Local Fiscal Pressure and Enterprise Environmental Protection Investment under COVID-19: Evidence from China

Abstract

Increasing local fiscal pressure and insufficient enterprise environmental protection investment are considerable problems in China. Based on the data of A-share-listed companies in heavily polluting industries in China from 2015 to 2021, this paper uses COVID-19 as an exogenous shock of local fiscal pressure and investigates the impact of local fiscal pressure on enterprise environmental protection investment by the continuous DID method. The study found that local fiscal pressure significantly reduces enterprise environmental protection investment and has a greater impact on non-state-owned enterprises, large enterprises, enterprises located in the “two-control-zone” cities and enterprises located in cities with low fiscal self-sufficiency rates. The mechanism of analysis shows that local governments ease fiscal pressure by “increasing revenue” and “cutting expenditure”. The former increases the non-tax burden of enterprises, while the latter reduces enterprise environmental protection subsidies, which leads to a decrease in environmental protection investment. The findings of this paper indicate that it is necessary to focus on solving the local fiscal pressure dilemma to increase green investment and achieve green development.

1. Introduction

Since China’s tax-sharing reform in 1994, the relationship between the central government and the local government has shown a trend of revenue centralization and expenditure decentralization, which has led to increasingly severe fiscal pressure on the local government [1]. Local fiscal pressure is often measured by the proportion of the gap between local fiscal revenue and expenditure to local fiscal revenue in previous studies. Figure 1 shows the change in local fiscal pressure from 1994 to 2020. After major public finance system reforms, such as the reform of income tax sharing in 2002, the reform of abolishing agricultural tax in 2005 and “five-to-five” sharing reform of VAT in 2016, the gap between local fiscal revenue and expenditure increased significantly, and local fiscal pressure showed a trend of fluctuating growth from 1994 to 2020.

The outbreak of the COVID-19 epidemic further reduced fiscal revenue and increases expenditure for local governments, making the already poor local finance worse. Additionally, the gap in fiscal revenue and expenditure widened. Due to the impact of COVID-19, local fiscal pressure rose from 0.38 to 0.435 [2]. On the one hand, local fiscal revenue decreased significantly. The epidemic had a great impact on China’s economy. In 2020, China’s GDP growth rate was 2.3%, the lowest level since 1978. During the COVID-19 epidemic, it was very difficult for enterprises to operate. To reduce the burden on enterprises, local governments stepped up the implementation of tax and fee reduction policies. In 2020 and 2021, tax and fee reductions exceeded CNY 2.5 trillion and CNY 1.1 trillion, respectively. Against the background of the economic downturn and tax and fee reduction, local fiscal revenue fell by 0.9% in 2020. On the other hand, the pressure on local fiscal expenditure was relatively high. Local governments undertook the “three guarantees” responsibility of “ensuring basic people’s livelihood, ensuring wages and ensuring operation.” The central government’s demands were implemented to ensure the priority of the “three guarantees” expenditure in local fiscal expenditure, so the “three guarantees” expenditure was a rigid policy. In 2020 and 2021, local fiscal expenditure on education, science and technology, social security and employment, and medical and health care totaled CNY 9.1 trillion and CNY 9.4 trillion, up 8.8% and 3.6% year on year, respectively.

Since the reform was implemented and opening up, and the sustained and high-speed growth of China’s economy, many negative problems caused by environmental pollution have become increasingly prominent. Industrial activities are the main source of pollution. According to the Statistical Annual Report of Ecological Environment of China in 2020, the proportions of sulfur dioxide emissions, nitrogen oxides emissions and particulate matter emissions from industrial sources in China were 80%, 41% and 66%, respectively. Environmental pollution has negative externalities, and environmental governance has positive externalities. Enterprises have no incentive to actively participate in environmental governance activities without regulation [3,4]. Enterprise environmental protection investment has been insufficient for a long time in China. According to statistics, the proportion of environmental protection investment in enterprise assets is always less than 3% [5].

Increasing local fiscal pressure and insufficient enterprise environmental protection investment are considerable problems in China. However, there is no literature on the relationship between local fiscal pressure and the environmental protection investment of enterprises. How will local fiscal pressure affect enterprise environmental protection investment? To answer this question, this paper uses COVID-19 as the exogenous shock of local fiscal pressure, applying the continuous DID method to investigate the impact of local fiscal pressure on enterprise environmental protection investment. The research objectives of this study include the following two aspects. The first is to reveal the relationship between local fiscal pressure and enterprise environmental protection investment from theoretical and empirical perspectives, focusing on the heterogeneity of firms and cities regarding fiscal pressure on the environmental protection investment of enterprises. The second is to explore the influence channels of local fiscal pressure on enterprise’s environmental protection behaviors.

The remainder of this paper is arranged as follows: the second part summarizes the relevant literature and presents the theoretical hypothesis of this paper; the third part is the research design, including the introduction of the theoretical model and the description of the main variables and related data; the fourth part is the empirical results; the fifth part is further analysis; the sixth part is the discussion; and the final section is the conclusion and policy implication.

2. Literature Review and Theoretical Hypotheses

2.1. Literature Review

COVID-19 has had enormous economic and social impacts. The impact of the pandemic on government finances has been widely analyzed [6,7], and this article is closely related to these studies. Because local governments are key actors in COVID-19 crisis management [8], local budgets will suffer more due to COVID-19 than budgets at the central level [9]. Local governments were found to adopt widely varied action strategies in response to COVID-19 [10,11], including containment and closure, economic response and health systems [12], which significantly impact local government finances on both the expenditure and revenue sides of their budgets. On the one hand, the epidemic led to a decline in local fiscal revenue. Aiming at counteracting the effects of COVID-19, local governments were willing to exempt businesses from various fees and taxes [11], and grant allowances, discounts and deferrals [10]. Not only were tax revenues falling, but also parking fees, public transport fares, local accommodation fees, public hospital charges, fees for the use of public spaces and rental incomes [9,13]. On the other hand, the epidemic led to an increase in local fiscal expenditure. Local governments significantly increased public health expenditure for epidemic governance [12]. At the same time, additional expenses for supporting residents and enterprises also increased sharply [11]. The COVID-19 pandemic led to serious financial burdens for local governments due to falling incomes, rising expenditure and problems with budget stabilization [14].

At present, scholars have not yet reached an agreement on how to measure local fiscal pressure. In the research, most scholars use “the gap between fiscal revenue and expenditure” to measure the fiscal pressure of local governments, including the difference method and ratio method. The former measures local fiscal pressure using the difference between regional fiscal expenditure and fiscal revenue, while the latter measures fiscal pressure by the ratio of the difference between regional fiscal expenditure and revenue to total fiscal revenue [2]. However, this method of directly using statistical data to calculate the local fiscal gap ignores the role of the transfer payment system, and can easily cause endogenous problems in the research. To solve this problem, scholars have measured fiscal pressure through quasi-natural experiments of policy shocks. For example, scholars have used the rotated minister system [15], the reform of abolishing agricultural tax [16,17,18], the reform of income tax sharing [19], the “business tax to VAT” policy [20], the real estate purchase restriction policy and the reform of educational authority as exogenous shocks of fiscal pressure [21,22], and constructed quasi-natural experiments to measure the changes in fiscal pressure of local governments.

Regarding the environmental effect of fiscal pressure, scholars have mainly discussed the impact of local government behavior on environmental quality from the perspective of fiscal pressure. Under the promotion incentive mechanism of “political tournament” in China, the local government tended to prioritize to economic development over environmental protection [23,24]. When faced with fiscal pressure, to promote local economic development and thus increase fiscal revenue, the local government is motivated to relax environmental control and adopt the “race to the bottom” strategy in exchange for economic growth [25,26]. Specifically, the increase in local fiscal pressure leads to an increase in the local industrial pollution level [27], a significant increase in local carbon emissions [22] and a deterioration in local air quality [28]. The greater the local fiscal pressure, the more obvious the boundary effect of air pollution [29]. In addition, under fiscal pressure, the local government relaxes environmental regulations. On the one hand, it can attract polluting enterprises with a high output value to settle down and develop locally [30]. On the other hand, it is beneficial to support local industrial enterprises with key tax sources to expand their production scale. This behavior is conducive to achieving economic growth and fiscal revenue targets, but it also leads to environmental pollution. It also reduces the environmental governance efficiency of the local government. The greater the fiscal pressure, the lower the local government’s environmental governance efficiency [31].

Based on the related literature, it can be observed that the existing research mainly has the following two shortcomings. On the one hand, the existing research on the environmental effects of local fiscal pressure was conducted mostly from a macro-enterprise perspective and rarely from the perspective of micro-enterprises. On the other hand, most related studies have paid attention to the pollution effect of fiscal pressure from the perspective of pollution emission and little attention has been paid to pollution control. Compared with the previous literature, the potential contributions of this paper are mainly reflected in the following aspects.

First, from the perspective of micro-enterprises, based on the data of A-share-listed companies from heavily polluting industries in China, this paper systematically studied the impact of local fiscal pressure on the environmental protection investment of enterprises and its internal mechanism. The existing literature has paid little attention to the relationship between local fiscal pressure and enterprise environmental protection investment. This paper innovatively discusses the relationship between them, enriching the related research on local fiscal pressure and environmental governance.

Second, it can be seen from previous studies that COVID-19 has had a significant impact on local public finance. The COVID-19 epidemic has been used as an exogenous shock to reflect the change in local fiscal pressure, which further enriches the measurement method of fiscal pressure. This work not only overcomes the endogenous problem of measuring local fiscal pressure, but also reflects the current situation with the latest data.

Third, this paper further verifies the key role of public finance in environmental governance by placing local fiscal pressure, government behavior and enterprise environmental protection investment into the unified analysis framework. The conclusion of this paper is significant to the future environmental governance of China, and it is necessary to focus on solving the local fiscal pressure dilemma to increase green investment and achieve green development.

2.2. Theoretical Hypotheses

In the face of fiscal pressure, local governments often choose the way of “increasing revenue and cutting expenditure” to alleviate the pressure from the income side and the expenditure side, respectively. In terms of revenue, scholars generally believe that the local government’s market access to fiscal revenue is the main approach to deal with fiscal pressure. Research has investigated in depth the relationship between fiscal pressure and local government financial resources cultivation, local government debt issuance, tax collection and administration, non-tax revenue collection and management. These behaviors are common choices for local governments to alleviate fiscal pressure.

First of all, as for the cultivation of financial resources, under the “pressure-type” fiscal incentives, local governments increase fiscal revenue and expand the tax base by vigorously developing the secondary industry and even relaxing environmental regulations to introduce enterprises with overcapacity [32]. The cultivation of financial resources by promoting local economic growth, the so-called “releasing water to raise fish”, does help local governments to fundamentally solve the fiscal pressure problem, but it often takes a long time to cultivate financial resources, which cannot alleviate the fiscal pressure caused by the COVID-19 epidemic in a short time.

Secondly, in terms of debt issuance, there are two limitations in dealing with fiscal pressure by issuing local government debt. On the one hand, the local government debt scale is subject to limited management, and it cannot exceed the approved limit. On the other hand, there is a certain lag in alleviating fiscal pressure through debt expansion [33].

Thirdly, as for the land transfer system, some scholars consider that land transfer is an important channel for local governments to ease fiscal pressure [34], but some scholars hold different opinions. The study by Fan (2015) found that the real reason for local governments to sell land is not fiscal pressure, but rather the impulsive investment behavior of local governments [15]. Faced with fiscal pressure, local governments will even reduce the transfer of industrial land [35].

Fourthly, in terms of tax collection and management, strengthening the tax collection and management of enterprises and transferring fiscal pressure to the enterprises within their jurisdiction are important ways for local governments to increase fiscal revenue and ease fiscal pressure [16]. However, strengthening tax collection and management undoubtedly increase the tax burden of enterprises, which is contrary to the tax reduction policy that China is vigorously promoting at present.

Finally, in the collection and management of non-tax revenue, local governments have the authority to set up and collect non-tax revenue, and can adjust the scale and intensity of non-tax revenue collection and management according to the current fiscal revenue and expenditure situation [36]. It can even be said that non-tax revenue is one of the areas where the local governments have the highest flexibility, the strongest intervention ability and the least restriction on collection and management among all financing means [37]. Therefore, non-tax revenue has become an important channel for local governments to relieve fiscal pressure [19].

To sum up, this paper considered that extending the “grab hand” to enterprises and strengthening the collection and management of non-tax revenue are effective policy tools for local governments to alleviate fiscal pressure in the short term. If the local government narrows the gap between fiscal revenue and expenditure by strengthening the collection and management of non-tax revenue, it will undoubtedly increase the non-tax burden on enterprises [19]. Non-tax expenditure, as an important part of enterprise expenditure flow, will reduce the cash flow and profitability of enterprises. The enterprises will coordinate the relationship between private marginal income and marginal cost according to their own profit target and financial situation. Due to the typical positive externality of environmental protection behavior, the private marginal income of environmental protection investment will be lower than the private marginal cost for enterprises. Therefore, enterprises are incentivized to distort the investment proportion of their production factors, tilt the production factors to the economic activities and reduce the investment in non-economic activities, such as environmental protection investment. In other words, strengthening the collection and management of non-tax revenue will lead enterprises to adjust the allocation of resources between their economic activities and non-economic activities, resulting in a crowding-out effect on environmental protection investment. Based on this, the following hypothesis is presented:

Hypothesis 1.

Local governments will strengthen the collection and management of non-tax revenue under fiscal pressure, leading enterprises to reduce their environmental protection investment.

In terms of fiscal expenditure, local governments can also ease the fiscal pressure by “cutting expenditure”. Especially in western countries, because of the flexibility of the budget system, local governments have more room to adjust the scale of fiscal expenditure flexibly. Scholars have found that, under fiscal pressure, the scale of local government expenditure shows a significant downward trend [38,39,40]. However, in China, the incentive from the economic growth target under the “political tournament” and the budget constraint caused by the increasing public demand make it difficult to flexibly change the local fiscal expenditure level. The fiscal expenditure of local government has a significant downward rigidity [41]. Therefore, in the face of fiscal pressure, the space for local governments to control simply by reducing the scale of fiscal expenditure is relatively limited, and they are more dependent on the adjustment of the fiscal expenditure structure [42].

First of all, fiscal pressure leads to the distortion of the public expenditure structure of local governments, showing a bias towards “emphasizing production and neglecting people’s livelihood” [43]. Fiscal productive expenditure is conducive to promoting the development of the local economy, thus bringing more tax sources. Therefore, for local governments, productive expenditures are often characterized by exclusive marginal income, high economic efficiency and a short investment cycle. On the contrary, the expenditure on people’s livelihood shows a significant income spillover effect, and it takes a long time to become effective. Under fiscal pressure, the local governments shift their limited financial funds from the people’s livelihood to the production field by adjusting the expenditure structure, prioritizing the realization of economic and tax targets. Secondly, among all types of people’s livelihood expenditures, the supply order of public goods, such as education, medical care and employment, still takes precedence over environmental governance [44]. Therefore, to alleviate fiscal pressure, local governments will further reduce environmental-protection-related expenditures and reduce environmental subsidies to enterprises.

Environmental protection subsidies are important tools for environmental regulation in China. It is a special subsidy provided by the government to encourage enterprises to actively invest in environmental protection, which is beneficial to internalizing the externalities of environmental governance [45]. According to the “Several Provisions on Strengthening the Management of Environmental Protection Subsidy Funds”, the environmental protection subsidy funds are earmarked for the environmental protection investment projects of enterprises, such as key pollution sources control and comprehensive environmental management. Environmental subsidies provide direct financial support for enterprises to participate in environmental governance activities and reduce the cost of environmental protection investment [46]; at the same time, they reduce the occupation of production funds by environmental protection investment and alleviate the shortage of funds for enterprises. The environmental subsidies are conducive to internalizing the positive externalities of enterprise environmental governance, thus encouraging enterprises. Once the government reduces the expenditure of environmental protection subsidies under fiscal pressure, the cash flow that enterprises can directly invest in environmental protection investment is reduced, and the enterprises are expected to reduce environmental protection investment. Based on this, the following hypothesis is proposed:

Hypothesis 2.

Local governments will reduce environmental subsidies under fiscal pressure, leading enterprises to reduce their environmental protection investment.

3. Research Design

3.1. Model Setting

To identify the causal relationship between local fiscal pressure and enterprise environmental protection investment, we applied the difference-in-differences (DID) method, which is commonly applied in microeconomic research to estimate policy effects when certain groups are exposed to a policy shock and others are not. The externality of policy is one of the necessary conditions for using DID. Before the outbreak of COVID-19, no one could have predicted its appearance. Therefore, the outbreak of COVID-19 provides a superexcellent exogenous shock to capture the sharp change in local fiscal pressure, and the perfect opportunity to estimate the causal effect of fiscal pressure. Because the epidemic affected various industries in various regions, the traditional DID method is not applicable, as it strictly distinguishes the treatment group and the control group according to whether they are treated or not, respectively. Referring to Nunn and Qian (2011) [47], this paper used the continuous DID method to solve this problem, replacing the binary dummy variable used to indicate whether it is treated or not with the continuous variable with different degrees of treating. The detailed design of the model is as follows:

$${invest}_{it}={\beta }_0+{\beta }_1{confirmed}_i\times {after}_t+{\beta }_2{X}_{it}+{\theta }_i+{\mu }_t+{\epsilon }_{it}$$

(1)

where the subscript i represents the enterprise and t is the year. The dependent variable invest is the environmental protection investment of the enterprise. ${confirmed}_i\times {after}_t$ is the core explanatory variable, where confirmed is the severity of the COVID-19 pandemic and after is a dummy variable in the time dimension. X is a series of control variables. θ_i is the individual fixed effect, μ_t is the time fixed effect and ε_it is the random error term. The interaction coefficient β₁ is the core of this paper, and its size and significance reflect the impact of fiscal pressure on environmental protection investment of enterprises during the COVID-19 pandemic. In Equation (1), if the estimated interaction coefficient β₁ is significantly negative, it shows that local fiscal pressure leads to reducing the environmental protection investment of enterprises.

3.2. Variable Definition

3.2.1. Explained Variable

In this paper, the COVID-19 epidemic was used as an exogenous shock variable to measure local fiscal pressure. Because the epidemic broke out in 2020, there are only data in 2020 and 2021 in China at present, so this paper had to choose the explained variable to which enterprises could quickly respond and that they could measure. Facing fiscal pressure under the impact of an epidemic, local governments’ behaviors of “increasing revenue and reducing expenditure” affect the environmental protection decision-making process of enterprises. Increasing or decreasing investment in environmental protection is a decision-making response that enterprises can make in a short time. Therefore, this paper chose environmental protection investment as the explained variable. The environmental protection investment of enterprises is the sum of capitalized expenditure and expensed expenditure. Among them, capitalized expenditures include environmental protection technology research and development expenditure, investment in environmental protection facilities and renovation expenditure, and expensed expenditures include environmental protection facilities operating expenses, greening expenses and sewage charges [48].

3.2.2. Core Explanatory Variable

The core explanatory variable is the interaction term $confirmed\times after$, where ‘confirmed’ is the severity of the COVID-19 pandemic, which is the cumulative diagnosis rate of a given city where the enterprise is located. The cumulative diagnosis rate is the ratio of the cumulative number of confirmed cases to the number of resident populations in each city. As for the dummy variable ‘after’, after the outbreak of the COVID-19 epidemic (that is, after 2020), it was ascribed the value of 1, and before that, it was assigned the value of 0.

3.2.3. Control Variables

To control other factors that affect the environmental protection investment of enterprises, this paper introduced the following control variables referring to the research of Zhang et al. (2019) [48]: financial leverage ratio, return on assets, enterprise size, cash flow from operations, fixed asset ratio, state-owned enterprise, per capita GDP of cities and government intervention level and controls individual fixed effects and year fixed effect. Detailed definitions of the variables are shown in

Table 1. Definition of the main variables.

Variable Type	Variable Symbol	Variable Name	Definition
Explained variable	invest	environmental protection investment	the logarithm of the total environmental protection investment
Explanatory variables	confirmed	the severity of the COVID-19 pandemic	cumulative number of confirmed cases/the number of resident populations in each city
	after	time dummy variable	a dummy variable that equals to 1 if the time is 2020 and beyond; otherwise, it equals to 0
Control variables of the enterprise dimension	lev	financial leverage ratio	liabilities/total assets
	roa	return on assets	net profit/total assets
	size	enterprise size	the logarithm of the total assets
	cashfo	cash flow from operations	net cash flow from operations/total assets
	density	fixed asset ratio	fixed asset/total assets
	soe	state-owned enterprise	a dummy variable that equals to 1 if an enterprise is state-owned; otherwise, it equals to 0
Control variables of the city dimension	lnpgdp	per capita GDP	the logarithm of per capita GDP
	gov	government intervention level	expenditure in local fiscal budget/local GDP

.

3.3. Sample Selection and Data Sources

In this paper, A-share-listed companies in heavily polluting industries in China from 2015 to 2021 were considered as the research samples. According to the Guide for Environmental Information Disclosure of Listed Companies issued by the Ministry of Environmental Protection in 2010 and the industry classification standard published by the China Securities Regulatory Commission (CSRC) in 2012, the listed companies in heavily polluting industries identified in this paper included: mining, textile, paper and paper products, petroleum, chemical industry, chemical fiber, ferrous (non-ferrous) metal smelting and processing, pharmaceutical industry, rubber and plastic, and fur products. First, this paper chose the listed companies in heavily polluting industries as the research sample because of the availability of data. The Guide for Environmental Information Disclosure of Listed Companies emphasizes that listed companies in heavily polluting industries should disclose environmental information accurately, timely and completely, without false records, misleading statements or significant omissions. Secondly, the total pollutant discharge of heavily polluting enterprises is generally much higher than that of non-heavily polluting enterprises, because the environmental damage caused by insufficient investment in environmental protection is more serious. Finally, heavily polluting industries are mostly manufacturing industries with a higher output value and more vital tax-creating ability, which are both large polluter and large taxpayer characteristics [49]. The environmental protection investment decisions of these enterprises are more easily influenced by the choice of behavior of local governments under fiscal pressure. Therefore, the research on environmental protection investment in heavily polluting industries has more practical significance and theoretical value. The selection of the starting year of the sample in this study was mainly due to the following considerations: the newly revised “Environmental Protection Law of People’s Republic of China” came into force on 1 January 2015, and its enforcement and supervision are stricter than before, which increases the cost of sewage discharge for polluting enterprises. To eliminate the possible impact of policy intervention on corporate environmental governance, this paper chose 2015 as the starting year of the sample.

Based on the original data, according to the usual practices in the literature, this paper considered the original data as follows: (1) eliminate ST and *ST companies; (2) eliminate sample companies with missing critical indicator data; and (3) control the influence of extreme values, with all continuous variables being winsorized at the 1% and 99% levels. This paper finally obtained 2427 sample observations from 621 listed companies.

The data in this paper came from the following sources. (1) The data on environmental protection investment came from the notes of listed companies’ annual reports, which we collected manually. The capitalized expenditure is directly related to environmental protection, and the data originated from “construction in progress”. The expensed expenditure data were from “management expenses” in the income statement of the listed companies. (2) Epidemic data and other data at the company level were from the CSMAR database and CNRDS database. (3) The municipal data mainly came from the China City Statistical Yearbook, and the data in 2021 came from the Statistical Bulletin on National Economic and Social Development, which were collected manually.

4. Empirical Results

4.1. Descriptive Statistics

The results of descriptive statistics are reported in

Table 2. Descriptive statistics of the main variables.

Variable	N	Mean	Sd	P10	P50	P90
invest	2427	16.060	2.268	13.21	16.120	18.990
confirmed	2427	0.566	2.033	0.035	0.132	1.412
after	2427	0.275	0.447	0	0	1
lev	2427	0.423	0.216	0.159	0.410	0.689
roa	2427	0.044	0.175	−0.008	0.041	0.115
size	2427	22.540	1.288	21.020	22.400	24.320
cashfo	2427	0.065	0.080	−0.006	0.063	0.146
density	2427	0.316	0.151	0.125	0.306	0.523
soe	2427	0.425	0.494	0	0	1
lnpgdp	2427	11.290	0.512	10.580	11.310	11.970
gov	2427	0.168	0.078	0.101	0.149	0.249

. The average of environmental protection investment of the listed companies in heavily polluting industries was 16.060. It was 16.167 before 2020 and 15.770 after the outbreak of the COVID-19 epidemic. Thus, we learn that enterprise environmental protection investment decreased after the epidemic. The standard deviation was as high as 2.268, which shows that the scale of environmental protection investment among enterprises in China is quite different. This also means that it is necessary to encourage enterprises to actively participate in environmental governance through macro-control by the government. The descriptive statistical results of the other control variables are consistent with the existing studies, and the range of values was reasonable.

4.2. Basic Regression Results

The basic regression results are shown in

Table 3. Basic regression results.

	(1)	(2)	(3)
confirmed × after	−0.066 ***	−0.070 ***	−0.070 ***
	(−2.78)	(−3.37)	(−3.37)
lev		0.902 *	0.903 *
		(1.74)	(1.75)
roa		0.350 *	0.349 *
		(1.76)	(1.76)
size		0.977 ***	0.981 ***
		(5.50)	(5.53)
cashfo		0.281	0.273
		(0.57)	(0.55)
density		−0.508	−0.503
		(−0.74)	(−0.73)
soe		0.119	0.118
		(0.32)	(0.31)
lnpgdp			0.218
			(0.53)
gov			0.547
			(0.67)
Constant	16.144 ***	−6.234	−8.888
	(4482.65)	(−1.54)	(−1.42)
Company FE	YES	YES	YES
Year FE	YES	YES	YES
Observations	2291	2291	2288
R-squared	0.715	0.729	0.729

Note: *** and * are significant at the levels of 1% and 10%, respectively, with the value of T in brackets. This also applies to the following table.

. The columns indicate the different choices of fixed effects and control variables. The coefficient of confirmed × after is negative and significant at the 1% level. This result shows that local fiscal pressure caused by the COVID-19 epidemic has a significant negative impact on the environmental protection investment of enterprises.

From the analysis of the coefficients of the control variables, the enterprise size, financial leverage ratio and return on assets have significant positive effects on the environmental protection investment of enterprises. The larger an enterprise is, the more resources it has, and the more likely it is to invest in environmental protection. When the financial leverage ratio is high, the financing constraints faced by enterprises are small. When the capital is sufficient, it is conducive to increasing enterprise environmental protection investment. For enterprises with a high return on assets, it is obvious that the better their profitability, the stronger their ability to invest in environmental protection.

4.3. Robustness Test

4.3.1. Parallel Trend Test

The premise of the validity of the estimated results of the DID model is to meet the parallel trend hypothesis, that is, the explained variables of the treatment group and the control group must have the same development trend before the policy influence; otherwise, the policy effect obtained by the difference is partly caused by the difference between the treatment and control groups. Therefore, this paper used the event study method to test the parallel trend hypothesis. The detailed model was designed as follows:

$${invest}_{it}=\mathsf{\alpha }+\sum _{j=2016}^{2021}{\beta }_j{confirmed}_i\times {time}_t^j+\gamma X_{it}+{\theta }_i+{\mu }_t+{\epsilon }_{it}$$

(2)

where ${time}_t^j$ is a dummy variable representing the year of the COVID-19 epidemic outbreak and the years before and after the epidemic outbreak. To avoid the multicollinearity problem, we chose 2015 as the base year. The definitions of other variables are the same as mentioned above.

Figure 2 shows the common trend results with a 95% confidence interval. The coefficients of the interaction term are statistically insignificant and close to zero, indicating that no significant difference exists between the treatment and control groups in the environmental protection investments of enterprises in the pre-epidemic period. This provides evidence to support the hypothesis of common trend in this paper.

4.3.2. Placebo Test

Although we included control variables and fixed effects, the results may also relate to other unobservable factors. In order to solve this problem, we conducted a placebo test by changing the time point of the COVID-19 outbreak. Based on the sample data before 2020, this paper assumed the outbreak time was 2016, 2017, 2018, and 2019, and empirically tested the environmental effects of government fiscal pressure according to model (1). The empirical results are shown in

Table 4. Placebo test results.

Variables	(1)	(2)	(3)	(4)
	2016	2017	2018	2019
confirmed × after	−0.082	0.037	−0.030	−0.082
	(−0.72)	(0.41)	(−0.34)	(−0.76)
Constant	−8.437	−7.699	−8.299	−8.710
	(−1.15)	(−1.05)	(−1.13)	(−1.19)
Controls	YES	YES	YES	YES
Company FE	YES	YES	YES	YES
Year FE	YES	YES	YES	YES
Observations	1654	1654	1654	1654
R-squared	0.773	0.773	0.773	0.773

. No matter how many years the outbreak time of the COVID-19 epidemic moves forward, the coefficients of confirmed × after fail to pass the significance test, which shows that no significant impact from unobservable factors exists, and the identification strategy is valid.

4.3.3. Replace the Explained Variable

In the basic regression, the explained variable is measured by the logarithm of the total environmental protection investment. Because different amounts of ecological protection investment reflect different environmental protection investment intensities, for smaller enterprises, less environmental protection investment may also represent greater environmental protection investment intensity. Therefore, referring to Liu et al. (2021) [50], this paper re-measured the environmental protection investment of enterprises by using two proportional indicators: the logarithm of environmental investment divided by the logarithm of total assets (EPinvest1) and the logarithm of environmental investment divided by the logarithm of main business income (EPinvest2). The reason why total assets and main business income were chosen is that total assets is a static accounting indicator reflecting the financial situation of an enterprise on a specific date. Contrastingly, main business income is a dynamic accounting indicator reflecting the operating results of an enterprise in a certain accounting period. The two do not only form a contrast, but also complement each other. The results are shown in columns (1) and (2) of

Table 5. Robustness test for the substitution of the explained variables and the exclusion of other policy effects.

Variables	(1)	(2)	(3)	(4)
	EPinvest1	EPinvest2	Environmental Protection Tax Law	Central Government Environmental Inspection
confirmed × after	−0.003 ***	−0.003 ***	−0.070 ***	−0.071 ***
	(−3.45)	(−3.39)	(−3.38)	(−3.39)
EPT			−1.311 ***
			(−6.23)
CGEI				−0.068
				(−0.59)
Constant	0.318	0.108	−8.606	−9.107
	(1.16)	(0.38)	(−1.39)	(−1.46)
Controls	YES	YES	YES	YES
Company FE	YES	YES	YES	YES
Year FE	YES	YES	YES	YES
Observations	2288	2287	2288	2288
R-squared	0.623	0.623	0.729	0.729

Note: *** is significant at the levels of 1%.

, respectively. The coefficients of the interaction term are significantly negative at the 1% level, indicating that the increase in local government fiscal pressure under the impact of the COVID-19 epidemic indeed led to the reduction in enterprise environmental protection investment.

4.3.4. Exclusion of the Impact of Other Policies

• People’s Republic of China (PRC) Environmental Protection Tax Law

The Environmental Protection Tax has been formally implemented since 1 January 2018, and since then, polluting enterprises have changed from paying sewage charges to paying environmental protection tax (hereinafter referred to as “Environmental Protection Fee-to-Tax”). Compared with sewage charges, the environmental protection tax is a compulsory, executable and supervisory tool for environmental regulation. The cost and regulatory pressures it brings will force enterprises to participate in environmental governance [51]. In this paper, the “Environmental Protection Tax” policy implementation variable EPT was introduced to eliminate the impact of “Environmental Protection Fee-to-Tax” on the environmental protection investment of enterprises. Before 2018, the dummy variable EPT equals 0; otherwise, it equals 1. The results are shown in column (3) of Table 5, and the coefficient of the interaction term is negative and significant at the 1% level. This shows that, under the impact of the COVID-19 epidemic, the negative impact of local fiscal pressure on enterprise environmental protection investment is not affected by the implementation of the “Environmental Protection Fee-to-Tax”.

2.

Central Government Environmental Inspection

The work of the Central Government Environmental Inspection began in 2016. As a compulsory environmental regulation tool, the Central Government Environmental Inspection brings local governments into the scope of supervision, focuses on the implementation of local environmental policies and strengthens the environmental responsibility awareness of local governments [52]. Under the background of the Central Government Environmental Inspection, local governments transfer environmental protection pressure to enterprises, thus affecting the environmental protection investment decisions. In this paper, CGEI, the policy implementation variable of Central Government Environmental Inspection, was introduced to eliminate the policy impact. Considering that the Central Government Environmental Inspection has established a long-term continuous tracking mechanism, if the enterprise is in the area where the inspector first enters, CGEI = 1; otherwise, CGEI = 0. The results are shown in column (4) of Table 5, and the coefficient of the interaction term is negative and it is significant at 1%. This shows that, under the impact of the COVID-19 epidemic, the negative impact of local fiscal pressure on enterprise environmental protection investment is not affected by the Central Government Environmental Inspection, which proves, once again, that the regression results of this paper are robust.

4.4. Mechanism Test

As can be seen from the above empirical analysis, the basic conclusion of this paper is that, under the impact of the COVID-19 epidemic, local fiscal pressure leads to the decline in enterprise environmental protection investment. Subsequently, this paper investigated the internal mechanism. In the mechanism analysis, this paper followed the following logical route: when the local government faces great fiscal pressure, it eases fiscal pressure through the behavior of “increasing revenue” on the revenue side and “cutting expenditure” on the expenditure side, thus affecting the decision-making process of environmental protection investment of enterprises.

4.4.1. Local Governments Increase Revenue

Referring to Park (2017) [53], this paper used penalty expenditure (Pen) to measure the non-tax burden of enterprises. In this paper, the total assets of the company were used to standardize the penalty expenditure, and then were multiplied by 100 to improve the readability of the regression coefficient. Faced with fiscal pressure, the local governments will strengthen the collection and management of non-tax revenue. Imposing fines on enterprises is an important means for local governments to alleviate their fiscal pressure. In China, penalty funds are included in local budgets, and revenue and expenditure are managed separately. Penalty funds are an important source of local government revenue [54], and are characterized by high flexibility and having a large space for collection and administration. As for the penalty funds, the local governments have a certain decision-making power in revenue and expenditure.

The results are shown in columns (1) and (2) of

Table 6. Results of the mechanism analysis.

Variables	(1)	(2)	(3)	(4)	(5)
	Pen	Pen	Taxrt1	Taxrt2	Sub
confirmed × after	0.018 **	0.021 **	0.000	−0.014	−0.068 ***
	(2.20)	(2.75)	(1.50)	(−0.51)	(−3.00)
Constant	0.399 ***	6.364	0.028	6.991	2.914 ***
	(39.69)	(1.06)	(1.71)	(0.45)	(161.17)
Controls	NO	YES	YES	YES	NO
Industry FE	YES	YES	YES	YES	YES
Year FE	YES	YES	YES	YES	YES
Observations	2291	2288	2365	2365	2522
R-squared	0.376	0.383	0.588	0.192	0.582

Note: *** and ** are significant at the levels of 1% and 5%, respectively.

. The coefficients of confirmed × after are positive regardless of whether the control variables are added, which are significant at the level of 5%, indicating that local fiscal pressure under the impact of the COVID-19 epidemic has a significant and positive effect on enterprise penalty expenditure. In other words, strengthening the collection and management of non-tax revenue is indeed an important way for local governments to ease the fiscal pressure, which leads to a decline in the environmental protection investment of enterprises, and verifies the rationality of Hypothesis 1.

In addition, this paper also further examined the impact of fiscal pressure on corporate tax burden. Because the collection procedures of the turnover tax, such as consumption tax and value-added tax, are relatively perfect, the local governments have very limited space for the collection of turnover tax. Therefore, the discussion on tax collection and management mainly focused on corporate income tax. In this paper, the ratio of corporate income tax payable to total assets (Taxrt1) and the ratio of corporate income tax payable to pre-tax profit (Taxrt2) were used to measure the corporate income tax burden. From columns (3) and (4) of Table 6, we can observe that local fiscal pressure has no significant impact on corporate income tax burden. That is to say, local governments have not alleviated the fiscal pressure by strengthening tax collection and management, which is consistent with the current policy background of reducing the corporate tax burden at a large scale in China.

4.4.2. Local Governments Cut Expenditure

From the above theoretical analysis, it can be seen that, to alleviate fiscal pressure, local governments will further reduce environmental subsidies to enterprises. In this paper, enterprise environmental subsidy (Sub) was used as the explained variable, and the influence of fiscal pressure on enterprise environmental subsidy was empirically tested. Because there are a considerable number of zero values in the enterprise environmental protection subsidies, and these zero values have a meaning, we added “1” to the amount of enterprise environmental protection subsidies and then used the natural logarithm.

The results are shown in columns (5) and (6) of Table 6, and the coefficients of the interaction item confirmed × after are negative regardless of whether the control variables are added, which are significant at the level of 1%. Local fiscal pressure has a significant negative impact on enterprises’ environmental subsidies. This shows that the reduction in enterprise environmental subsidy expenditure is another important channel for local governments to alleviate fiscal pressure, which in turn leads to the reduction in enterprise environmental investment. The rationality of Hypothesis 2 is verified.

5. Further Analysis

This paper analyzed heterogeneity from the two dimensions of enterprise and city. Among them, we mainly considered the differences in enterprise equity nature and enterprise size in the enterprise dimension, and the differences in environmental regulation and public fiscal self-sufficiency rate in the city dimension.

5.1. Enterprise Equity Nature

To further test the robustness of the regression results, based on the enterprise equity nature, this paper divided the samples into two groups: state-owned enterprises and non-state-owned enterprises. The regression results of state-owned enterprises are shown in column (1) of

Table 7. Results of the heterogeneity of the enterprise dimension.

Variables	(1)	(2)	(3)	(4)
	State-Owned Enterprise	Non-State-Owned Enterprise	Large Enterprise	Medium and Small-sized Enterprise
confirmed × after	−0.1	−0.069 ***	−0.067 ***	−0.14
	(−0.78)	(−4.57)	(−3.31)	(−0.70)
Constant	−5.112	−15.854 *	−7.011	−9.931
	(−0.55)	(−1.78)	(−1.04)	(−0.43)
Controls	YES	YES	YES	YES
Company FE	YES	YES	YES	YES
Year FE	YES	YES	YES	YES
Observations	993	1283	1916	339
R-squared	0.723	0.738	0.709	0.737

Note: *** and * are significant at the levels of 1% and 10%, respectively.

. The coefficient of the interaction term confirmed × after is negative, but it is not statistically significant. The regression results of non-state-owned enterprises are shown in column (2) of Table 7. The coefficient of the interaction term is negative and significant at 1%. Under the influence of the COVID-19 epidemic, the fiscal pressure of the local governments has a significant negative impact on the environmental protection investment of non-state-owned enterprises, but the impact on state-owned enterprises is not significant.

This paper mainly analyzed the possible reasons from the following two aspects. On the one hand, there are differences in environmental protection investment capacity. The close relationship between state-owned enterprises and governments affords enterprises a comparative advantage in financing. For state-owned enterprises, the market competition mechanism is not completely applicable. Therefore, state-owned enterprises are more capable of investing in environmental protection. Compared with state-owned enterprises, non-state-owned enterprises are at a disadvantage in terms of financing and market competition, and their ability to invest in environmental protection is weak.

On the other hand, there are differences in environmental protection investment motivation. With the growing awareness of environmental protection, China has incorporated environmental protection indicators into the performance evaluation system of local officials. In order to achieve political and economic goals, the local governments can intervene in the pollution behavior of state-owned enterprises that are closely related to the government. Therefore, the “altruistic” motivation of state-owned enterprises is stronger. State-owned enterprises adopt energy conservation and emission reduction as their main tasks to meet the requirements of green development, and have more incentives to carry out environmental management, promote green technology and product innovation, and reduce pollution emissions. However, non-state-owned enterprises are more likely to maximize their profit and have a stronger “self-interest” motivation. Therefore, the increasing local fiscal pressure leads to a sharp decline in the environmental protection investment of non-state-owned enterprises.

5.2. Enterprise Size

To further test the robustness of the regression results, this paper grouped enterprises by scale. According to the latest enterprise classification standard of the Ministry of Industry and Information Technology, enterprises are divided into small- and medium-sized enterprises and large enterprises. Among them, those with less than 1000 employees and an operating income of less than CNY 2 billion are small- and medium-sized enterprises; otherwise, they are large enterprises. The regression results of large enterprises are shown in column (3) of Table 7. The coefficient of the interaction term confirmed × after is negative and significant at the level of 1%. The regression results of the small- and medium-sized enterprise samples are shown in column (4) of Table 7. The coefficient of the interaction term confirmed × after is negative but not statistically significant. In other words, the fiscal pressure of local governments significantly reduces the environmental protection investment of large enterprises, but it has no obvious impact on small- and medium-sized enterprises.

The possible reason is that, on the one hand, large enterprises contribute with a higher industrial output value and tax revenue in the process of production and operation, so they are more likely to be affected by local government activities. On the other hand, the production activities of large enterprises are more frequent, and the motivation for pollution discharge is stronger. Therefore, large enterprises are more sensitive to the choice of government behavior. Small- and medium-sized enterprises may be more limited by the scale of their production activities than by external factors, such as government activities. Therefore, when local governments face fiscal pressure, the behavior choices of increasing revenue and reducing expenditure have a more significant impact on the environmental protection investment decisions of large enterprises.

5.3. Environmental Regulation

Environmental regulation is an important factor forcing enterprises to participate in environmental governance. Referring to Liu et al. (2021) [55], this paper divided cities into “two-control-zone” cities and “non-two-control-zone” cities. The regression results of the enterprises located in the “two-control-zone” cities are shown in column (1) of

Table 8. Results of the heterogeneity of the city dimension.

Variables	(1)	(2)	(3)	(4)
	Two-Control Zones	Non-Two-Control Zones	High Fiscal Self-Sufficiency Rate	Low Fiscal Self-Sufficiency Rate
confirmed × after	−0.064 ***	−0.195	0.049	−0.191 *
	(−4.14)	(−1.48)	(0.20)	(−1.85)
Constant	−7.751	−20.509	−27.434 **	0.739
	(−1.04)	(−1.59)	(−2.26)	(0.10)
Controls	YES	YES	YES	YES
Company FE	YES	YES	YES	YES
Year FE	YES	YES	YES	YES
Observations	1864	423	1127	1123
R-squared	0.752	0.637	0.759	0.708

Note: ***, ** and * are significant at the levels of 1%, 5% and 10%, respectively.

, and the coefficient of the interaction term confirmed × after is negative, which is significant at the level of 1%. The regression results of enterprises located in “non-two-control-zone” cities are shown in column (2) of Table 8. The coefficient is negative, but it is not statistically significant. In other words, the fiscal pressure of local governments under the impact of the epidemic has a significant negative impact on the environmental protection investment of enterprises located in the “two-control-zone” cities, but has little impact on the enterprises located in the “non-two-control-zone” cities.

A possible reason is that different cities face different levels of environmental regulation. Enterprises located in the “two-control-zone” cities face stronger environmental regulations. Strict environmental supervision and law enforcement force enterprises to increase investment in environmental protection, which leads to the amount of environmental protection investment of enterprises greatly exceeding their desired level. When local governments face great fiscal pressure, regional economic development may become the priority target compared to environmental protection. Therefore, the local governments relax their environmental supervision. For the “two-control-zone” cities, there is more room for local governments to relax environmental supervision. Once the local governments relax environmental supervision, the enterprises located in the “two-control-zone” cities respond quickly, reducing the amount of environmental protection investment to their initial desirable level to maximize profit. Therefore, local fiscal pressure has a more significant impact on the decision-making process of environmental protection investment of enterprises located in the “two-control-zone” cities.

5.4. Fiscal Self-Sufficiency Rate

Due to differences in the level of economic development and other factors, there are significant differences in fiscal revenue and expenditure gap and fiscal self-sufficiency rate among cities in China [56,57,58]. According to the median fiscal self-sufficiency rate of the cities where the enterprises are registered, the samples were divided into high fiscal self-sufficiency rates and low fiscal self-sufficiency rates. Referring to Gu et al. (2022) [59], the fiscal self-sufficiency rate was defined as the ratio of local fiscal revenue to expenditure. The regression results of enterprises located in cities with high fiscal self-sufficiency rates are shown in column (3) of Table 8. The coefficient of the interaction term confirmed × after is negative but not statistically significant. The regression results of enterprises located in cities with low fiscal self-sufficiency rates are shown in column (4) of Table 8. The estimation coefficient of the interaction term confirmed × after is significantly negative. In other words, local fiscal pressure has a significant negative impact on the environmental protection investment of enterprises located in cities with low fiscal self-sufficiency rates, but has no significant impact on the enterprises located in cities with high fiscal self-sufficiency rates.

Prior studies defined fiscal self-sufficiency as the capability of local governments to raise their own income to fund their functions rather than relying on transfers from the central government [56,60]. Fiscal self-sufficiency rate is an indicator that is often used to measure decentralization [61]. According to fiscal federalism theories, a high fiscal self-sufficiency rate means that local governments have higher economic prosperity, greater fiscal capacity and lower dependence on the transfers from the central government. Local governments with abundant financial resources are more capable of coordinating economic development and environmental protection. Therefore, local fiscal pressure has no significant impact on the environmental protection investment of enterprises located in cities with high fiscal self-sufficiency rates. On the contrary, local governments without fiscal sufficiency struggle to address issues of capital independently. They must rely on transfer payments to boost the economy, but it reduces the efficiency of public service provision [58]. Therefore, when faced with fiscal pressure, local governments with low fiscal self-sufficiency rates are more motivated to relax environmental regulations and transfer their fiscal pressure to enterprises, thus having a more significant negative impact on enterprises’ environmental investment within their jurisdictions.

6. Discussion

In terms of research methods, this paper used the continuous DID method to explore whether and how local fiscal pressure influences the environmental protection investment of enterprises. The continuous DID method has been widely used in studies of fiscal pressure [16,20,22]. The difference is that we used the COVID-19 epidemic as an exogenous shock to reflect the sharp change in local fiscal pressure. On the one hand, due to the impact of COVID-19, local fiscal pressure increased significantly. On the other hand, before the outbreak of COVID-19, no one could have predicted its appearance. Therefore, the choice of using COVID-19 as an exogenous shock of local fiscal pressure in this paper is reasonable.

For the research conclusion, this study found that local fiscal pressure has a significant negative impact on enterprise environmental protection investment. It is closely related to the research by Zhong et al. (2022) [62], who found that enterprises tend to crowd out their investment in environmental conservation when local governments are under high pressure to address growth targets [62]. Although the research perspectives are different, we reached similar conclusions. When local governments face economic pressure, whether it is fiscal or economic growth target pressures, government intervention will hinder the environmental performance of enterprises. Therefore, this study provided a deeper insight into the role of government intervention in enterprise environmental investment decisions.

7. Conclusions and Policy Implications

Based on the data of A-share-listed companies in heavily polluting industries in China from 2015 to 2021, this paper used COVID-19 as the exogenous shock of local fiscal pressure and investigated the impact of local fiscal pressure on enterprise environmental protection investment by the continuous DID method. The study found that local fiscal pressure significantly reduced the environmental protection investment of enterprises. The mechanism of analysis showed that local governments alleviate their fiscal pressure through “increasing revenue” and “cutting expenditure”. The former increases the non-tax burden of enterprises, while the latter reduces the environmental protection subsidy of enterprises, which leads to the decrease in the environmental protection investment of enterprises.

To further analyze the results, we analyzed the heterogeneity of two dimensions: enterprise and city. From the enterprise perspective, local fiscal pressure has a significant negative impact on environmental protection investment of non-state-owned enterprises and large enterprises. From the city perspective, local fiscal pressure leads to a significant decline in the environmental protection investment of enterprises located in the “two-control-zone” cities and cities with low self-sufficiency rates. According to the research conclusions of this paper, the following policy suggestions are proposed.

First, local fiscal pressure should be relieved. In order to increase green investment and achieve green development, China should focus on solving the local fiscal pressure dilemma. On the one hand, issuing local government debts is a feasible way to alleviate local fiscal pressure. On the other hand, reasonably dividing the fiscal resources and expenditure responsibilities of the central and local governments and establishing a unified fiscal relationship between the central and local governments are necessary.

Second, the results may provide insights for policymakers to focus more on non-state-owned enterprises and large enterprises. They tend to avoid environmental responsibility more than state-owned enterprises and small- and medium-sized enterprises. Policies and action plans should be proposed to encourage and supervise such enterprises’ green investment behaviors. As the negative effect is more significant for enterprises located in “two-control-zone” cities, policymakers must balance the relationship between economic development and supervising enterprises’ environmental activities. Finally, the negative impact is significant for enterprises located in cities with low fiscal self-sufficiency rates. Improving the government’s fiscal self-sufficiency rate is conducive to achieving coordinated economic and environmental development.

There are also some limitations to this study. First, this paper only used China’s A-share-listed companies in heavily polluting industries as a sample. Although the research data were sufficient to support the research conclusions, future studies can select samples from other industries to further enhance the representativeness of the conclusions. Second, this paper only explored the impact of local fiscal pressure on enterprises’ environmental protection investment. The impact on environmental performance and financial performance can be further investigated in the future.