Do Firms That Are Disadvantaged by Unilateral Climate Policy Receive Compensation? Evidence from China’s Energy-Saving Quota Policy

Abstract

Inequities caused by a unilateral climate policy may threaten the sustainability of CO₂ emission reduction efforts by countries and firms, thus endangering sustainable development for humans and the eco-environment. However, few studies have conducted ex-post evaluations on whether environmentally regulated firms receive external compensation such as subsidies, tax reductions, and loan support. Thus, this study investigates whether firms experiencing inequitable conditions under China’s Energy-Saving Quota Policy (ESQP) are financially compensated. It develops a balanced panel of data from 6189 ESQP-regulated and 6189 unregulated firms from 2010 to 2013, and combines a probit model with the difference-in-differences method to conduct empirical analysis. The results show that ESQP-regulated firms receive more subsidy income and lower tax rates than unregulated firms. Of the ESQP-regulated firms, companies with higher energy-saving burdens receive larger subsidies and lower financial expense ratios than those with lower burdens. Additionally, firms that complete their energy-saving quotas are compensated with larger subsidies and/or lower financial expense ratios and tax rates than those that fail to complete them. Finally, state-owned firms receive more subsidies than private ones. Unlike the emission trading schemes implemented worldwide that formulate an exemption mechanism (i.e., free or over-allocated allowances), the ESQP does not exempt regulated firms from their energy-saving responsibilities. Rather, regulated firms receive a greater amount of external compensation in exchange for their reductions in energy consumption.

1. Introduction

Global warming caused by CO₂ emissions has resulted in drought, flooding, and other dreadful consequences, which endangers the sustainable development for humans and the eco-environment. Minimizing energy consumption can reduce CO₂ emissions and support the mitigation of global warming and the achievement of sustainable development goals. Many countries have been trying to reduce carbon dioxide emissions through energy conservation policies, such as carbon taxes in Denmark and most other European countries, China’s energy-saving and low-carbon action of 16,078 firms (or institutions). Climate change caused by CO₂ emissions is a global problem, but climate change policies are mostly unilateral or sub-global [1]. Policymakers and industry representatives are concerned that these policies may produce at least two types of inequity [2,3]. First, energy-intensive and trade-exposed (EITE) firms may lose competitiveness in world markets due to policies inducing higher costs compared to their international rivals from countries with less stringent climate policies [4,5]. We refer to this phenomenon as international inequity. Second, firms regulated by a climate policy may lose their comparative advantage in domestic markets when competing with unregulated firms in the same industry [6,7]. We call this phenomenon domestic inequity. Domestic enterprises may move abroad if these inequities are not rectified, thereby resulting in the loss of GDP and welfare, and threatening the sustainability of CO₂ emission reduction efforts by specific countries and firms. It is, therefore, important to study whether enterprises regulated by a climate policy receive more compensation, the findings of which can provide important policy implications.

Compensatory policies to mitigate the inequities of unilateral climate policies have been subject to intense political and academic debates [8]. A large number of important climate policies have compensation mechanisms in place [9]; almost all countries design and constantly update their policy packages. These include: free or over-allocated emission allowances in the European Union (EU) Emission Trading Schemes (ETSs) [10]; subsidies for renewable energy in the United States [11]; soft loans, tendering, and other incentive programs in Germany [12]; Salix loans and the Industrial Emissions Directive in the United Kingdom [12]; subsidies for investments of abatement equipment in France [13]; tax reductions [14]; restricted entry of competitors [15]; and environmental funds or grants in countries such as the United States, China, Japan, Korea, France, Denmark, Russia, and Australia [16,17]. These compensatory policies (with the exception of free- or over-allocation, which only applies to ETS-regulated firms) can be enjoyed by every enterprise based on its energy conservation and emission abatement activities. In order to optimize policy packages, policymakers need to determine whether enterprises that are regulated by a special climate policy can obtain more financial incentives than unregulated ones. To date, there are few empirical studies investigating this question.

Previous studies have confirmed that it is efficient to mitigate the inequities caused by emission regulations by combining them with compensatory instruments. Scholars have looked at the combination of an ETS or emission tax with subsidies for research and development (R&D) activities on energy efficiency or renewable sources [18,19]. They have also examined the EU ETS and free/over-allocated emissions allow for exemptions [20]. These studies can provide suggestions for compensatory policies implemented in a unilateral climate policy, but cannot ease the worries of regulated firms by telling them they can get more compensation, mainly because scholars have not examined whether firms that are regulated by a unilateral climate policy receive more compensation compared to unregulated firms.

Our study seeks to fill this gap by studying the Energy-Saving Quota Policy (ESQP) in China. This policy assigned an energy-saving quota to 16,078 firms from 2011 to 2015. We introduced four types of inequities caused by this unilateral policy and designed three proxy variables for compensatory policies (subsidy, financial expense ratio, and tax rate). Then, we analyzed a balanced panel data of 12,378 firms (including 6189 regulated and 6189 unregulated firms) from the manufacturing sectors in China for the period from 2010 to 2013. We used a matching technique combined with a difference-in-differences estimator to search for the most similar firms from unregulated ones for each of the ESQP-regulated firms, which can overcome the endogeneity problem and the self-selection bias. The results show that the ESQP-regulated firms received larger subsidies and paid lower tax rates than the unregulated firms. Within the ESQP-regulated firms, state-owned firms, firms with high environmental burdens, firms that fully met the energy-saving quota requirements, and exporting firms obtained more compensation.

This article provides additional insights into the inequity issue of climate policy. First, drawing on Chinese firm-level data, we conduct an ex-post analysis of whether firms that are regulated by a climate policy can receive more external assistance, such as subsidies, tax reductions, and loan support. The mathematical analysis and numerical simulations conducted in ex ante studies cannot reveal whether regulated firms are better compensated than unregulated ones with certainty. We overcame this deficiency by determining the differences in compensation between regulated and unregulated firms after the implementation of the ESQP, which can provide more reliable conclusions for policymakers and enterprises. Second, based on the unique data constructed by combining the three databases, we further investigated the inequities within regulated firms. These include the inequities between high-burden and low-burden firms, between “complete” (fully meeting the quota) and “incomplete” firms, and between state-owned and private firms. These three inequities have not been fully discussed in the previous literature. Our results indicate that the firms that are disadvantaged by the three inequity types do receive more compensation, which has implications for regulated firms’ design of abatement strategies. Third, while most existing studies on the inequities caused by environmental policies were conducted in developed countries, this study focuses on China, the world’s largest GHG emitter, with a rapidly growing economy. Thus, our findings can be applied to other developing countries that are implementing unilateral climate policies and compensatory policies. China’s attempts to enhance local governments’ motivation for environmental protection can serve as important lessons learned. These include the incorporation of the energy consumption index into the performance evaluation index system for local officials as well as the central environmental supervision of local governments.

The remainder of this study is organized as follows. Section 2 provides a literature review. Section 3 elaborates on the background of the ESQP, describes the four inequities induced by the ESQP and compensatory policies, describes the datasets and variables used in this study, and describes the difference-in-differences matching estimator. In Section 4, we report and discuss the empirical results. Conclusions and their implications follow in Section 5.

2. Literature Review

A growing body of academic literature has examined whether inclusive and complementary policies can alleviate the domestic and international inequities caused by a unilateral climate policy.

The first issue addressed by scholars is whether compensatory policies can offset the disadvantages experienced by firms subject to climate policies. Scholars introduce the general equilibrium model, mathematical simulation method, and other policy simulation methods to simulate the economic and welfare effects under different policy scenarios, then judge the necessity and feasibility of designing compensation policies by comparing the policy effects of every scenario. The policy scenarios designed by scholars include but are not limited to the following: only implementing a climate policy (e.g., energy tax); scenario with a climate policy and subsidies (or a refunding scheme or one of other supportive policies); combination of a climate policy and multiple supportive policies. Obviously, the measures introduced as compensation policies have been increased and refined. Fischer and Newell [21] suggest that an optimal portfolio of policies (such as an emissions trading scheme, a renewable energy subsidy, and an R&D subsidy) could help the United States electricity production sector achieve emissions abatement at a significantly lower cost than any single policy. Cato [22] argues that a policy mix combining an emission tax, a refunding scheme, and an entry-license tax can effectively alleviate inequities. Fischer [23] find that an emissions tax combined with output-based refunding can alleviate the output and under-provision of regulated firms in an imperfect market. Fischer and Fox [24] consider four policies that could be combined with unilateral emissions pricing to counter effects on international competitiveness, including a border rebate for exports and domestic output-based rebating. Christiansen and Smith [25] confirm that an emissions tax combined with the direct regulation of abatement technology (or an abatement technology investment subsidy) can help reduce the firm’s uncertainty regarding abatement costs. Ćetković and Buzogány [26] claim that energy and climate policies should be integrated with industrial and innovation policies. The conclusions of these studies, however, are based on mathematical analysis. Based on the computable general equilibrium (CGE) model and GTAP-E data, Böhringer et al. [27] suggest that industry exemptions are effective in mitigating inequities. Antimiani et al. [28] use a dynamic CGE model to assess the leakage-avoiding and competitiveness-protecting effects when supplementing an ETS (or energy tax) with subsidies for R&D activities in energy efficiency or renewable sources and find that the two effects are positive for energy-intensive industries in EU member countries. Böhringer et al. [2] show that this is welfare-improving for a country that implements emission pricing along with output-based rebates. Corradini et al. [29] conducted similar numerical simulations, confirming that the EU’s policy portfolio can help its unilateral carbon mitigation target without economic losses. Hagem et al. [30] demonstrate that both output-based and expenditure-based refunding lead to higher output and higher investments in the emission abatement equipment of polluting firms than the standard tax system. Kruse-Andersen and Srensen [19] point out that the optimal leakage-adjusted tax-subsidy scheme can create a social welfare accounting for about 0.5% of the national income compared to a single uniform emissions tax.

The results from the mathematical analysis and simulation experiments are suggestive, that is, they can inspire policymakers to supplement a unilateral climate policy with compensatory policies that place domestic and foreign companies on an equal footing [31]. For that, almost all countries have already designed and constantly updated their compensation mechanisms, and the practical guidance of these ex ante studies is very limited. In fact, policymakers pay more attention to whether firms that make more efforts to reduce energy consumption and CO₂ emission can get more compensation under the existing compensation mechanisms. Nevertheless, few ex-post empirical studies have been published on whether regulated firms do indeed receive more assistance outside the climate policy. Ex-post studies are important because they can provide evidence that would help policymakers to encourage firms to adhere to environmental regulations by telling them with conviction that they would receive more compensation. This being the case, our empirical study will systematically investigate whether environmentally regulated firms receive external compensation such as subsidies, tax reductions, and loan support.

The second issue addressed by scholars is whether specific complementary policies can reduce the inequities caused by unilateral CO₂ emissions policies. Taking the EU ETS as an example, several anti-leakage measures have been implemented or proposed, including the free allocation of emission allowances, Joint Implementation and Clean Development Mechanisms. Scholars have investigated the effects of the EU ETS on regulated firms’ employment [32], revenue [6], value added and profit margins [33], unit material costs [34], profitability [35], renewal of installed capital stock [36], and stock prices [37]. Surprisingly, scholars did not find any negative effects on firms’ competitiveness during the first three phases of the EU ETS. Other ex-post studies on all the ETS-regulated industries [9,38] and on the aluminum [39], oil refinery [40], steel [41], pulp-and-paper [42], and cement [20] industries similarly concluded that the EU ETS did not reduce international competitiveness. Scholars consider the over-allocation of free emissions allowances to be the main explanation for the absence of negative effects on competitiveness in the EU ETS20 [43]. This indicates that this complementary policy can successfully offset domestic and international inequities. ETS-regulated firms receive emission allowances free of charge based on their historical emissions (i.e., the grandfathering principle). Thus, they do not need to pay additional costs in the EU ETS (e.g., costs to buy permits or to take actions on production processes), which prevents them from losing their competitiveness [44]. Allevi et al. [20] and Genovese and Tvinnereim [38] further argue that allowance grandfathering can prevent ETS-regulated firms from relocating abroad. As over-allocation is an internal compensatory mechanism for an ETS, we cannot judge whether ETS-regulated firms are better compensated by external favorable policies [41].

In sum, much of the literature focuses on equity in international markets and equity for regulated industries and firms. In addition, a small number of studies have considered the inequities within the EU ETS-regulated firms, which have contributed to new perspectives. These studies constructed an allocation factor (measured by the quotient of allowances allocated to the verified emissions of EU ETS-regulated firms) to reflect the differences in abating pressure or compensation among regulated firms and to identify their economic effects in European countries [6,32,33,35]. Nevertheless, these studies only assess the inequity-mitigating effect of the internal compensation mechanisms (over-allocation). The question of whether enterprises with greater abatement pressure receive more outside compensation has not been fully answered.

3. Research Design

3.1. Energy-Saving Quota Policy

The Energy-Saving Quota Policy (ESQP), also known as the “Energy-saving and Low-carbon Action of 16,078 Firms (or Institutions)”, was proposed by the Chinese government in 2010, approved in December 2011, and became effective in May 2012. It was implemented as a classical command-and-control regulation for emission reductions, assigning energy-saving quotas to 16,078 firms and entities by 2015. The National Development and Reform Commission of China (NDRCC) allocated an energy-saving quota to each of the 31 provinces, which then allocated its own quota to each of the firms within its administrative boundaries, covered by the regulation of achieving the provincial goal mandated by the NDRCC. The total annual energy consumption of firms included in the ESQP was about 2000 million tonnes of standard coal equivalent (Mt SCE), covering around 87% of China’s industrial energy consumption and 60% of the total national energy consumption. Serving as China’s primary policy instrument to achieve its emission abatement targets, the ESQP’s goal was to reduce 250 Mt SCE by 2015, approximately 38% of China’s energy-saving target for the period 2011–2015. This goal was achieved by the end of 2013, indicating that the vast majority of regulated firms had taken effective energy-saving measures.

Similar to the EU ETS, there are several criteria for firms to be included in the ESQP. One criterion is total annual energy consumption. Only firms that consumed 10,000 t SCE or more in 2010 were included in the ESQP. This participation eligibility provides an opportunity to create treatment and control groups within the same industry in the following section. Additionally, the ESQP covered almost all energy-intensive industries across the 31 provinces of China, including power and heat generation, cement, iron and steel, oil refineries, coke ovens, glasses, lime, bricks, ceramics, and pulp and paper products.

In practice, the Chinese government required the regulated firms to sign an energy-saving responsibility letter with the local governments, and to divide the energy-saving quota into annual energy-saving goals, which would be verified each year. There is no exemption or free quotas for ESQP-regulated firms. This is different from the EU ETS, which provides free or over-allocated emission allowances as a compensation for the regulated firms. Complementary incentives were provided for the regulated firms that achieved their annual goals and penalties for those that failed to achieve their goals. Both financial and non-financial incentives or penalties were designed. For example, firms with outstanding achievements in energy conservation would be commended and rewarded with subsidies. Failing firms, on the other hand, could face at least one of three penalties. First, they would be ordered to meet their goals within a limited time period and disqualified from all awards and honorary titles; second, the evaluation and approval of their new high-energy-consuming projects would be postponed; third, they would be at a disadvantage in terms of credit rating, credit access, credit exit management, and loan delivery. Moreover, the achievement of the energy-saving quota was included in the performance evaluation of provincial leaders, which might spur local governments to vigorously compensate the ESQP-regulated firms to encourage them to enhance energy-saving activities.

3.2. Four Inequities Induced by the ESQP

The ESQP is a “command-and-control” regulation in China aiming to address the market failures (negative externalities) associated with carbon emissions. However, resistance to this regulation emerged as soon as the ESQP was implemented in 2012, in part because the policy would potentially be sensitive to free-riding [45]. Enterprises mainly complete energy-saving tasks or environmental responsibilities by purchasing or improving environmental protection equipment and production equipment, and improving production processes, and optimizing product design. Firms interviewed by us frequently complain about the negative impacts of these environmental costs on profitability and market competition. Moreover, the higher intensity of energy-saving tasks of regulated firms, the more capital investment they need, and the more disadvantaged they are in the market competition. Using surveys and other available data on regulated firms, we classified the inequities stemming from the ESQP into four categories.

The first type (Inequity I) is the inequity between regulated and unregulated firms. As mentioned previously, regulated firms may lose their competitiveness due to the additional abatement costs compared to unregulated firms in the same industry. The remaining three categories are inequities among the ESQP-regulated firms. The second inequity (Inequity II) arises because the energy-saving burdens (measured by the ratio of energy-saving quota to gross output in 2010) among the regulated firms are different. We divided the regulated firms into high-burden firms, whose energy-saving burdens are larger than the average of all firms’ energy-saving burdens, and low-burden firms for all other firms. Compared to the firms with a similar output or energy-saving quota, the high-burden firms will bear higher marginal costs because they require more investments in energy-saving technology than other firms. The third inequity (Inequity III) is between firms that fully meet their energy-saving quota requirements and the firms that fail to meet their quotas. In December 2014, the NDRCC announced that 14,119 firms had met their energy-saving targets for 2013 [46]. The remaining 1959 firms were not verified because they were restructured, closed, relocated, or eliminated. By 2013, 1191 firms were classified as “incomplete” based on the data released by the NDRCC [47]. We assume that the firms that met the quotas were subject to higher marginal costs than the firms that did not meet their quotas. The last inequity (Inequity IV), which may be unique to China, is between state-owned and private firms. In a survey of the regulated firms in the Fujian Province, most of the respondents in state-owned firms indicated that they were at a disadvantage as a result of the ESQP. The main reasons for this are as follows: The executives of state-owned firms are government officials appointed by the local (provincial) governments, who have political motivations (e.g., for personal promotion) to overstate their energy consumption and production. In contrast, private firms tend to under-report their energy consumption because they seek lower energy-saving quotas. Moreover, state-owned firms have traditionally enjoyed some privileges over private companies in terms of loan and market access, lower interest rates, and negotiation powers [48].

3.3. Compensatory Policies and Their Measurements

Firms’ efforts on reducing energy consumption and CO₂ emission have positive externalities. Compensations that offset firms’ additional environmental costs can promote firms to increase investment in energy conservation, thus better avoiding free-riding problems. As outlined in Section 3.1, policymakers have designed countermeasures to alleviate the inequities caused by the ESQP. These mainly include various subsidies, tax incentives, and loan support. These policies are common mitigation measures for environmental policy inequity in China as well as in other parts of the world [49]. In particular, tax incentives and loan support may help regulated firms to achieve lower tax rates and financial expense ratios. Accordingly, we employed subsidy, financial expense ratio and tax rate as the measurements for compensatory policies (see

Table 1. Definitions and sample means of the variables.

Variables		Definitions	2010		2013
			Control	Treated	Control	Treated
outcome variables	subsidy	log of subsidy income (CNY)	5.490	4.688	5.104	7.270
	financial expense ratio (%)	financial expense to total liability	3.332	3.120	5.432	4.405
	tax rate (%)	(sum of income tax, value-added tax and business tax) to main business revenue	1.280	1.357	1.685	1.651
control variables	lnoutput	log of total output	12.460	12.780	12.550	13.340
	age	2013—founding year	18.750	20.300	22.260	23.170
	lnlabor	log of number of employees	5.446	6.304	6.157	6.667
	tfp	total factor productivity calculated using Solow’s productivity residua	2.866	1.330	2.052	1.494
	lncapital	log of the ratio of total fixed assets to the number of employees	4.069	4.749	4.422	5.238
	lnexports	log of exports	2.999	3.084	3.591	3.456
	loa	ratio of total liabilities to total assets	0.630	0.652	0.550	0.602
	soa	ratio of revenue to total assets	5.583	8.259	2.319	1.771
	deficit	1 for a negative profit and 0 otherwise	0.096	0.122	0.147	0.219
	state-own	1 for state-owned firms and 0 otherwise	0.054	0.067	0.033	0.046
	foreign-own	1 for foreign firms and 0 otherwise	0.200	0.131	0.185	0.123
	HKMT-own	1 for firms belonging to a parent company in Hong Kong, Macao, and Taiwan and 0 otherwise	0.113	0.100	0.116	0.097
	lnpgdp	log of per capita GDP in each province of China
	lnfiscal	log of fiscal revenue per unit land area in each province of China
matching variables	lnoutput	log of total output	12.460	12.780
	state-own	1 for state-owned firms and 0 otherwise	0.054	0.067
	lnassets	log of total assets	11.160	11.540
	subsidy2008	log of subsidy income in 2008	2.110	2.443
	tax rate 2008-10 (%)	average of tax rate during 2008–2010	1.296	1.516

Notes: Descriptive statistics on matching variables are only presented in 2010, as the matching procedure was conducted on the data of firms in this year.

for detailed definitions).

Data released by the Ministry of Finance of the People’s Republic of China (MFPRC) showed that the financial expenditure on energy conservation and environmental protection reached CNY 744.4 billion (Chinese yuan) in 2019, of which the expenditure on energy conservation and CO₂ emission reductions increased by 48.6% over 2018 [50]. The ESQP-regulated firms may receive more subsidies than others, as they benefit from various incentive-based environmental policies and other preferential policies. In China, however, there is no firm-level database available on environmental- or energy-related subsidies, tax incentives, and loan support. In this paper, subsidy income is the sum of environmental subsidies (e.g., subsidies for R&D of energy-saving technology and environmental investments), production subsidies (e.g., subsidies for production equipment investments and technological innovations), R&D subsidies (e.g., subsidies for scientific research projects, patent application fees, and new product development), rewards or honor-related subsidies (e.g., brand awards, demonstration company awards, national new product awards), and export subsidies (e.g., cash incentives and export rebates).

Several preferential tax regimes were also designed by the Chinese government to encourage enterprises to save energy and reduce CO₂ emissions, including a preferential tax base, tax rate, tax amount, tax payment time, and tax deduction. For instance, 10% of the investments in special equipment for environmental protection and energy conservation can be offset from the enterprises’ current tax year. If the current tax year is insufficient for generating offsets, it can be carried forward in the next five years [51]. The income of enterprises engaged in projects that qualify as environmental protection and energy conservation are exempted from income tax from the first year they earn income to the third year, and are allowed a 50% reduction from the fourth year to the sixth [51]. Companies producing high-purity carbon dioxide products with industrial waste gas as a raw material can obtain an immediate refund of value-added tax. Unlike with other instruments, the preferential tax payment time (e.g., accelerated depreciation and delayed tax payment) cannot reduce enterprises’ total tax burden. However, it can speed up their recovery funds and ease financial difficulties.

Furthermore, China implemented a green credit policy in 2007 [52], which was enhanced by the Green Credit Guidelines released in 2012 [53]. This policy applies to commercial banks and other financial institutions. It restricts loans to companies and projects with a poor environmental performance, promotes loans for environmentally friendly or energy-saving firms or projects [54], and guarantees financing and discount-interest loans. Data published by the People’s Bank of China (PBC) showed that the loans provided by commercial banks towards major projects to improve environmental quality, save energy, and reduce emissions amounted to CNY 10.22 trillion in 2019 [55]. The green credit policy may help ESQP-regulated firms to achieve easier and cheaper access to loans.

In practice, Chinese government branches at all levels are accustomed to allocating resources to firms regulated by a special policy. That is, ESQP-regulated firms may receive more environment-related subsidies, as well as other subsidies, such as those for R&D and production. Thus, it is reasonable to carry out an empirical analysis based on data on firms’ total subsidy income (overall tax rate and financial expense ratio) rather than their environment-related subsidy income (tax incentives and loan support), even if firm-level data on environmental- or energy-related subsidies, tax incentives, and loan support can be obtained. To summarize, we considered ESQP-regulated firms to receive compensation as long as they received more subsidies and had lower financial expense ratios or tax rates than unregulated firms. Therefore, our research puts forward the following hypotheses:

Hypothesis 1 (H1).

The ESQP-regulated firms receive compensation.

Hypothesis 2 (H2).

Among the regulated firms, the firms with a higher energy-saving burden or more active implementation of energy conservation activities receive more compensation than firms with a lower burden or less active implementation.

3.4. Methods and Variables

There are regulated and unregulated firms within a regulated industry, and data on these firms are available for both pre- and post-ESQP periods, allowing for a semi-natural experiment to be mimicked, using the observational data to test the effects of the ESQP based on comparisons between regulated and unregulated firms and between the pre- and post-ESQP periods [56]. We applied a difference-in-differences matching technique to achieve our research objectives.

3.4.1. Difference-in-Differences Model

We used dummy variables to distinguish the before/after and treatment/control structure of the semi-natural experiment. The difference-in-differences model can be written as:

$$COM{P}_{it}={\alpha }_0^{}+{\alpha }_1^{}trea{t}_i^{}\times tim{e}_t^{}+\beta Z+{\alpha }_i^{}+{\delta }_t^{}+{\epsilon }_{it}^{}$$

(1)

where COMP_it is the outcome variable representing the subsidy, financial expense ratio, and tax rate received by firm i in year t; treat is a binary variable indicating whether a firm is regulated by the ESQP (1 for the regulated firms, the treatment group and 0 for the unregulated firm, the control group); time is also a binary variable with 1 for the year 2012–2013 (post-ESQP) and 0 for the year 2010–2011 (pre-ESQP); ε_it is the error term and assumed to be mean zero; α_i is the firm’s fixed effect adjusted by all constant, unobserved determinants of firm i; δ_t is the year’s fixed effect; α₁ is the effect of the ESQP, i.e., the average treatment effect. To control for unobservable firm characteristics, we estimated this difference-in-difference model as a fixed-effects (FE) model. To consider observed and unobserved heterogeneities across the regulated and unregulated firms, we also added a set of control variables Z: lnoutput, age, lnlabor, tfp, lncapital, lnexports, loa, soa, deficit, state-own, foreign-own, HKMT-own, lnpgdp, and lnfiscal [57,58,59,60,61]; see Table 1 for detailed definitions. Note that tfp was used to capture firms’ innovation ability [62], because we could not obtain the data on R&D and patent fees.

The validity of the difference-in-differences model relies on the parallel trend assumption, that is, the trends of the outcome variable in the sample period (2010–2013) would be the same for both the treatment and control groups in the absence of the ESQP regulation. The assumption is most plausible when the control and treatment groups are very similar in the pre-ESQP period (2010–2011). As such, we constructed the control group by including all other firms in the same two-digit industry as the treatment group, since they share common industrial characteristics [35]. The treatment group may have stronger operational abilities than the control group. If this is the case, the COMP_it values of the treatment and control groups may differ systematically, even in the absence of the ESQP [63,64]. This difference in trends is referred to as the self-selection bias that is expected to provide upward-biased estimates for the competitiveness effect of the ESQP.

3.4.2. Propensity Score Matching

We used propensity score matching to reduce the selection bias and validate the parallel trend assumption. The self-selection bias was reduced when the unregulated firms were as similar as possible to the regulated firms. Thus, the one-to-one nearest neighbor matching without replacement was applied to find the nearest sample of the unregulated firms (control group) for each of the regulated firms (treatment group) in the same industry. That is, each regulated firm was matched to an unregulated firm with the minimal distance in the propensity score [65]. Essentially, the propensity score, denoted by P_i, is the estimated probability of being regulated by the ESQP, $P(X)=P\mathrm{r}(D=1\left|X\right.)$, for a given set of observable characteristics, X, of regulated and unregulated firms in 2010. These characteristics are the matching variables in Table 1. The matching between regulated and unregulated firms from 2011 to 2013 followed the matching results in 2010. We estimated the propensity score and match firms on pre-treatment characteristics (2010–2011), to obtain covariates that were not influenced by the treatment. Finally, we estimated the propensity score of every firm by using a logit regression on the following variables (X): lnoutput, lnassets, subsidy2008, tax rate2008-10, and state ownership (see Table 1 for detailed definitions). The reasons for designing the above variables are as follows: (1) size is an important factor for inclusion in the ESQP, while firms receiving more subsidies and incentives in the past may be required to assume more environmental responsibilities; and (2) state-owned firms are more likely to be included in climate policies, as mentioned in Section 3.1.

The validity of the estimation equation for the propensity score lays a foundation for satisfying the conditional independence assumption [64]. In other words, after the matching based on the propensity score, the difference between the values of the observed covariates for treated firms and control firms is not statistically significant [65]. If this assumption is false, the equation measuring the propensity score should be reset. We carried out a balance test to assess the quality of the matching by verifying whether there was no significant difference in the average value of the matching variables between the treatment and control groups, and by calculating the standard bias of the matching variables before and after matching. As shown in Figure 1, the matching result is sufficient because the matching variables’ absolute value of standard bias remains below 10% [65]. In addition, we conducted a common support test to ensure that matching was performed on a common support, where the distributions of the propensity score in the treatment and control groups overlap.

We conducted propensity score matching by industry, so there were 30 estimation results of the logit regression model, which was used to estimate the propensity score. These estimation results are not provided, but some evidence is provided to judge the matching quality. First, according to the kernel density distribution of the propensity score shown in Figure 1, there was a significant improvement in the balance between the treated and control groups after matching in the pre-ESQP period (only in 2010). Second, Table 1 demonstrates that the observed values of five matching variables in the treated and control firms are very close after matching.

Figure 2 again shows almost no difference in the mean value of matching variables between the two groups after matching, as the absolute standard bias of the four matching variables is less than 10%. Thus, the matching technique reduces the sample selection bias and ensures that we estimated the DID estimator under parallel trends assumptions. Thus, we believe that the estimated results of the DID model based on the sample data constructed by us are reliable.

3.4.3. Procedures for Testing Compensation for Inequities Ⅱ~Ⅳ

Here, we take inequity Ⅱ (the difference in energy-saving burden) as an example to explain the steps for testing whether firms are compensated for ESQP-induced inequities. First, the energy-saving burden of an ESQP-regulated firm is measured as the ratio of its energy-saving quota to its gross output annually, from 2010 to 2013. Second, a dummy variable, ESB, was introduced to separate the high-burden firms (ESB = 1), whose energy-saving burdens are larger than the average of all firms, from the other firms (ESB = 0). Thus, ESB×ESQP depicts the interaction between the dummy ESB and ESQP (treat×time). Third, treat×time in Equation (1) was substituted by ESB×ESQP. Fourth, ESB was regarded as a moderating variable that may affect the amount of compensation received by ESQP-regulated firms. If the coefficient of ESB×ESQP on subsidy (financial expense ratio or tax rate) is significantly positive, we can claim that high-burden firms receive more compensation.

Similarly, we used another dummy variable, complete, to distinguish the firms that accomplished their energy-saving quotas by 2013 (complete = 1) from the “incomplete” firms that failed to accomplish their quotas (complete = 0). Furthermore, two interactive items, complete×ESQP between complete and ESQP and state-own×ESQP between state-own and ESQP, were included in the model. This allowed us to investigate the moderating effects of the dummy variables, complete and state-own, respectively.

3.5. Data

We estimated our econometric model using a dataset that we constructed by combining three main databases. The first and central database is the “List of Energy Saving Task of Each Firm (or institution) in Energy Saving and Low Carbon Action (i.e., the ESQP-regulated firms)”, which provides information regarding firms’ names, unique legal-person codes (similar to taxpayer identifiers) and energy-saving tasks or quotas [46]. The second database is the list of 1191 “incomplete” firms that failed to meet their energy-saving quotas by the end of 2013, as reported by the NDRCC [47]. These firms are merged with the first database to distinguish the firms that completed their tasks from the others that were not completing their tasks. Third, we used the information from the Chinese Industrial Enterprise Database (CIED) to add company data to the above dataset [66]. CIED covering 1998–2013 includes all state-owned firms and private firms with an annual main revenue of over CNY 5 million during 1998–2010 (increasing to 20 million CNY in other years) [67], which was collected by the National Bureau of Statistics of China (NBSC) and widely used by scholars due to its large sample. CIED contains data on a variety of indicators such as gross output, revenue, subsidy income, and income tax for individual firms. We merged the CIED data with the ESQP data to identify firms that are regulated by the ESQP. As the data on subsidy income for 2009 and 2010 are missing, we used the 2008 subsidy income to substitute for that in 2010. In addition, we kept the manufacturing firms whose observed values of gross output, revenue, total assets, total fixed assets, and number of employees were greater than 0.

We obtained a total of 6189 ESQP-regulated firms and 111,211 unregulated firms after balancing the unbalanced panel data from 2010 to 2013. To ensure that an industry has sufficient samples to scientifically estimate the predicted probability (propensity score) of being an ESQP-regulated firm, we only kept the enterprises in 30 manufacturing industries (see Table 6 for details). Then, we chose the unregulated firm that best resembled each regulated firm based on one-to-one nearest-neighbor matching without replacement. Finally, our operative sample for fitting the difference-in-differences model was composed of 6189 ESQP-regulated firms and 6189 unregulated firms observed over the period 2010–2013. Data on the fiscal revenue, land area, and per capita GDP of each province in China were obtained from the Statistical Yearbook of China.

The detailed definitions and descriptive statistics of the control and treatment groups, and post-ESQP and pre-ESQP, are shown in Table 1. Except for two matching variables such as lnoutput and state-own, there were a number of differences in control variables between the control and treatment groups. Hence, it is important to control for these underlying differences, since they may affect firms acquiring compensation. In addition, matching variables showed no significant difference in the mean value between treatment groups and control groups after matching, indicating that the sample selection bias was overcome.

4. Results and Discussions

4.1. Average Compensation Received by ESQP-Regulated and Unregulated Firms

As shown in

Table 2. Compensation of ESQP-regulated and unregulated firms.

	Subsidized Firms		Average Subsidy Income of Subsidized Firms (CNY 104)		Subsidy		Financial Expense Ratio (%)		Tax Rate (%)
	Control	Treated	Control	Treated	Control	Treated	Control	Treated	Control	Treated
2010	3034	2017	224	692	5.490	4.688	3.332	3.120	1.280	1.357
2011	2997	4012	154	364	4.558	6.058	5.162	4.098	2.012	3.298
2012	3219	4466	267	657	5.147	7.185	6.066	4.747	3.500	3.654
2013	3201	4489	269	604	5.104	7.270	5.432	4.405	1.685	1.651

, when viewed as a whole, the ESQP-regulated firms (treated) were more likely to receive subsidies and to be subsidized more, as their average subsidy income was greater than that of the unregulated firms (control) during 2010–2013. The above disparities were further widened in the post-ESQP era. This phenomenon also occurred in the outcome variables, such as subsidy, tax rate, and financial expense ratio. Thus, the ESQP seemed to help regulated firms to obtain more compensation than unregulated firms. Of course, this needs to be further verified by subsequent econometric analysis. Moreover, the “trends” of the three outcome variables were almost the same for both the treated and control firms, suggesting the validity of the difference-in-differences method.

As shown in

Table 3. Compensation received by different sub-groups of ESQP-regulated firms.

Subgroup	Sample Size	Subsidy Income (CNY 104)	Financial Expense Ratio (%)	Tax Rate (%)
high-burden firms	761	492	2.891	2.529
low-burden firms	5428	218	4.824	2.673
“complete” firms	452	465	4.526	2.673
“incomplete” firms	5737	358	5.212	2.395
state-owned firms	385 (in 2012) 284 (in 2013)	1070	2.355	5.631
private firms	5804 (in 2012) 5905 (in 2013)	424	4.702	2.482
exporting firms	1841 (in 2012) 1816 (in 2013)	901	3.254	2.135
non-exporting firms	4348 (in 2012) 4373 (in 2013)	270	5.130	2.872

Notes: The figures above are the means of sample values for 2012 and 2013. The numbers of high-burden firms and “complete” firms remain unchanged for the two-year period.

, within the ESQP-regulated firms, high-burden firms, on average, received more subsidies and lower tax rates and financial expense ratios than low-burden firms. Exporting firms enjoyed the same privileges compared to non-exporting firms. In addition, state-owned firms obtained more subsidies and had lower financial expense ratios but paid higher tax rates than private companies. Those that did not completely meet the quota requirement were punished with smaller subsidies and higher tax rates and financial expense ratios.

4.2. Compensation for Inequity Ⅰ between Regulated and Unregulated Firms

Table 4. Estimated results of compensation for Inequity I between regulated and unregulated firms.

	Subsidy	Financial Expense Ratio	Tax Rate
treat×time (the ESQP)	0.677 (0.043) ***	−0.280 (0.578)	−0.306(0.096) ***
lnoutput	0.125 (0.025) ***	−1.476 (0.334) ***	−0.138(0.055) **
lnasset	−0.097 (0.036) ***	−1.267 (0.484) ***	0.590 (0.080) ***
age	−0.002 (0.002)	−0.001 (0.020)	0.002 (0.003)
lnlabor	−0.131 (0.034) ***	1.070 (0.455) **	−0.018 (0.075)
tfp	−0.001 (0.001)	0.004 (0.017)	−0.002 (0.003)
lncapint	−0.152 (0.024) ***	0.921 (0.317) ***	0.170 (0.053) ***
lnexports	0.011 (0.007)	0.075 (0.089)	−0.036 (0.015) **
loa	0.003 (0.011)	−0.572 (0.148) ***	−0.047 (0.025) *
soa	−0.001 (0.001)	−0.001 (0.007)	0.001 (0.001)
deficit	−0.158 (0.049) ***	−0.493 (0.655)	−0.852 (0.109) ***
state-own	−0.046 (0.128)	−0.560 (1.733)	0.790 (0.287) ***
foreign-own	−0.049 (0.155)	0.512 (2.094)	0.222 (0.347)
HKMT-own	0.030 (0.160)	0.098 (2.159)	0.298 (0.358)
lnpgdp	−5.317 (0.528) ***	16.860 (7.126) **	−0.210 (1.180)
lnfiscal	2.615 (0.330) ***	−6.434 (4.452)	0.593 (0.737)
_cons	22.220 (−1.718) ***	−59.64 0(−23.180) **	−9.961 (−3.840) ***
Number of obs	49,512	49,512	49,512

Noting: Standard errors in parentheses; *, **, *** indicate significance at the 10%, 5%, and 1% levels, respectively.

shows the results estimated using a balanced panel dataset of 6189 ESQP-regulated and 6189 unregulated firms from 2010 to 2013. The estimated coefficients of treat×time (the ESQP) on firms’ subsidy and tax rate are 0.677 and −0.306, respectively. Both are significant at the 99% confidence level, whereas the effect of ESQP on the financial expense ratio is −0.280 but insignificant. The results suggest that ESQP-regulated firms receive larger subsidies and lower tax rates compared to unregulated firms, if everything else remains the same. This indicates that the Chinese government does help regulated firms to alleviate their extra burdens through compensation. Therefore, hypothesis H1 is accepted.

There are three main explanations for these results. First, the Chinese government has designed various incentive-based environmental policies, as well as other preferential policies. Therefore, one or more energy-saving projects of an ESQP-regulated firm can receive multiple stacked subsidies and preferential policies from the Development and Reform Commission, the Science and Technology Administrative Department, the Environmental Protection Administrative Department, the Financial Department, and the Department of Commerce, among others. Second, China has established institutions aimed at enhancing the regional enforcement of environmental regulations. These institutions also encourage local governments to compensate environmentally friendly or energy-saving firms or projects, for example, whether they had reached the energy-saving quota was included in the performance evaluation of provincial leaders. This might spur local governments to compensate the ESQP-regulated firms to encourage them to enhance their energy-saving activities [46]. The Chinese government also incorporated a mandatory environmental index (including energy consumed per 10,000 CNY of GDP) into the performance evaluation index system of governors, mayors, and county magistrates in 2007. An operation system, including statistics, monitoring, and the assessment of energy consumption, was established afterwards, which affects the promotion of local government officials. Hence, local government officials have a strong internal motivation to support ESQP-regulated firms to achieve energy-saving tasks. They gain political capital with better performance evaluations, which helps them obtain future promotions [68]. (In 2016, China set up the Central Environmental Protection Supervision Committee to reduce local protectionism in environmental regulation. The Committee supervises the environmental protection activities of local governments and their relevant departments. Meanwhile, China carried out a pilot reform in environmental monitoring institutions under the municipal/county environmental protection departments, which led them to be vertically managed by the provincial departments). Third, the spending subsidies for enterprises are supervised by strict auditing and accountability systems in China. To some extent, such strict systems ensure that subsidies are granted to ESQP-regulated firms according to their energy conservation responsibilities.

Why do subsidies and tax rates differ significantly between the ESQP-regulated and unregulated firms while their financial expense ratios show no statistically significant difference? There could be several reasons for this. First, given the subsidy budget and tax preferential policies, local governments have the decision-making power and flexibility to grant subsidies and determine which firms enjoy preferential tax treatments. This is due to the lack of detailed and specific implementation guidelines for public subsidies and preferential tax policies [69]. Provincial and lower-level governments may set tendentious or implicit clauses to distribute more subsidies and preferential tax policies to ESQP-regulated firms to encourage them to finish energy-saving quotas. This can help provincial governments to accomplish the energy-saving tasks mandated by the Central Government. Second, the demand-side competition of bank-based capital markets in China is fierce. Commercial banks have sufficient autonomy and negotiation power to choose projects that fit their lending strategy and have the lowest default risk. This is confirmed by the fact that the coefficient for lnasset is significantly negative (−1.267,

Table 5. Estimated results of compensation for inequities II–IV among regulated firms.

Inequity	Independent Variables	Subsidy	Financial Expense Ratio	Tax Rate	Subsidy Intensity
Inequity Ⅱ	ESB×ESQP	0.593 ***	−3.395 ***	0.025	0.254 ***
		(−0.086)	(−1.124)	(−0.182)	(−0.044)
Inequity Ⅲ	complete×ESQP	0.269 ***	−1.606 ***	−0.424 ***	0.0530 **
		(−0.043)	(−0.564)	(−0.091)	(−0.022)
Inequity Ⅳ	state-own×ESQP	0.433 ***	−0.850	0.561 *	0.080
		(−0.146)	(−1.906)	(−0.308)	(−0.074)
export×ESQP (exporting or not)		0.056 ***	−0.022	−0.025 **	0.016 ***
		(−0.005)	(−0.072)	(−0.012)	(−0.003)

Noting: Standard errors in parentheses; *, **, *** indicate significance at the 10%, 5%, and 1%. There are estimated results of 16 models, but only the estimated coefficients of key explanatory variables in each model are listed.

). This also implies that the green credit policy is not successfully implemented [70], or is at least not well coordinated with the ESQP. The wide-ranging impact on energy-intensive industries, unclear implementation standards, and lack of environmental information are the main problems in the implementation of the green credit policy in China [71].

4.3. Compensation for Inequities II–IV among Regulated Firms

The results regarding whether the inequities within the ESQP-regulated firms are compensated are shown in Table 5.

First, for inequities II–IV, the estimated coefficients for ESB×ESQP, complete×ESQP, and state-own×ESQP on subsidy are all significantly positive, at 0.593, 0.269, and 0.433, respectively. Hence, among the ESQP-regulated firms, those that have high energy-saving burdens, complete energy-saving quotas, or are state-owned receive larger subsidies than their counterparts. This indicates that the ESQP’s complementary financial incentives (e.g., funds to reward advanced or up-to-standard firms in terms of energy conservation) or penalties are effectively implemented. Second, for inequities II–III, the treatment effects of ESB×ESQP and complete×ESQP on the financial expense ratio are negative (−3.395 and −1.606, respectively) and significant at the 99% confidence level. However, we did not find a significant effect of state-own×ESQP on the financial expense ratio. As described in the previous section, local governments and commercial banks can only help a small subset of regulated firms by granting easy access to loans at a lower cost. This may also be attributed to the green credit policy, where local governments and commercial banks are required to jointly loan a certain amount of green loans to qualifying projects and companies. Third, the coefficient for complete×ESQP on the tax rate is negative (−0.424) for Inequity III, while the coefficient is positive (0.561) for inequity IV. State-owned firms are generally high taxpayers in a region and rarely engage in tax avoidance, which is more common among private firms. Many managers in state-owned firms consider their firms as “a springboard for becoming bureaucrats” [72]. They even seek to pay more taxes, which scores them more points for their future promotion. Consequently, the status of state-owned firms does not necessarily help them reduce tax rates. In summary, hypothesis H2 is almost accepted.

Approximately 88.03% of high-burden firms come from five industries, identified as energy-intensive industries. These include the petroleum, refining, and coking industry, the raw chemical materials and chemical products industry, the non-metal mineral products industry, the smelting and pressing of the ferrous metals industry, and the smelting and pressing of non-ferrous metals industry [73]. This indicates that the Chinese government has not exempted energy-intensive industries, although our findings indicate that they receive more compensation.

4.4. Effects of Other Characteristics of Firms on Compensation

First, we designed an interactive item (export×ESQP) between the dummy export (1 for exporting firms and 0 for others) and the dummy ESQP. This allowed us to examine whether exporting firms received more compensation than ESQP-regulated firms without exports. According to Table 5, exporting firms received more subsidies and tax benefits than non-exporting ones. This finding echoes those of Dai and Cheng [58] and Duch et al. [74], suggesting that existing policies in China help to enhance or maintain the exporting capacity of ESQP-regulated firms.

Second, we use subsidy intensity (the ratio of subsidy income to revenue) to substitute subsidy income and normalize the impact of firm size [75]. Our results are consistent with the above findings. ESQP-regulated firms receive larger subsidy amounts than unregulated firms (not presented in the paper). Regulated firms with a high energy-saving burden or those that accomplish their energy-saving quotas, as well as exporting firms, appear to receive more subsidies (Table 5).

Third, manufacturing sub-sectors face different market conditions for their products, face different environmental regulations, and have different cost pass-through abilities [15,76]. Thus, the compensation provided to ESQP-regulated firms might vary across sub-sectors [77]. For this reason, we estimated Equation (1) at the two-digit sector level to clarify the industrial heterogeneity of the average treatment effect. We focused on the sectors with more than 100 ESQP-regulated firms. Their two-digit sector code includes 13~15, 17, 22, 25~28, 31~33, 35, 37, and 40 (

Table 6. Estimated results by sector.

Two-Digit Sector	Subsidy	Financial Expense Ratio	Tax Rate
13-Food Processing	1.062(0.207) ***	−2.539(1.889)	−0.901(0.907)
14-Food Production	0.562(0.337) *	−0.592(0.865)	−0.217(0.838)
15-Beverage Industry	0.573(0.269) **	−2.237(1.998)	−0.66(0.682)
16-Tobacco Industry	-	-	-
17-Textile Industry	0.131(0.129)	0.609(0.988)	−0.175(0.222)
18-Garments and Other Fibre Products	-	-	-
19-Leather, Furs, Down and Related Products	-	-	-
20-Timber Processing	-	-	-
21-Furniture Manufacturing	-	-	-
22-Papermaking and Paper Products	−0.002(0.173)	0.677(0.599)	−0.511(0.361)
23-Printing and Record Medium Reproduction	-	-	-
24-Cultural, Educational and Sports Goods	-	-	-
25-Petroleum Refining and Coking	0.436(0.192) **	−0.94(1.012)	−0.549(0.419)
26-Raw Chemical Materials and Chemical Products	0.946(0.114) ***	−0.018(1.020)	−0.252(0.206)
27-Medical and Pharmaceutical Products	0.601(0.343) *	1.360(0.768) *	−1.045(0.662)
28-Chemical Fibre	0.769(0.347) **	−0.024(0.223)	−1.065(0.286) ***
29-Rubber Products	-	-	-
30-Plastic Products	-	-	-
31-Nonmetal Mineral Products	0.948(0.105) ***	−0.776(1.530)	−0.206(0.256)
32-Smelting and Pressing of Ferrous Metals	0.651(0.132) ***	−0.517(0.967)	−0.184(0.214)
33-Smelting and Pressing of Nonferrous Metals	1.009(0.446) **	−0.659(0.964)	0.311(0.426)
34-Metal Products	-	-	-
35-Ordinary Machinery	0.622(0.416)	−5.823(6.315)	−1.283(0.908)
36-Special Purposes Equipment	-	-	-
37-Transport Equipment	1.507(0.350) ***	0.275(0.441)	0.296(0.716)
39-Other Electronic Equipment	-	-	-
40-Electric Equipment and Machinery	0.777(0.413) *	−0.257(0.535)	0.63(0.949)
41-Electronic and Telecommunications	-	-	-
42-Instruments and meters	-	-	-
43-Waste Resources and Waste Materials Recycling and Processing Industry	-	-	-

Noting: Standard errors in parentheses; *, **, *** indicate significance at the 10%, 5%, and 1%. The numbers of treated firms and control firms after matching in each industry are the same, due to each treated firms being matched to a control firm from the same industry. The industries without results are those with less than 100 regulated firms.

). Except for the firms in the textile industry (sector code 17), paper-making and paper products (22), and ordinary machinery (35), ESQP-regulated firms from the sub-sectors above receive more subsidies than the unregulated firms in the same sectors. However, the tax benefits received by the regulated and unregulated firms in those sub-sectors were significantly different, except for the medical and pharmaceutical products sub-sector (27). Hence, subsidies are a very common compensatory measure used by the Chinese government to mitigate losses from uneven energy saving and emission reduction burdens.

5. Conclusions and Inspiration

5.1. Conclusions

Using the Energy-Saving Quota Policy in China as a case study, we investigated whether the ESQP-regulated firms are compensated by more subsidies and/or lower financial expense ratios and tax rates. We also investigated whether inequities within the ESQP-regulated firms are rectified. We used the balanced panel data of 6189 ESQP-regulated and 6189 unregulated firms from 2010 to 2013 and a matching technique combined with a difference-in-differences estimator. Our results show that the ESQP-regulated firms receive larger subsidies and benefit from lower tax rates than the unregulated firms in the same industry. Within the ESQP-regulated firms, those with higher energy-saving burdens are subsidized more and achieve lower financial expense ratios. Firms that complete their energy-saving quotas benefit from more subsidies, lower financial expense ratios, and lower tax rates. State-owned firms receive more subsidies, and exporting firms receive more subsidies and lower tax rates. hypothesis H1 and H2 are almost accepted. We found that the most popular compensation type is subsidies, followed by tax benefits and loan support. Finally, our results reveal that the Chinese government offers energy-intensive industries more compensation, rather than exempting them from the ESQP, as 88.03% of high-burden firms are from the five most energy-intensive industries. In conclusion, firms regulated by the Chinese energy-saving quota policy received appropriate compensation, which provides an impetus for firms to reduce energy consumption and CO₂ emission in a sustainable manner.

5.2. Practical Inspiration

This research conclusion provides practical guidance for firms to reduce energy consumption and CO₂ emission in a sustainable manner and is of great practical significance to policy makers improving compensation policies beneficial to the mitigation of global warming and the sustainable development of the eco-environment.

From a firm manager’s perspective, our findings suggest that the Chinese government provides compensation for firms suffering from inequities induced by energy regulation. In this case, the regulated firms should increase investments in energy-saving activities to improve energy end-use efficiency and reduce CO₂ emissions. This includes the introduction of energy-saving technologies or ultra-low emissions techniques, the installment of energy-saving and emission-reduction equipment, and the elimination of outdated production capacity. Firm managers should actively apply for environmental, energy-related, and other general subsidies, tax incentives, and loan support programs to offset their energy-saving costs. Most ESQP-regulated enterprises are pillar enterprises (e.g., large taxpayers and large employers), which makes them more likely to be monitored. Thus, information regarding supporting policies is provided to them in a timely and comprehensive manner by government officials. However, small and medium-sized enterprises may face information asymmetry because they are often overlooked by local officials due to their smaller economic contribution. Hence, small and medium-sized enterprises need to make more efforts to gather information about, and apply for, supporting policies.

The results of our study indicate compensatory policies accompanying unilateral climate policies can really benefit regulated firms and those with higher burdens, which guides policymakers to design or optimize their compensatory policies and increase the intensity of compensation for firms to alleviate their concerns about the loss of competitiveness. Thus, sustainable development of CO₂ emission reduction efforts can be realized, ultimately expanding the economic development space of a country or a region. Of course, supervision (such as the central environmental supervision for local governments and the auditing system in China) of the use of compensatory policies should be strengthened to ensure that they are used to promote energy-saving activities and other environment-friendly activities.

Future studies could evaluate the moderating effects of compensation on regulated firms’ economic performance, such as revenue, profit, exports, and employment. Due to the lack of data on the energy consumption amount saved by the regulated firms, we were unable to determine whether the compensation received by these firms could cover the cost of their energy-saving activities. This could be another topic for future research.