’s heavily polluting enterprises innovation; China’s heavy
from 2010 to 2019, this article analyzes the synergistic effect of pollution enterprise;
government subsidies and executive incentives on green compensation incentive;
innovation. It is found that government subsidies significantly equity incentive
promote the output of innovation results, and this effect has a
time lag. Executive compensation incentives will weaken the
current policy effect, and the benefit convergence effect
produced by equity incentives can reverse executives’ short-
sighted thinking. Although the current impact of equity
incentives is not significant, it can play a substantial synergistic
effect in the first and second lag periods. It is proved that the
green innovation of heavy pollution enterprises can not only

consider the impact of exogenous policy but ignore the

perfection of internal governance mechanism.

1. Introduction and literature review

According to the ‘2021 Interim Report on the State of the Global Climate’ released by the
World Meteorological Organization (WMO), the frequency and intensity of global
climate extremes are rising rapidly, which may form a ‘dangerous compound effect’
with the impact of the epidemic and economic recession. The problem of environmental
degradation deserves the vigilance of all countries in the world. After more than 40 years
of rapid development of reform and opening up, China’s environmental carrying
capacity is gradually approaching the upper limit, and ecological governance is immi-

nent. China’s 2060 carbon neutrality strategy emphasises the importance of technology
in improving climate and environmental change, and clarifies that emission reduction
needs to shift from end-to-end governance to source control, and the growth model
from factor-driven to innovation-driven. Therefore, green innovation has received a
lot of attention from Chinese academia. Regarding green innovation, there is currently
no generally accepted definition in academia. Early scholars mostly believed that green
innovation is a new technology or new product with the premise of protecting the eco-
logical environment and the pursuit of economic benefits. With the extension and devel-
opment of the concept of green innovation, green management innovation (such as
green marketing, green supply chain, etc.) is gradually included in the category of

CONTACT Hu Liu Shaanxi Normal University, International Business School, 620 West
Chang’an Street, Xi’an 710062, People’s Republic of China
© KOSIME, ASIALICS, STEPI 2022

ASIAN JOURNAL OF TECHNOLOGY INNOVATION 535

green innovation. However, considering the research purpose of this article, the green
innovation referred to in this article is more inclined to green technology innovation.
Green equipment innovation, green product innovation, and green production material
innovation all belong to the category of green technology innovation. Compared with
general innovation, green innovation pays more attention to environmental benefits,
that is, whether the damage to the environment can be reduced or avoided.

From the perspective of life cycle, green innovation should include the whole process
from idea formation to market launch. Green R&D is the capital invested for green inno-
vation activities, and it is the beginning of enterprise green innovation activities. Accord-
ing to Hamamoto’s (2006) research and combined with the data of the ‘China Science
and Technology Statistical Yearbook’, the green R&D investment in China from 2000
to 2019 can be calculated, as shown in Figure 1. The green R&D investment generally

shows an upward trend, with an average annual growth rate of 15.27%. The annual
growth rate of general R&D investment is 18.41%, which is slightly higher than that of
green R&D investment, but there is a large order of magnitude difference between the
two. At this stage, China’s green R&D investment seriously underinvested. Although
green innovation is considered to be the first driving force to lead high-quality economic
development and enterprise transformation (Guo et al., 2019a), the dual externalities of
green innovation and the profit-seeking nature of capital lead to the fact that enterprises
tend to prefer economies of scale and short-term economic benefits while ignoring the
environment protection in the process of development, lack of enthusiasm and initiative
to increase green R&D investment. In addition, the lack of scientific research personnel,
technology and funds also hinders the improvement of enterprises’ green governance
and innovation capabilities. Since this article pays more attention to the results of enter-
prise green innovation activities and the feasibility of measurement indicators, the
number of green patents is selected to reflect the level of green innovation in Figure 2.
The reason for not using green R&D is that the greater the number of green R&D
does not mean the more green patents. According to statistics, the average annual
growth rate of green patents in China from 2000 to 2019 is 24.07%, which is higher

Figure 1. Trend chart of R&D investment for green innovation in China.

536 H. LIU ET AL.

Figure 2. Trend chart of green innovation and other types of innovations in China.

than that of common patents. The number of green patents in 2019 is about 60 times that
of 2000. However, there is still a big gap with leading countries such as the United States,
Japan and Germany.

To solve environmental problems, China has formed a governance system led by the
government, dominated by enterprises, and participated by social organisations and the

public. The government has been revising and promulgating laws, regulations, and policy
measures for a long time, so as to incorporate environmental factors into the production
function of enterprises, and encourage or force green transformation of enterprises. Gov-
ernment subsidies and environmental taxes are the two main means of current govern-
ment intervention. See Table 1 for details.

In order to alleviate the financing constraints faced by enterprises and reduce the
negative impact of ‘market failure’, the Chinese central government and local govern-
ments have widely adopted fiscal and taxation policies to intervene in innovation entities.

Table 1. Comparative analysis table of government subsidies and environmental taxes.

Type Government subsidies Environmental tax-independent taxes in China

Mechanism Resource Supplement Punitive pushback
Responsible Finance Department Tax Department

Department Wide range of objects Enterprises, institutions and other producers and
Subsidy object operators that directly discharge taxable
There is information asymmetry between the pollutants into the environment
Challenge government and enterprises, and it is difficult
for the government to directly supervise the Under the ever-increasing environmental
use of government subsidy funds by protection supervision, in recent years, there
enterprises, and it is even difficult to determine has been a ‘one-size-fits-all’ approach in
whether the enterprises that receive environmental supervision and law
government subsidies really have the enforcement, such as companies shutting
corresponding qualifications. down polluting production activities to deal
with the behaviour of the inspectors.

ASIAN JOURNAL OF TECHNOLOGY INNOVATION 537

Government subsidies are free funds allocated by governments at all levels to enterprises.
They are an ex-ante incentive for enterprises, and their smoothing effect on innovation
has also been questioned by many academic circles. Scholars of the incentive view believe
that the signalling effect of government subsidies makes enterprises more favoured by the
capital market (Xia & He, 2020), easy to obtain external input, reduce financing con-
straints (Li et al., 2019; Montmartin & Herrera, 2015), and higher expected returns
enhance their innovation willingness (Yu et al., 2021a); secondly, the ‘leverage effect’
hypothesis believes that the various review and evaluation mechanisms formulated by
the government around the subsidy policy are also conducive to improving the R&D
investment and economic performance of enterprises, and alleviating the opportunism
of enterprises in innovation incentives (Ding & Xie, 2021; Yu et al., 2021b); furthermore,
the government allocates the risk of innovation failure through R&D funds, affects the
allocation of innovation resources, and then promotes R&D innovation (Du & Zhang,
2020; Hu & Deng, 2019). Scholars who hold the inhibition view believe that the horizon-
tal competition of local governments can easily lead to enterprise policy arbitrage, which
leads to policy failure and waste of public resources. The serious ‘patent bubble’ problem
also makes government subsidies and enterprise innovation choices fall into the ‘prison-
er’s dilemma’ (Hu & Jin, 2021). The influx of a large number of government subsidies will
also reduce the risk-taking spirit of entrepreneurs and inhibit innovation performance.
There are also differences in the investment goals of the government and enterprises,
and the funds used for green R&D activities may be misappropriated or directly
crowded out of enterprises self-owned R&D funds (Liu et al., 2019). Different from
the resources allocated by other market mechanisms, subsidies lack corresponding
value appeals and are ‘inactive’ in innovation activities (Jourdan & Kivleniece, 2017).
From the perspective of market supply, the increase in financial subsidies may also
lead to an increase in resource prices, a decline in marginal benefits, and an adverse
impact on innovation output. The promotion and inhibition of government subsidies
and green innovation may occur successively (Huang et al., 2016; Wang & Wang,
2020b; Wu & Zhang, 2021; Zhang, 2020), property rights (Jin et al., 2018), scale (Pere,

2013), type of government subsidies (Lou et al., 2021), system quality (Bianchini et al.,
2019), the policy attention of enterprises will also cause differences in research results
(Fu & Gao, 2021).

In fact, the output of green innovation not only depends on the resource endowment
of enterprises, but also is affected by differences in internal governance of enterprises.
From the principle of the dialectical relationship between internal and external causes,
we can see that the internal cause is the fundamental cause of the development of
things, the external cause is the condition of change, which acts through the internal
cause. Therefore, only relying on external policies cannot fundamentally improve the
green innovation performance of enterprises and promote the ‘greening’ process of
enterprises. In recent years, scholars have also begun to pay attention to the impact of
internal governance differences on green innovation. For example, the shareholding of
state-owned enterprises can promote the green governance of private heavily polluting
enterprises (Wang et al., 2022), and non-state-owned equity participation in state-
owned enterprises also plays a positive role (Zhao et al., 2022); the stronger the execu-
tives’ dual environmental awareness, the better at identifying and grasping the market
opportunities brought by green innovation, they also actively reflect on the shortcomings

538 H. LIU ET AL.

of the enterprise’s own green development, and implement a green innovation strategy
(Xi & Zhao, 2022); influenced by morality and ethics, green investors are willing to
reduce interest demands, which can prompt enterprises to implement green actions,
increase green expenditures, and improve green governance performance (Jiang et al.,
2021). In modern enterprises, senior executives hold a large part of the management
rights of the enterprise and become the makers and implementers of enterprise
decision-making. In the case of information asymmetry, the existence of agency pro-
blems makes executives more likely to focus on arbitrage for personal gain when
making decisions. Risk aversion, high innovation failure costs, and lack of innovation

compensation mechanisms further weaken executives’ green innovation tendency. Com-
pared with stimulating innovation, stimulating green innovation is a more challenging
proposition. How to use the executive incentive mechanism to enable enterprise man-
agers to actively carry out green innovation activities, improve the policy effect of govern-
ment subsidies, and realise the ‘green’, ‘ecological’ and ‘innovative’ of enterprise, has both
theoretical and practical significance.

Therefore, this article takes heavily polluting enterprises as samples to explore the
relationship between government subsidies and green innovation and the moderating
effect of executive incentives on the two. It may enrich and deepen the existing research
in the following aspects: (1) The empirical research conclusions on the effect of government
subsidies are quite different, and the existing literature on government subsidies and enter-
prise green innovation is still relatively small, the effect of government subsidies in heavily
polluting enterprises needs more evidence to support. (2) ‘The Suggestions of the Central
Committee of the Communist Party of China on Formulating the Fourteenth Five-Year
Plan for National Economic and Social Development and the Long-term Goals for 2035’
pointed out that the green transformation of key industries and key areas should be pro-
moted in the future. Therefore, this article takes the heavily polluting enterprises as the
research object, and considers the macro-policy perspective (government subsidies) and
the micro-perspective of enterprises (executive incentives) to study the green innovation
of enterprises, which can provide some enlightenment for how to promote the construction
of ecological civilisation in China by improving the coordination of internal and external
mechanisms. (3) This article analyzes the effect of government subsidies under different
executive incentive methods, which will help relevant government departments to effectively
identify enterprises before subsidies are issued, so as to reduce supervision costs and trans-
action costs, and reduce the possibility of enterprises breaking the invisible contract.

2. Theoretical analysis and research hypothesis

2.1. Government subsidies and green innovation

Dual externalities are the main reason for the low enthusiasm for green innovation in the
current heavily polluting enterprises. From the externality theory, it can be known that in
economic activities, if enterprises cannot enjoy all the benefits of decision-making or
needs to bear all the costs of decision-making, externalities will occur. Knowledge and
technology have positive externalities, the green innovation achievements of heavily pol-
luting enterprises are also prone to spillover, leading to the widespread phenomenon of
‘free riders’ in the industry. If competitors quickly imitate and reproduce at low cost, it

ASIAN JOURNAL OF TECHNOLOGY INNOVATION 539

will greatly shorten the time for the enterprise to enjoy the technological advantages,
market position and high economic returns brought by the green innovation achieve-
ments, and even cannot make up for the enterprise’s early innovation investment,
which greatly dampen the enterprise’s enthusiasm for green innovation. This phenom-
enon will become more prominent when the mechanism is relatively weak. The negative
externality of environmental pollution means that the pollution costs incurred by enter-
prises in the production and operation process are not fully borne by the enterprises
themselves. Without being punished, enterprises lack the initiative to adopt green tech-
nologies. It can be deduced from this that without corresponding policy incentives and
institutional constraints, the development and use of green technologies will be uneco-
nomical for heavily polluting enterprises that are currently rational economies. In
addition, innovation activities require stable financial support, but the endogenous
financing of heavily polluting enterprises is generally difficult to meet innovation
needs (Cao et al., 2021). China’s financial market is also seriously lagging behind indus-
trial development, and the allocation of market resources has failed. Therefore, govern-
ment departments need to effectively intervene in the innovation activities of enterprises
by using government subsidies as an innovative resource supplement mechanism.

Based on the resource-based view, government subsidies, as an ‘ex-ante incentive’

measure, can provide enterprises with certain resources and directly alleviate the
financing constraints of enterprises in a way of nearly zero financing costs, which is con-
ducive to enhancing the green innovation ability and willingness of enterprises to inno-
vate, leading them to increase the intensity of innovation investment, manage R&D
activities in a refined manner, and accelerate the innovation process. From the perspec-
tive of ‘signaling’ and ‘authentication effect’, government subsidies can reduce the nega-
tive effects of information asymmetry. In order to maintain a competitive advantage and
prevent the leakage of core secrets, enterprises will strictly control the leakage of infor-
mation before the final results are formed. The complexity and opacity of the innovation
process make it impossible for outside investors to judge the expected return of project
investment, resulting in increased financing constraints (Hall & Lerner, 2010). Govern-
ment subsidy is equivalent to sending a signal to the market as an intermediary, that is,
the enterprise is recognised and focused by the government, which is conducive to
enhancing investor confidence, expanding enterprise financing channels, ensuring the
necessary capital investment for green innovation activities, reducing the marginal
cost of R&D activities, and preventing the crowding of normal operating cash flow.
From the perspective of cost and risk sharing, local governments use government subsi-
dies to provide financial support for R&D activities, and share the risk of innovation
failure with enterprises. Enterprises in the initial stage of transformation may need to
do some preparations for R&D activities, such as purchasing certain new equipment,
hiring scientific researchers, and forming R&D teams. Government subsidies are used
to reduce the sunk costs of enterprises and ease the pressure on capital flow. In the
process of ‘learning by learning’ and ‘learning by doing’, the knowledge reserve of enter-
prises increases dynamically and improves the possibility of successful innovation activi-
ties in the future. For enterprises that have already carried out R&D activities,
government subsidies can reduce the marginal cost of enterprises, encourage more
R&D investment, compensate for losses caused by technology spillovers, bridge the
gap between private benefits and social benefits. At the same time, it helps enterprises

540 H. LIU ET AL.

share the risks of innovation activities and reduces the probability of bankruptcy of enter-
prises after innovation fails. In addition, the government will strengthen the regulation
and supervision of enterprises that receive subsidies, further improving the efficiency
of innovation output of enterprises (Wang et al., 2017). Based on this, the first research
hypothesis of this article is put forward:

H1: Government subsidy plays a positive role in promoting green innovation of heavy pol-
lution enterprises.

2.2. Moderating effects of executive incentive

The incentive mechanism of executives stems from the first type of principal-agent
problem. The income function of enterprise executives and shareholders is not consist-
ent. Shareholders can diversify their shares to diversify risks and obtain benefits. The
wealth accumulation and career prospects of senior management depend on the
success or failure of the enterprise. If the incentive contract cannot provide enough
incentives to offset the negative impact of the failure of green innovation activities, execu-
tives may be ‘lazy’ and ‘inaction’ in innovation activities due to career concerns. Because
the value-creating effect of green innovation is difficult to translate into observable per-
formance of enterprises in the short term, executives’ ‘short-sightedness’ and lack of
awareness of green innovation will also affect enterprises’ decision-making choices and
investment rankings. How to implement effective incentives for senior management
has always been a hot topic in the theoretical and practical circles. If the internal incentive
mechanism of the enterprise is ineffective, the implementation and operation costs of
government subsidies will be increased, and it is difficult to achieve the best ‘greening’
effect. Innovation may even become a tool for executives to capture resources and
pursue self-interest (Tong et al., 2014). According to existing research, executive incen-
tives are mainly divided into compensation incentives and equity incentives. The adjust-
ment effects of different incentive methods are as follows.

2.2.1. The moderating effect of compensation incentives
The executive compensation incentive mechanism is to establish executive compensation
on the basis of certain performance standards. At present, Chinese executives are in the
stage of wealth accumulation. Compensation incentives link executives’ monetary
returns with the short-term interests of the enterprise, and are believed to satisfy execu-
tives’ wealth needs and curb their risk aversion tendencies, so that executives can perform
their duties diligently, make scientific and effective business decisions (Wang & Wang,
2020a), stimulate enthusiasm for work, and concentrate on strengthening the manage-
ment of R&D activities of enterprises. Ultimately, the efficiency of using government sub-
sidies is improved, while the ‘crowding out’ effect on the original R&D funds of
enterprises is weakened, and the probability of successful green innovation activities is
increased. Although green innovation activities are beneficial to enterprises, it has
large investment scale, long payback period and high risk, and cannot generate cash
flow during the tenure of senior executives, which may have a negative impact on com-
pensation incentives based on short-term performance. The preference of executives for
short-term benefits such as salary and bonuses will reduce their risk tolerance, and even

ASIAN JOURNAL OF TECHNOLOGY INNOVATION 541

stick to conservative investment decisions in order to maintain the existing monetary
income, and the development of core technologies will be locked back to low-end orien-
tation. Secondly, green innovation activities will increase the current cost of enterprises
and reduce investment in productive and profitable projects. When business perform-
ance declines, stakeholders such as shareholders will question the operational capabilities
of executives, and the career prospects and personal reputation of the executives will also
be damaged. As a result, executives may respond negatively to the green innovation
activities of the company. In the absence of strong support from senior executives, the
internal green innovation power of enterprises is insufficient, project operations lack
planning, and innovation efficiency is low. Therefore, the role of government subsidies

in promoting green innovation in enterprises is weakened. Accordingly, the second
research hypothesis of this article is put forward:

H2: The executive compensation incentive policy plays a significant negative role in regulat-
ing the relationship between government subsidy and green innovation.

2.2.2. The moderating effect of equity incentives
It is generally believed that equity incentives enable executives to obtain the distribution
rights of residual income through shareholding, resulting in a convergence effect of inter-
ests. The long-term value of the enterprise and the personal value of the executives are
unified, which enhances the sense of protagonist of the executives, encourages the execu-
tives to reduce adverse choices in strategic decision-making, and make strategic decisions
that are in line with the sustainable development of the enterprise. Due to information
asymmetry, it is difficult for the government subsidy support mechanism to rely on exter-
nal orders to carry out orderly development. The green innovation funds given to enter-
prises by government departments may be misappropriated to other projects, or have a
‘crowding out’ effect on the enterprise’s own R&D investment. Even if there is external
supervision, the use of subsidy funds and the management of green innovation activities
are difficult to achieve an effective state, which increases the cost of supervision and
reduces the efficiency of government subsidies. When there is a reasonable equity incen-
tive contract within the enterprise, it will increase the executives’ tolerance for short-term
green innovation failures, and give incentive objects generous remuneration in the long
run, so that executives have strong green innovation motivation and strong willingness to
green governance. Not only government subsidies are used for green innovation activi-
ties, but additional R&D investment will be added. Therefore, under the synergy of equity
incentives, the positive effect of government subsidies on green innovation of enterprises
is magnified. In recent years, environmental policies have become increasingly strict, and
the new Environmental Protection Law has introduced information disclosure mechan-
isms for public supervision, which has improved the transparency of enterprise environ-
mental information and significantly reduced the opportunistic behaviour of enterprises

to conceal environmental information (Wang et al., 2020). Local governments have also
tightened penalties for environmental violations. Because the frequent exposure of enter-
prise violation information will accelerate the depreciation of executives’ human capital,
executives can make strategic choices to maintain enterprise reputation out of risk aver-
sion considerations. At the same time, the accumulation of negative environmental news
has led to a rising risk of stock price collapse (Yu & Bi, 2021). In order to increase stock

542 H. LIU ET AL.

Figure 3. Theoretical hypothetical framework.

prices to achieve asset preservation and appreciation, executives will be more proactive in
carrying out green innovation activities. Therefore, the third research hypothesis of this
article is proposed:

H3: Executive equity incentive policy plays a significant positive role in regulating the
relationship between government subsidy and green innovation.

Based on the above theoretical analysis and research hypotheses, a research frame of
the relationship is constructed in Figure 3.

3. Research and design

3.1. Sample selection and data source

This article selects 2010–2019 A-share listed companies in China’s Shanghai and Shenzhen
stock markets with heavy pollution industries as the research object, and the heavy pol-
lution enterprises are defined according to the sixteen subdivided industries such as
thermal power, iron and steel involved in the Guide to Environmental Information Disclos-
ure of Listed Companies issued by the Ministry of Environmental Protection in 2010.

According to the division standard of the Guidance of Industry Classification of Listed
Companies issued by the CSRC in 2001, the corresponding secondary industries are com-
pared, and then whether the listed companies have the attribute of heavy pollution industry
is judged. In order to ensure the quality of the research data, the article further screened the
samples as follows: (1) Exclude enterprise samples that were ST and *ST during the sample
period; (2) Exclude the enterprise samples listed in 2011 and later; (3) Exclude enterprise
sample with missing financial data. After the above treatment, this article finally obtained
536 heavy pollution listed enterprises 5360 balance panel data. The green patent application
data of listed companies comes from CNRDS, and other enterprise governance and
financial data come from CSMAR. At the same time, in order to avoid the influence of out-
liers on the research conclusions, this article conducts Winsorize at the upper and lower 1%
level for all continuous variables. The data was processed using Excel2010 and Stata16.0.

3.2. Variable design

3.2.1. Dependent variable
Green innovation activities of enterprises include green innovation input and green inno-
vation output. Most researches on green innovation use energy consumption or new

ASIAN JOURNAL OF TECHNOLOGY INNOVATION 543

product output value to measure. However, this method cannot accurately meet the need for
research on the level of green innovation at the microcosmic individual level of enterprises
(Xiao et al., 2021). The cost of green R&D input has not been measured and disclosed in the
financial report, and it is difficult to separate from the R&D input of enterprises. Therefore,
this article selects the number of green patent applications to represent the green innovation
capability of enterprises. In the process of data collation, this article can distinguish green
patents and non-green patents by checking whether keywords such as ‘energy saving’ and
‘emission reduction’ appear in documents such as corporate annual reports and patent appli-
cation materials. The reason for not choosing the number of green patent grants is that cur-

rently a patent usually takes 1–2 years from application to grant, and it is easily affected by
external interference during the granting process, which is unstable. In comparison, the
enterprise green patent application process has already indicated that the enterprise is carry-
ing out innovation activities. So, the number of patent applications will more timely and
reliably reflect the real green innovation level of enterprises than the number of patent
grants. This article uses the sum of the current green invention patents and green utility
patents of enterprises to measure green innovation, and adds 1 to the number of green
patent applications in the regression model to take the logarithm.

3.2.2. Independent and moderator variables

(1) Independent variables. In this article, the total amount of government subsidy disclosed
in the annual report of listed companies is used to measure the subsidy intensity.

(2) Moderator variables. For compensation incentive, this article draws on the method
of Guo et al. (2019b), and uses the ratio of the total compensation of the top three
executives to the total compensation of all executives. Due to the generally low share-
holding level of executives in some listed companies, the total number of shares held
by executives is added by 1 and the natural logarithm is taken to measure equity
incentives (Yin et al., 2018).

3.2.3. Control variables
(Table 2).

Table 2. Variable definition table.

Type Name Symbol Definition
Gpatent
Dependent Green innovation LN(1 + The number of green patent applications in the
variable Subsidy current period)

Government subsidies
Independent Payinc LN(1+ Total amount of government subsidies for the current
variable Compensation incentive period)
Shainc
Moderator Equity incentive Roa Top three executive payroll/Executive payroll
variable Rate of return on total
RD LN(1 + Total number of shares held by executives)
Control variable assets Age Net profit/Total asset balance
R&D input
Business Age Soe R&D input/Operating income
The number of years from the enterprise’s founding period
Nature of property right TobinQ
to the sample period
Ability to grow Virtual variable: 1 for state-owned enterprises and 0 for non-

state-owned enterprises
(Stock market value + Net debt)/Total assets

544 H. LIU ET AL.

3.3. Model design

In order to test the relationship between government subsidy and enterprise green inno-
vation, and the moderating effect of executive incentive on the relationship, under the
condition of controlling the main influencing factors of enterprise green innovation,
the following regression model is constructed in sequence:

Gptentit = a0 + a1Subsidyit + a2Xit + mi + gt + 1it (1)

Gptentit = b0 + b1Subsidyit + b2Subsidyit × Payincit + b3Payincit

+b4Xit + mi + gt + 1it

Gptentit = l0 + l1Subsidyit + l2Subsidyit × Shaincit + l3Shaincit
+l4Xit + mi + gt + 1it

Among them, subscript i represents enterprise, t represents year; Gpatent represents
the enterprise green innovation variable, Subsidy represents the government subsidy
variable, Payinc and Shainc represent executive compensation incentives and equity
incentive variables, respectively. X are a set of control variables, μ, γ, and ε represent
firm fixed effects, year fixed effects, and stochastic disturbance term. The key parameters
of this article are α1, β2 and λ2, which reflect the specific impact of government subsidies
on green innovation and the moderating role of executive incentive.

4. Empirical results

4.1. Descriptive statistics

Descriptive statistics of the primary variables for this article are presented in Table 3. The
average value of green innovation is 3.100, the standard deviation is 9.500, the minimum
value and the median value are zero, indicating that more than half of the sample enter-
prises do not carry out substantial green innovation activities, resulting in a generally low
green patent output level and a large degree of dispersion. The maximum value of gov-
ernment subsidy is 849741992 RMB (about US$126,016,737.41), the minimum value is
zero, the mean value and median value are 49273866 RMB (about US$7,307,314.33)
and 13028658 RMB (about US$1,932,149.98) respectively, which indicates that the
subsidy amount obtained by different sample enterprises varies greatly. The minimum
value of executive compensation incentive is 0.22, the maximum value is 0.89, and the
standard deviation is 13.600. The minimum and median of senior executives’ equity

Table 3. Descriptive statistics of main variables.

Variable name Sample size Mean Median Standard deviation Min. Max.

Green innovation 5360 3.100 0 9.500 0 72
13028658 118339666 0 849741992
Government subsidies 5360 49273866 22
44.300 13.600 0 89
Compensation incentive 5360 46.300 0 66613486 −0.238 404612837
0
Equity incentive 5360 24303059 0.037 0.065 4.170 0.227
0.010 0.022 0 0.107
Rate of return on total assets 5360 0.041 5.460 0.870 30.400
18 0.500
R&D input 5360 0.019 1 1.390 1
1.650 8.430
Business Age 5360 17.800

Nature of property right 5360 0.522

Ability to grow 5360 2.120

ASIAN JOURNAL OF TECHNOLOGY INNOVATION 545

incentive are all zero. It can be seen that there are relatively few listed enterprises in the
heavy pollution industry implementing equity incentive, the average value of variable is
24303059, and the standard deviation is 66613486, indicating that the number of senior
executives’ equity incentive is generally low and there is obvious difference among
different enterprises. All other variables are within the reasonable value range, and
details are not described herein.

4.2. Regression analysis

4.2.1. Regression analysis results of government subsidy on green innovation of
enterprises
Table 4 reports the regression results of government subsidy and enterprise green inno-
vation, columns (1) and (2) report the estimation results of fixed effect and random effect
models respectively, and column (3) reports the model regression results of two-way
fixed effect. According to the Hausman test, the corresponding P value is 0.000, so the
fixed effect model is selected. Then, the individual fixed effect and time fixed effect are
constructed by F statistics. It is found that the corresponding P value is 0.000, and the
goodness of fitting of the model with the two-way fixed effect is better, so the two-way
fixed effect model is finally selected for analysis. The regression coefficient of government
subsidies is 0.007, and it is significant at the significance level of 10%, which indicates that
government subsidies play a significant role in promoting enterprise green innovation.
Thus hypothesis H1 is verified. This shows that government subsidies, as an important
supplementary mechanism of innovation resources, can share innovation risk for enter-
prises and enhance their willingness to innovate green. At the same time, because gov-
ernment subsidies play the role of ‘signal transmission’ in capital market, it can
effectively widen the external financing channel of enterprises and reduce the degree

Table 4. Regression analysis of government subsidies to enterprise green innovation.

Variable (1) (2) (3)
Gpatent Gpatent Gpatent

FE RE FE

Subsidy 0.008** 0.022*** 0.007*
(0.004) (0.004) (0.004)
Roa 0.242 0.285* 0.192

(0.171) (0.170) (0.169)
RD 2.280*** 3.493*** 2.559***
(0.754) (0.684) (0.747)
Age 0.059*** 0.042*** −0.067
(0.003) (0.003) (0.044)
Soe 0.157** 0.166*** 0.179***
(0.069) (0.047) (0.068)
TobinQ −0.023** −0.043*** −0.030***
(0.009) (0.009) (0.010)
Constant term −0.659*** −0.572*** 1.038*
(0.092) (0.088) (0.595)
R2 0.095 0.088 0.134
F 84.50 49.52
Firm fixed effect Control No control Control
Year fixed effect No control No control Control
Number of samples
5360 5360 5360

Note: Standard errors in parentheses; ***, **, * represent significance levels of 1%, 5%, and 10%, respectively.

546 H. LIU ET AL.

of financing constraint. In addition, the heavy pollution enterprises’ heavy dependence
on government subsidies also restricts the possible arbitrage behaviour of enterprises.

Considering that government subsidies may have a lag and cumulative effect on
green innovation activities of enterprises, based on model (1), this article regresses gov-
ernment subsidies and all control variables with a lag of one and two periods, respect-
ively. The regression coefficient of government subsidies in the first lag period is 0.008,
which is significant at the significance level of 5%. The regression coefficient of govern-

ment subsidies in the second lag period is 0.007, which did not pass the significance
test. This shows that, for the heavy pollution enterprises, the promotion of government
subsidies to the green innovation of enterprises, not only has a certain time-delay
characteristics, but also has a stronger influence coefficient and a higher significance
level, that is, the government subsidies obtained in the current period significantly
improve the green innovation level of enterprises in the current and next period.
The lagging effect of government subsidies may be because enterprises often need to
go through a certain period from R&D investment to patent output, on the other
hand, because the policy can play a certain guiding role, the local government’s
support for green innovation will enhance the confidence of enterprises, encourage
heavily polluting companies to allocate more resources to innovation activities later.
Therefore, hypothesis H1 is further validated (Table 5).

4.2.2. The analysis of the moderating effect of executive incentive
The differential effects of executive compensation incentive and equity incentive on the
moderating effect of government subsidies and enterprise green innovation are shown
in Table 6. The first column is the empirical results of the moderating effect of execu-
tive compensation incentives. The regression coefficient of the interaction term
between compensation incentive and government subsidies is −0.019, which passed
the significance test at the level of 5%. It shows that the compensation incentive
restrains the positive influence of government subsidies on the green innovation of
enterprises. Because the final effect of green innovation activity is difficult to reflect
on the financial performance in a short time, the compensation incentive based on
short-term performance evaluation will enhance the motivation of executives to
pursue private interests and strengthen the short-sightedness effect, so that they
divert the fund originally used for innovation to some quick profitable projects to
achieve the effect of smooth performance, resulting in the crowding out of R&D invest-
ment. The highly uncertain and risky results of green innovation activities also enhance
executives’ awareness of risk aversion, resulting in the failure of compensation incen-
tives to effectively resolve the mismatch between executives’ risk-taking and benefit

acquisition. To sum up, it is assumed that H2 is supported. The second list is the
empirical result of senior executives’ equity incentive moderating function. The
regression coefficient of the interaction term between equity incentives and govern-
ment subsidies is 0.095, but it has not passed the significance test. This may be
because the feedback of equity incentive effect also needs a certain time and cannot
be reflected in the current period.

Because it has been confirmed that there is a time lag in the effect of government sub-
sidies, this article believes that it is necessary to further explore the long-term and short-
term effects of different executive incentives. On the basis of the models (2) and (3), the

ASIAN JOURNAL OF TECHNOLOGY INNOVATION 547

Table 5. Regression analysis of lag effect.

Variable (1) (2)
Gpatent Gpatent

Subsidyt−1 0.008** 0.007
(0.004) (0.005)
Roat−1 0.240 0.336*
(0.184) (0.203)
RDt−1 1.772** −0.073
(0.817) (0.891)
Aget−1 −0.083 −0.093
(0.056) (0.068)
Soet−1 0.121 0.117
(0.075) (0.081)
TobinQt−1 −0.036*** −0.032***
(0.010) (0.011)

Subsidyt−2 1.597*
1.360* (0.919)
Roat−2 (0.760) 0.105
0.118
RDt−2 33.73
40.70 Control
Aget−2 Control Control
Control 4288
Soet−2 4824

TobinQt−2

Constant term
R2
F
Firm fixed effect
Year fixed effect
Number of samples

Note: Standard errors in parentheses; ***, **, * represent significance levels of 1%, 5%,
and 10%, respectively.

government subsidies, executive incentive and all control variables are regressed in the
first and second phases, respectively. The regressive results are shown in Table 7. Accord-
ing to column (1) and column (2), the regression coefficients of the reciprocal items of
compensation incentive and government subsidy are negative, but both failed the signifi-
cance test. This shows that the compensation incentive is a kind of short-term incentive,
and the short-term pressure will only crowd out the government subsidies in the current
period. According to the empirical results of column (3) and column (4), the regression
coefficient of the interaction term between equity incentives and government subsidies in

the first lag period is 0.023, and it is significant at the 10% significance level; the
regression coefficient of the interaction term in the second lag period is 0.035, which
is significant at the 5% significance level. This shows that equity incentive is a long-
term incentive mechanism, and its profit-driven effect reverses the short-sighted thinking
of executives, enhances the green innovation power of executives, and then promotes the
allocation of enterprise resources to innovation activities. Since the substantial synergy of
equity incentives also amplifies the signalling function and certification effect of govern-
ment subsidies, heavily polluting enterprises will attract more external capital injections.
Hence, this mechanism makes the policy effect closer to the expectations of the local gov-
ernment. H3 is verified.

548 H. LIU ET AL.

Table 6. Regression analysis of incentive moderating effect of senior
executives.

Variable (1) (2)
Gpatent Gpatent

Subsidy 0.010** 0.009**
(0.004) (0.004)
Subsidy × Payinc −0.019**
(0.008) 0.018
Payinc −0.002** (0.013)
(0.001) −0.003
Subsidy × Shainc (0.003)
0.199 0.209
Shainc (0.169) (0.169)
2.507*** 2.529***
Roa (0.747) (0.748)

−0.073* −0.067
RD (0.044) (0.044)
0.168** 0.173**
Age (0.068) (0.068)
−0.0291*** −0.031***
Soe (0.010) (0.010)
1.176** 1.029*
TobinQ (0.600) (0.595)
0.136 0.135
Constant term
R2 44.36 43.95
F Control Control
Firm fixed effect Control Control
Year fixed effect 5360 5360
Number of samples

Note: Standard errors in parentheses; ***, **, * represent significance levels of 1%, 5%,
and 10%, respectively.

4.3. Robustness test

4.3.1. Instrumental variable method
In practice, enterprises may get government subsidies because of green innovation activi-
ties. Although this article has delayed the variables by one and two periods to avoid the
endogenous problems of mutual causality, it may still face endogenous problems such as
missing variables. Therefore, the industry mean of government subsidies excluding the
enterprise itself is selected as the instrumental variable, and the endogenous influence
is excluded by 2SLS. See Table 8 for the regression results. The Cragg-Donald Wald F
statistic in Table 8 shows that there is no weak instrumental variable in the instrumental
variable, and the regression result of the instrumental variable is basically consistent with

that of the baseline model, that is, there is a significant positive correlation between gov-
ernment subsidies and green innovation of enterprises, which further illustrates the
robustness of the regression result in this article.

4.3.2. Replacing regression model

Because the median of green patent is zero, there is a large proportion of zero value in the
data, which forms a mixed distribution of ‘zero value accumulation’ and ‘continuous
positive value’ coexistence. Referring to the research of Tong et al. (2018), this article
uses Tobit model to re-examine the main basic regression. Table 9 reports the regression
results of Tobit model. After replacing the regression model, this article finds that

ASIAN JOURNAL OF TECHNOLOGY INNOVATION 549

Table 7. Hysteresis regression analysis of senior executive incentive moderating effect.

Variable (1) (2) (3) (4)
Gpatent Gpatent Gpatent Gpatent

Subsidyt−1 0.009** 0.007 0.011** 0.010**
Subsidyt−1* Payinct−1 (0.005) (0.005) (0.004) (0.005)
Payinct−1 −0.006 −0.003
Subsidyt−1×Shainct−1 (0.009) (0.010) 0.023* 0.035**
Shainct−1 −0.001 0.001 (0.014) (0.015)
Roat−1 (0.001) (0.001) −0.002 −0.001
RDt−1 (0.003) (0.003)
Aget−1 0.241 0.341* 0.260 0.356*
Soet−1 (0.184) (0.204) (0.185) (0.203)
TobinQt−1 1.757** −0.064 1.735** −0.149
Subsidyt−2 (0.817) (0.891) (0.817) (0.891)

Subsidyt−2* Payinct−2 −0.086 −0.093 −0.082 −0.093
Payinct−2 (0.056) (0.068) (0.056) (0.068)
Subsidyt−2×Shainct−2 0.115 0.118 0.116 0.112
Shainct−2 (0.075) (0.082) (0.075) (0.081)
Roat−2 −0.035*** −0.033*** −0.038*** −0.034***
RDt−2 (0.010) (0.011) (0.010) (0.011)
Aget−2 1.567* 1.546*
Soet−2 1.453* (0.924) 1.335* (0.919)
TobinQt−2 (0.765) 0.105 (0.760) 0.106
Constant term 0.118 0.119
R2 29.23 29.67
F 35.76 Control 35.89 Control
Firm fixed effect Control Control Control Control
Year fixed effect Control 4288 Control 4288
Number of samples 4824 4824

Note: Standard errors in parentheses; ***, **, * represent significance levels of 1%, 5%, and 10%, respectively.

government subsidies still play a significant role in promoting the green innovation of
enterprises in the heavy pollution industry, and the compensation incentive still plays
a significant negative moderating role in it, while the moderating role of equity incentive
is not significant. The results verify the consistency of the main conclusions in this article.

4.3.3. Replacing variable
In order to prevent the problem of collinearity, this article excludes the influence of
financial subsidies on the profitability, adjusts the total net interest rate of assets,

550 H. LIU ET AL.

Table 8. Regression results for instrumental variables.

Variable Column A
Gpatent

Instrumental variable First-order lag term
Subsidy 0.1408***
(0.054)
Roa
−0.0257
RD (0.030)
0.9793*
Age (0.510)
−0.2139
Soe (0.274)
0.2325***
TobinQ (0.079)
0.0115
Firm fixed effect (0.0154)
Year fixed effect Control
Cragg-Donald Wald F statistics Control
Number of samples
34.99
5220

Note: Standard errors in parentheses; ***, **, * represent significance levels of 1%, 5%,
and 10%, respectively.

measures the total net asset interest rate with ‘(net profit-financial subsidy)/average total
asset’, and tests it with the basic regression model. The regression results are shown in
Table 9. As can be seen from Table 10, the regression results did not materially

change, so the hypothesis in this article is considered to be robust.

5. Main research conclusions and enlightenment

5.1. Main research conclusions

How to lead the green production and operation of enterprises and get rid of the tra-
ditional development path of high pollution and high energy consumption has
become an important issue that cannot be ignored in the process of promoting eco-
logical civilisation construction and economic transformation. China has tried to use
government subsidies to drive green innovation of enterprises, but the effect of the
policy is controversial. This article takes heavily polluting enterprises as research
samples to study the relationship between government subsidies and green innovation,
and analyzes the coordination between executive incentives and government subsidies
in the process of influencing enterprise green innovation. The research results show
that: First, government subsidies significantly promote green innovation in heavily
polluting companies. Second, executive compensation incentives are not an effective
way to enhance the role of government subsidies in promoting green innovation,
but significantly weaken the positive relationship between them. Compensation incen-
tives cannot balance the risk-taking and benefit-acquisition of green innovation activi-
ties, and the risk aversion tendency of executives is still strong, which reduces the
effect of government subsidies. Third, executive equity incentives have formed a posi-
tive synergy effect on government subsidies. Although the moderating effect of equity
incentives in the current period is not significant, but after the first and second
periods lag, it is found that government subsidies can produce more green innovation

ASIAN JOURNAL OF TECHNOLOGY INNOVATION 551

Table 9. Regression results of tobit model.

Variable (1) (2) (3)
Gpatent Gpatent Gpatent

Subsidy 0.016*** 0.020*** 0.017***
Subsidy × Payinc (0.005) (0.006) (0.005)
`Payinc −0.029***
Subsidy × Shainc 0.229 (0.009) 0.018
Shainc (0.156) −0.003*** (0.013)
Roa 2.064*** (0.001) −0.005*
RD (0.745) (0.003)
Age −0.022** 0.238* 0.252
Soe (0.009) (0.132) (0.164)
TobinQ 0.255*** 1.986** 2.110***
Constant term (0.063) (0.863) (0.716)
Sigma_u −0.044*** −0.021*** −0.024***
Sigma_e (0.010) (0.007) (0.008)
Firm fixed effect 0.299** 0.237*** 0.237***
Year fixed effect (0.148) (0.055) (0.058)
Number of samples 0.708*** −0.042*** −0.046***
(0.039) (0.008) (0.007)
0.593*** 0.371*** 0.334**
(0.013) (0.138) (0.148)
Control 0.698*** 0.707***
Control (0.044) (0.046)
0.593*** 0.593***
5360 (0.016) (0.013)
Control Control
Control Control

5360 5360

Note: Standard errors in parentheses; ***, **, * represent significance levels of 1%, 5%, and 10%, respectively.

results. It can be seen that even in heavily polluting enterprises, equity incentives are
still an effective way to combine the personal interests of executives with the long-
term development of the company, which reverses the short-sighted thinking of
executives and helps to improve the positive effect of government subsidies on
green innovation.

5.2. Research enlightenment

First, the financing constraints of heavily polluting enterprises are still serious, and
insufficient funds have delayed the green innovation process of enterprises. Govern-
ments at all levels should strengthen financial support for heavily polluting enterprises,
ease the cost pressure on enterprises, and strengthen enterprises’ willingness to inno-
vate green. At the same time, we will eliminate discrimination in the financial industry
and guide more social funds to be injected into enterprises. Second, the financial
departments of governments at all levels should strengthen the evaluation of enter-
prises before issuing government subsidies, and can consider whether the enterprise
implements the executive equity incentive mechanism that is conducive to the
output of green innovation as an evaluation criterion. This will improve the accuracy
and effectiveness of fund allocation, reduce the problem of adverse selection between
government subsidies and green innovation to a certain extent, and prevent

552 H. LIU ET AL.

Table 10. Regression results for substituted variables.

Variable (1) (2) (3)
Gpatent Gpatent Gpatent

Subsidy 0.008* 0.011** 0.010**
(0.004) (0.004) (0.004)
Subsidy × Payinc −0.019**
0.014 (0.008) 0.018
Payinc (0.068) −0.002** (0.013)
2.520*** (0.001) −0.003
Subsidy × Shainc (0.747) (0.003)
−0.066 0.019 0.022
Shainc (0.044) (0.068) (0.068)
0.177*** 2.468*** 2.489***
Adjustroa (0.068) (0.746) (0.747)
−0.030*** −0.072* −0.066
RD (0.010) (0.044) (0.044)
1.030* 0.166** 0.171**
Age (0.595) (0.068) (0.068)
0.133 −0.028*** −0.031***
Soe (0.010) (0.010)
49.10 1.172* 1.022*
TobinQ Control (0.600) (0.595)
Control 0.135 0.133
Constant term 5360
R2 43.99 43.55
F Control Control
Firm fixed effect Control Control
Year fixed effect 5360 5360
Number of samples

Note: Standard errors in parentheses; ***, **, * represent significance levels of 1%, 5%, and 10%, respectively.

enterprises from using green innovation as a gimmick to defraud more external funds.
Third, governments at all levels should introduce supporting measures related to gov-
ernment subsidies. After the subsidy funds are released, relevant departments should
not only strengthen supervision to ensure that the use of funds is compliant and
reasonable, but also require enterprises to improve internal governance. Relying
solely on the direct supervision of government departments is too costly, and both
internal and external approaches can more effectively promote the construction of
China’s ecological civilisation. Fourth, relevant government departments should estab-
lish a dynamic assessment system and increase the proportion of green innovation in
the assessment system to encourage enterprises to introduce green innovation as a
performance goal into the executive incentive mechanism.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Funding

This work was supported by Shannxi Province Social Science Fund Program [grant numbers
2020D017, 2022HZ0540]; “Rural Revitalization” Special scientific research Project of Shaanxi
Normal University [grant number 22XCZX08].