How GDP Manipulation by Local Government Affects Corporate Greenwashing in China

Abstract

Firms frequently face a tradeoff between the advantages of upholding sustainability and ESG performance and the expenses associated with participating in ESG initiatives. This tension leads to an increase in greenwashing practices, which ultimately undermines genuine sustainability efforts and misleads stakeholders. Motivated by this trend, our study examines the influence of a macro-level factor, specifically local city-level governments’ GDP manipulation, on the extent of firms’ greenwashing, highlighting how government behavior can distort sustainable business practices. Using the data of the publicly traded Chinese manufacturing companies during the period of 2007–2019, we find a positive and significant relationship between the extent to which firms engage in greenwashing and the extent of local city-level governments’ GDP manipulation. Additional analysis reveals that firms’ financial constraints and external monitoring are the channels through which governments influence firms’ greenwashing. In addition, the finding of a positive association between firm greenwashing and government GDP manipulation is more pronounced in regions with a less developed marketization index, in periods before China’s anti-corruption campaign, in state-owned firms, and in firms at the business life cycle of the mature stage. Our study addresses a gap in the literature by demonstrating how government economic interventions influence firms’ sustainability performance.

1. Introduction

In recent decades, deteriorating environmental issues and the rise in corporate scandals have led to heightened public scrutiny of firms’ environmental, social, and governance (ESG) performance as key drivers of sustainability [1]. Investors’ assessments of firms have expanded beyond financial performance alone and extended to firms’ ESG performance, which plays a crucial role in achieving long-term environmental and social sustainability. Driven by this pattern, research on ESG performance has been flourishing. Previous research has established that engaging in ESG practices can foster trust between companies and investors [2], resulting in reduced capital costs [3,4] and ultimately improving the financial performance of companies [5]. Nevertheless, the advantages of engaging in ESG activities mainly include long-term benefits only. Companies continue to lack motivation to participate in ESG initiatives as a result of the absence of immediate financial benefits [6]. For example, a company can improve its brand reputation by consistently achieving higher environmental standards over time. However, the financial commitment to environmental protection is substantial, and it may not yield immediate profits for the company. Consequently, the investment would lead to a decrease in short-term performance. Because managers’ compensation is typically tied to the company’s short-term profitability, managers may be reluctant to initiate environmental investments, even if these investments contribute to long-term sustainability.

As a result, greenwashing practices started to emerge, posing a significant challenge to corporate sustainability. Greenwashing is defined in the literature as having two distinct features: (1) companies conceal their poor ESG performance and (2) companies overstate their ESG performance to deceive the public into believing that they have achieved a higher level of performance than they actually have [7,8,9]. Firms engage in greenwashing with the intention of benefiting from a positive reputation resulting from excellent ESG performance while also seeking immediate financial gains by avoiding significant expenses related to ESG activities. This deceptive practice not only misguides investors but also weakens broader sustainability efforts by diverting resources away from genuine ESG progress. A notable example of greenwashing is Volkswagen’s admission of tampering with emissions tests through the installation of a “defeat” device and software in multiple vehicles, which could alter their performance to reduce emissions specifically during testing.

Previous research has extensively documented numerous consequences of greenwashing [9,10,11,12,13]. Nevertheless, the factors that impact the extent to which firms engage in or practice greenwashing are quite restricted. Specifically, the factors that have been identified primarily pertain to firm-level factors, while broader macro-level factors have received limited attention. The objective of our study is to address this gap in the existing literature by investigating the association between a company’s degree of greenwashing and the level of gross domestic product (GDP) manipulation by the local city-level government in the region where the company is registered, specifically focusing on Chinese manufacturing firms. By doing so, we contribute to the sustainability literature by shedding light on how political and economic distortions can influence corporate ESG behavior and hinder meaningful sustainability initiatives.

We selected Chinese manufacturing firms for our study for three specific reasons. First, while developed nations have historically been the primary contributors to global warming and other environmental problems, the escalation of fresh pollution is currently being propelled by newly industrialized nations [14]. In 2018, China accounted for 26% of the total global greenhouse gas emissions, as reported by the European Union’s Emissions Database for Global Atmospheric Research. China’s significance as a research topic for addressing global environmental issues cannot be overstated. Second, manufacturing activities are the primary cause of pollution in China [15]. The 2015 China Statistical Yearbook on the Environment states that industrial emissions of SO2, oxynitride, and smoke (dust) make up 88.15%, 67.60%, and 83.65%, respectively, of the total emissions in society [15]. Thus, studying manufacturing firms is essential for addressing China’s environmental problems. Third, China is among the few countries that publicly admits GDP manipulation by local governments. Three provinces (Liaoning, Inner Mongolia, and Tianjin) have admitted publicly to manipulating fiscal and industrial data in recent years. Chen Qiufa, the governor of Liaoning Province, admitted to fabricating fiscal data from 2011 to 2014 in January 2017, and the proportion of inflated data increased annually from 2011 to 2014, with a 23% increase in 2014 (http://politics.people.com.cn/n1/2017/0120/c1001-29037119.html (in Chinese), accessed on 3 May 2024). Similarly, the local governments of the Inner Mongolia Autonomous Region and Tianjin acknowledged that the 2016 financial and economic data were likewise overstated. Inner Mongolia reduced its 2016 industrial output estimates by 40%, and the Tianjin Binhai New Area revised its 2016 GDP by nearly one-third in response to GDP manipulation (https://m.21jingji.com/article/20180120/herald/e940a46e802b7493bcef7e253c868586.html (in Chinese), accessed on 23 April 2020). In summary, China’s significant role in addressing ESG issues, along with its unique attributes, makes it a valuable case study for examining the correlation between corporate greenwashing and local governments’ manipulation of GDP. The intersection of corporate greenwashing and government economic misrepresentation presents an important yet understudied dimension of sustainability research.

We anticipate a positive relationship between the level of a company’s greenwashing and the level of manipulation of the local government’s GDP. Government officials manipulate GDP due to their concerns about the impact of GDP on their political careers [15,16]. The performance of China’s government officials is closely scrutinized based on the economic growth of their respective regions, and a strong growth rate is essential for the political advancement of local officials [17,18,19]. The existing literature documents empirical evidence about firms collaborating with governments in fulfilling their requirements for reporting GDP. During the course of assisting governments with their GDP objectives, firms have begun to recognize that governments give higher priority to GDP than to other factors, such as ESG. Consequently, governments’ oversight and implementation of local firms’ ESG practices will be weak. As previously mentioned, firms do not receive immediate benefits from engaging in ESG activities, and they are less likely to participate in such activities when they perceive weak monitoring. Nevertheless, given the enduring advantages of ESG practices, companies continue to strive for an environmentally friendly and socially responsible image. Consequently, companies may resort to increased greenwashing practices. Firms that are not mandated by governments to contribute to GDP figures may still participate in fewer genuine ESG activities and instead engage in more greenwashing. The reason for this is that according to institutional theory, companies tend to adopt the actions and practices of other companies. When firms witness others engaging in this behavior of greenwashing, they will also adopt it. Furthermore, the actions of governments have a significant impact on shaping the prevailing social norms in this region. Moreover, firms tend to emulate the dishonest behaviors exhibited by governments, as suggested by institutional theory. In conclusion, our forecast suggests that in a region where local governments exert a greater degree of manipulation over GDP figures, companies are likely to engage in more greenwashing.

Our study provides three contributions to the existing body of literature. This study contributes to the sustainability literature by examining the influence of local government behavior, specifically GDP manipulation, on corporate greenwashing practices in China [20,21,22,23,24]. By analyzing the relationship between governmental economic distortions and corporate environmental reporting, this research highlights the complex interplay between political incentives and corporate sustainability efforts. The findings suggest that local governments’ manipulation of economic data can undermine corporate commitments to genuine sustainability practices, leading to increased greenwashing. This aligns with research indicating that lack of government supervision and imperfect certification mechanisms contribute to corporate greenwashing behaviors [25].

To the best of our knowledge, our study is the first to investigate the impact of government behaviors on firm-level greenwashing. Despite the increasing prevalence of greenwashing, there is a lack of research on the role of the government in this issue. In theory, governments play a crucial role in influencing firms’ greenwashing practices, as they establish and enforce regulations regarding firms’ ESG activities. Our study validates the significant impact of government involvement on firms’ greenwashing practices. Specifically, we found that government manipulation of GDP is a determining factor that influences the extent to which firms engage in greenwashing.

Furthermore, we suggest alternative metrics for assessing greenwashing. To serve as a proxy for a firm’s greenwashing practices, prior studies often compare two ESG ratings, namely, disclosure and performance ratings. Nevertheless, the vulnerability of this measure lies in its dependence on the ratings provided by third parties, which can potentially be manipulated through a company’s public relations or lobbying efforts. Our measure utilizes a firm’s ESG investment as a proxy for its actual performance, providing a more accurate assessment of a firm’s greenwashing practices.

Moreover, our study is expected to captivate the attention of both shareholders and policymakers. As green investment grows, investors are interested in differentiating companies that genuinely demonstrate good ESG performance from those that engage in greenwashing. Our research findings can provide valuable insights into this matter. Similarly, society is keen to address and has extensively researched both issues: government manipulation of GDP and firm greenwashing. Nevertheless, their relationship is disregarded. The results of our study suggest that there is a positive correlation between the two variables, implying that implementing stricter government regulations on GDP manipulation could potentially decrease the occurrence of firm greenwashing. Policymakers should find this discovery intriguing.

2. Literature Review and Hypothesis Development

2.1. Greenwashing

The increased focus on companies’ ESG performance is driving the growth of ESG investment worldwide. An example of a note is the introduction of pan-ESG funds in the Chinese market. By December 2021, the total value of these funds had reached CNY 220 billion [1]. China’s rapid growth not only solidifies its position as a major player in the global financial arena but also signifies a significant shift toward sustainable investing within the country.

Due to the increasing popularity of ESG investing, an increasing number of companies are presently disclosing ESG-related information. This phenomenon is not limited to any particular geographic region but is observed worldwide. According to data gathered by the Governance and Accountability Institute, 86% of the companies listed in the S&P 500 Index have made a deliberate decision to disclose their ESG reports [1]. By 2021, more than 1400 listed companies in China had committed to releasing their ESG reports across various markets [1]. This indicates a substantial increase of approximately 40% in comparison to the previous year, demonstrating a widespread acknowledgment of the importance of ESG factors in corporate decision making.

However, despite the current positive progress, there are still persistent concerns about the occurrence of greenwashing—the dishonest act of presenting an environmentally and socially responsible image while engaging in unsustainable or harmful practices. The main issue arises from the lack of clarity regarding the precise definition of a genuine “green” initiative. This difficulty is intensified by the absence of consistent ESG disclosure criteria in various jurisdictions, thereby giving rise to opportunities for dishonest entities to take advantage of these criteria. Moreover, the divergence in ESG ratings across various rating agencies and methodologies adds additional complexities [20,26]. The volatility in the sustainability performance of companies presents a dilemma for investors in accurately assessing it. Furthermore, it provides companies with opportunities to manipulate their ESG ratings in order to present a more favorable image than they actually deserve.

The occurrence of greenwashing has ramifications that extend beyond financial concerns. Previous research highlights the adverse impacts of greenwashing on consumer behavior [27], employee performance [2], and risk perception [28], which all negatively affect creditability and company reputation [24,25]. Consequently, greenwashing diminishes a company’s capacity to compete in the market [29].

However, when viewed through an economic lens, there is mounting evidence suggesting that authentic sustainability initiatives can yield concrete benefits for businesses [27,30]. While greenwashing may offer immediate advantages such as cost reduction and increased profit margins, the potential long-term repercussions on brand reputation and market credibility are far more substantial than any short-term benefits.

2.2. GDP Manipulation

Recently, the Chinese central government has actively implemented administrative decentralization reforms, transferring responsibilities from the federal level to local governments. This approach has led to a governance structure that is marked by the concentration of political power and the decentralization of budgetary power. The evaluation of local government performance by the central authority plays a crucial role in determining the career progression of local officials [28,31].

The existing system of political promotion for local leaders, which is mainly linked to GDP growth, offers significant motivation for officials to prioritize local economic development and attain high growth rates [30,32,33,34]. Nevertheless, this system of rewards also encourages exploitative conduct among local administrations, such as the manipulation of GDP data [35,36,37,38].

Local governments possess considerable resources, including administrative approvals, financial subsidies, land acquisition, and tax incentives [18], which give them a significant informational edge regarding the actual conditions within their jurisdictions. The significant disparity in information between central and local governments creates a favorable environment for local officials to engage in opportunistic behavior.

Prior research emphasizes the widespread occurrence of opportunistic behavior motivated by promotion incentives [39]. The manipulation of GDP data by local governments exemplifies opportunistic behavior. Officials in a system where the success of a local government is linked to its GDP performance have been observed to engage in statistical manipulation to increase their likelihood of promotion [35,40].

2.3. Hypothesis Development

The existing literature outlines three strategies that governments can employ to boost their GDP: increasing government investments, leaving zombie firms, and exerting pressure on local businesses to manipulate financial records [18]. Although the first two options typically necessitate a financial investment, the final option can be achieved without any actual cash outflows. The accounting literature has presented empirical evidence to substantiate this technique. In their study, Fan and Song [19] discovered that Chinese companies under the control of the central government engage in earnings manipulation to decrease fluctuations in the central government’s GDP data. To achieve this objective, these companies resort to real earnings management. Another study by Cai et al. [18] revealed that companies under the control of local governments engage in earnings manipulation to artificially increase the GDP statistics of those local governments. These studies suggest that governments may directly or indirectly request that companies help achieve GDP objectives.

The motivation for firms to adhere to government requests lies in the advantages of political connections. Multiple studies have demonstrated that companies can derive numerous advantages from their affiliations with governmental entities, including but not limited to obtaining loans [41], enjoying reduced financing costs [42], facing fewer penalties [43], and achieving higher rates of success in securing government contracts [44]. Consequently, companies may be inclined to assist the government due to the allure of these advantages. Meanwhile, companies are beginning to comprehend that governments place greater importance on GDP than on other factors, such as ESG. When governments are confronted with a conflict between GDP and ESG considerations, it is probable that their implementation and enforcement of ESG policies will lack the strength to prioritize GDP growth. Moreover, companies with political connections have influence over the government such that even if they are discovered to be noncompliant with ESG regulations, their penalties will be mitigated as a result of their political affiliations. Thus, they may not necessarily have to participate in ESG activities that entail substantial financial expenses. In the meantime, companies may still be interested in appearing environmentally and socially conscious, driven by the advantages of an ESG image. Consequently, they will increase their greenwashing practices, and they are likely to reap the advantages of both political connections and a positive reputation for ESG.

Firms that are not required or implied by governments to assist with GDP figures may still exhibit similar behavior; specifically, they may engage in more greenwashing. Institutional theory suggests that firms consistently strive to emulate the actions of their peers or competitors. When companies observe certain peers engaging in greenwashing, which refers to making false claims about investing in ESG initiatives, they may be inclined to mimic this behavior. In doing so, they will sustain their competitive edge among their competitors. Otherwise, they might be compelled to raise their expenses for ESG initiatives, thereby putting themselves at a disadvantage in the competition.

Moreover, the actions of governments often play a role in shaping social norms. When governments establish a culture of deceit, companies are inclined to adopt the same behavior, as indicated by institutional theory. This will result in an increase in the practice of greenwashing by companies.

Our hypothesis is formally stated as follows:

 Hypothesis 1: 

There is a positive correlation between greenwashing by firms and GDP manipulation by local governments.

3. Research Design

3.1. Sample

Our sample construction commenced by encompassing all publicly traded firms in China listed on the Shanghai or Shenzhen stock exchanges between 2007 and 2019. Next, we narrowed our sample to include only manufacturing firms, as they are the primary contributors to pollution in numerous countries, including China. The year 2007 was chosen as the initial reference point due to its alignment with the adoption of China’s new accounting standards, which bear a striking resemblance to the International Financial Reporting Standards (IFRS). This guarantees that all financial data during our selected time frame were comparable. The sample ends in 2019 because of the revision of the GDP calculation methodology in 2020. There are two globally used models for computing GDP: centralized and decentralized. The centralized approach allows the central government or its affiliated bureaus to ascertain the GDPs at both the local and national levels. The United States, Canada, and Germany are among the countries that utilize the centralized approach. On the other hand, the decentralized model requires that local governments or their associated bureaus compute the GDP at the local level, while the central government or its associated bureaus compute the national GDP. Japan, Russia, and Vietnam are countries that utilize the centralized approach. China implemented a decentralized model from 1985 to 2019. Starting in 2020, the National Bureau of Statistics computed provincial figures, which is why our sample period ends in 2019. Subsequently, we eliminated any businesses whose names commenced with “ST” (special treatment) from our sample, in accordance with prior research. The addition of “ST” to the names of these firms signifies that they are at a high risk of being delisted from the stock exchange. ST firms may endeavor to improve their publicly disclosed financial information as a means to avoid being removed from listing. As a result, their behaviors may differ from those of other businesses.

The calculation of GDP manipulation relies on the utilization of nighttime lighting data obtained from the publicly accessible dataset of the National Oceanic and Atmospheric Administration (NOAA). The China Stock Market & Accounting Research (CSMAR) database is the primary source of financial data at the firm level. Data regarding businesses’ ESG ratings were obtained from hexuan.com. The statistics for cities were obtained from the China City Statistical Yearbook, with additional data from local yearbooks to fill in any missing information. After removing all variables with missing data, our analysis included a total of 2401 firm-year observations, which correspond to 1976 unique firms.

Initial analysis of the samples revealed the existence of outliers. To reduce the impact of extreme values, all continuous variables underwent winsorization. This process involved replacing values that were below the 1st percentile and above the 99th percentile with the values at those specific percentiles.

3.2. Variables

In our study, the primary variables of interest were the dependent variable of greenwashing, and the independent variable represented the degree to which local city-level governments manipulate GDP. Following the prior literature [45], we define greenwashing as the discrepancy between a company’s disclosed ESG performance and its genuine ESG performance:

Greenwash (GW) = Genuine ESG performance ranking − Disclosed ESG performance ranking

where the Genuine ESG ranking reflects a firm’s position based on its actual ESG performance across all observations, with rank 1 indicating the highest performance. This ranking is determined by a firm’s environmental investment as a percentage of total assets, based on the principle that strong ESG performance requires substantial investment—achieving high performance without meaningful investment is unlikely.

The disclosed ESG ranking represents a firm’s ranking based on its ESG disclosure ratings from the Hexun database, where rank 1 denotes the highest level of disclosure. This ranking captures the extent to which firms publicly report their ESG activities, independent of their actual ESG performance.

To measure disclosed ESG performance, we employed ESG disclosure ratings from the Hexun database as a proxy for a company’s publicly disclosed ESG performance. Hexun has independently released CSR evaluation products since September 2013. Its rating approach accounts for shareholder responsibility and industry differences, providing a more comprehensive assessment of corporate ESG performance.

We contend that a firm’s expenditure or investment in environmental conservation is a contributing factor to the firm’s ESG performance, and a higher investment is expected to result in superior performance. Additionally, this investment as an input to ESG performance is unlikely to be subject to greenwashing and can serve as a reliable indicator of actual performance. In order to acquire the data, we gathered information on expenses associated with desulfurization projects, sewage treatment, waste gas, dust removal, and energy savings from the annual reports and ESG reports of companies. The raw data were adjusted by dividing them by the total assets of each company to account for variations in firm size. Subsequently, we assessed and assigned a rank to the two ESG indicators acquired from each source, allowing us to ascertain the position of a company’s performance.

Afterward, we calculated the difference between the two rankings. Greater disparities between the firms’ stated ESG performance and their actual performance result in a greater level of greenwashing.

Based on recent research [36,39,40,41,42], which highlights the predictive power of nighttime lights in assessing economic activities, we utilized nighttime lights to estimate the actual economic progress of a specific area. Unlike the traditional GDP measure used to assess a country or region’s economic development, nighttime light data are unaffected by interregional price factors. Furthermore, light data encompass not only the tangible products and services of the market economy, as quantified by GDP, but also the worth of goods and services that are not exchanged in the market [46]. Hence, satellite nighttime light data can provide a more precise depiction of the actual economic progress of a nation or locality.

Chen et al. [47] used nighttime light data to estimate China’s real GDP and then compared it to the officially reported GDP to identify any manipulation of GDP. We contend that the rate of GDP growth is more significant than the absolute value of GDP. To improve upon the approach put forth by Chen et al. [47], we determine the actual growth rate of the GDP by computing the percentage change between the nighttime light of the present year and that of the preceding year. The reported GDP growth is calculated as the percentage change between the current year’s GDP, as reported by local governments, and the GDP data from the corresponding previous year. The disparity between the two figures indicates potential manipulation of GDP by governments.

To accurately measure real GDP growth, it is necessary to eliminate economic fluctuations from the calculation of the economic growth rate. HP filtering is commonly used in the academic literature for decomposing fluctuations in regional economies. However, this method is subject to controversy due to its numerous shortcomings. For instance, the application of the HP filtering technique can result in significant disparities between the filtered values at the conclusion and the midpoint of the sample. Additionally, the method is highly responsive to variations in the smoothing parameters. To address the limitations of the HP filtering method, Hamilton [48] introduced an alternative filtering approach known as Hamilton filtering. The Hamilton filtering method preserves the benefits of conventional HP filtering while addressing the aforementioned limitations of HP filtering. This paper utilizes the Hamilton filter to separate random variations from the developing trend in the light growth rate and in the reported GDP growth rate. To more accurately measure the sustainable growth of the economy, we extended the test period for the reported GDP growth rate and lighting growth rate to cover the years 1993–2019. Using this expanded sample, we calculated an economic volatility indicator and a reported GDP trend indicator using the Hamilton filter. Additionally, we obtained a light volatility indicator and a light growth trend indicator. The indicator of local government GDP manipulation, which signals the degree of opportunism, is determined by taking the absolute value of the difference between the light growth trend indicator and the reported GDP trend indicator from 2003–2017.

3.3. Model

To test our hypothesis, we utilize the following equation. We account for firm and year fixed effects and address the potential intercorrelations among firms located in the same city by clustering the standard errors by city.

$${\mathrm{G}\mathrm{W}}_{\mathrm{i},\mathrm{j},\mathrm{t}}={\mathsf{\beta }}_0+{\mathsf{\beta }}_1\mathrm{\ast }{\mathrm{G}\mathrm{D}\mathrm{P}\_\mathrm{M}\mathrm{a}\mathrm{n}\mathrm{i}\mathrm{p}}_{\mathrm{j},\mathrm{t}-1}+\mathsf{\beta }\mathrm{\ast }\mathrm{X}+\mathrm{f}\mathrm{i}\mathrm{r}{\mathrm{m}}_{\mathrm{i}}+\mathrm{y}\mathrm{e}\mathrm{a}{\mathrm{r}}_{\mathrm{t}}+{\mathsf{\epsilon }}_{\mathrm{i},\mathrm{j},\mathrm{t}}$$

(1)

GW stands for the degree of greenwashing, as defined in Section 3.2. GDP_Manip represents the extent to which the city-level government manipulates GDP in the city where firm $\mathrm{i}$ is registered. $\mathrm{X}$ indicates all control variables, including firm size, growth, leverage, ownership concentration, return on assets, cash, operating cash flow, firm age, and CEO duality, where i denotes firm i, while j and t denote city j and year $\mathrm{t}$, respectively. Definitions for all variables can be found in Appendix A. To avoid reserve casualty issues, we lagged all independent variables, including GPD_Manip, by one year.

4. Empirical Results

4.1. Summary Statistics

Table 1. Summary statistics.

Variable	Obs	Mean	Std. Dev.	Min	Max
City-level Variable
GDP_Manip	2407	0.079	0.038	0.004	0.202
Firm-level Variable
GW	2407	−128.403	262.568	−901.000	589.000
SIZE	2407	23.092	1.217	19.296	25.390
Growth	2407	0.162	0.386	−0.600	3.315
LEV	2407	0.515	0.187	0.051	1.007
OWN1	2407	37.669	15.166	8.810	74.650
OWN2_10	2407	20.935	12.802	2.060	57.290
ROA	2407	0.039	0.057	−0.292	0.211
CASH	2407	0.142	0.098	0.011	0.685
CFO	2407	0.058	0.069	−0.180	0.254
AGE	2407	2.873	0.311	0.693	3.611
INDEP	2407	0.372	0.055	0.231	0.667
DUAL	2407	0.843	0.364	0.000	1.000

presents the summary statistics for the key variables used in our baseline regression analyses. The sample size shrinks to approximately 2400 in the regression analysis due to missing data on the dependent variable. While this missing rate is comparable to that in other studies [45,49], we acknowledge this limitation and advise readers to interpret our findings with caution. The average value of GW is −128. The average value of GDP_manip is 0.079, indicating that local governments intentionally manipulate their GDP figures. The statistical characteristics of the control variables are similar to those found in previous studies.

Table 2. Pearson correlation.

	GW	GDP_Manip	SIZE	Growth	LEV	OWN1	OWN2_10
GDP_Manip	0.063 ***	1.000
SIZE	0.124 ***	0.036 *	1.000
Growth	0.000	0.038 *	0.023	1.000
LEV	0.065 ***	0.083 ***	0.450 ***	0.059 ***	1.000
OWN1	0.063 ***	0.090 ***	0.241 ***	−0.011	0.060 ***	1.000
OWN2_10	0.015	−0.067 ***	0.012	0.098 ***	−0.082 ***	−0.427 ***	1.000
ROA	0.066 ***	0.010	−0.018	0.177 ***	−0.415 ***	0.062 ***	0.081 ***
CASH	0.130 ***	0.032	−0.099 ***	0.005	−0.256 ***	−0.030	0.122 ***
CFO	0.018	−0.054 ***	0.017	0.046 **	−0.229 ***	0.106 ***	0.016
AGE	−0.024	−0.161 ***	0.212 ***	−0.048 **	0.074 ***	−0.143 ***	−0.001
INDEP	0.043 **	−0.009	0.046 **	−0.012	0.003	0.055 ***	−0.044 **
DUAL	0.077 ***	0.051 **	0.071 ***	−0.015	0.057 ***	0.091 ***	−0.075 ***
	ROA		CASH	CFO	AGE		INDEP
CASH	0.224 ***		1.000
CFO	0.432 ***		0.097 ***	1.000
AGE	−0.064 ***		−0.097 ***	0.020	1.000
INDEP	−0.021		0.000	0.010	0.014		1.000
DUAL	−0.035 *		−0.009	0.003	−0.009		−0.096 ***

Notes: * p < 0.1; ** p < 0.05; *** p < 0.01.

presents the Pearson correlation coefficients among these variables, based on the sample used in the baseline regression analyses. There is a positive and significant correlation between greenwashing and GDP manipulation, which provides some support for our first hypothesis. None of the correlations exceed 0.5, suggesting that multicollinearity is not a significant concern in our study.

4.2. Hypothesis Test

We test our hypothesis 1 by examining Equation (1) and present the results in

Table 3. Baseline regression.

	(1)	(2)
	OLS	2SLS
City-level Variable
GDP_Manip	256.030 *	707.614 *
	(144.053)	(428.626)
Firm-level Variable
SIZE	−8.992	−7.464
	(21.329)	(21.364)
Growth	1.638	0.912
	(12.690)	(12.636)
LEV	151.192 **	143.196 *
	(73.705)	(74.348)
OWN1	−0.596	−0.600
	(1.819)	(1.779)
OWN2_10	−0.256	−0.335
	(1.030)	(1.028)
ROA	79.302	81.390
	(151.313)	(149.573)
CASH	201.053 **	209.766 **
	(91.772)	(91.544)
CFO	97.972	85.789
	(85.431)	(86.153)
AGE	37.672	31.940
	(114.652)	(112.722)
INDEP	−8.837	−7.487
	(132.814)	(133.014)
DUAL	20.122	20.530
	(15.567)	(15.713)
Constant	−102.395	−154.502
	(575.471)	(569.381)
Firm-fixed effect	Included	Included
Year-fixed effect	Included	Included
Cluster	City
No. of Obs.	2407	2401
R-Squard	0.056	0.054

Notes: * p < 0.1; ** p < 0.05; standard errors in parentheses.

. In column 1, we include all control variables and control for both firm fixed effects and year fixed effects. The coefficient on GDP_Manip is 256.03 and is significant at the 10% level. This indicates that there is a positive relationship between firms’ greenwashing and local governments’ GDP manipulation. Increased manipulation of GDP by local governments results in a corresponding increase in greenwashing by local firms. When the manipulation of GDP increases by 1%, the degree of greenwashing increases by 25.6 (256.03 × 1%), which is economically significant because 25.6 is equivalent to more than 19% of the average level of greenwashing (25.6/129.211). In conclusion, the data support our first hypothesis.

4.3. Endogeneity Issue

The baseline results obtained from Table 3 may be affected by endogeneity problems. In other words, there are factors that simultaneously affect both firms’ greenwashing behavior (dependent variable) and governments’ manipulation of GDP (test variable). For example, when a culture promotes and supports the act of taking risks, it motivates both local businesses and local governments to engage in more opportunistic behaviors. Consequently, local governments could significantly manipulate their GDP figures, while local firms engage in a greater degree of greenwashing to enhance their ESG performance. In order to address the concerns arising from endogeneity issues, we employed a two-step least squares (2SLS) regression with an instrumental variable approach to re-evaluate hypothesis 1.

An optimal instrumental variable should exhibit a clear and direct relationship with the test variable while having no direct effect on the dependent variable except through the test variable. In order to achieve this objective, we employed the mean value of GDP manipulation by all other city-level governments in the same province and year as the instrumental variable in our testing of the government in question. Institutional theory posits that entities consistently strive to emulate the behaviors of others. When local governments observe other governments engaging in a significant level of GDP manipulation through their interactions, they will subsequently adopt a similar pattern, becoming more aggressive in their own GDP manipulation. Even if the government being studied is unaware of the manipulation of GDP by other neighboring governments, the high GDP figures reported by neighboring governments will exert pressure on it and lead to an increase in GDP manipulation by this government. Furthermore, there is no theoretical evidence suggesting that the manipulation of neighboring governments’ GDP by a government has a direct influence on the greenwashing behaviors of local firms.

To assess the validity of this instrumental variable, we initially performed a Kleibergen—Paap rk LM test. The resulting statistical value for the under-identification test was found to be 34.185, with a corresponding p-value of 0.000. The Cragg-Donald Wald F statistic result for the weak identification test is 222.248, which exceeds the critical threshold. These results indicate that there are no serious concerns of under-identification and weak instrumental variables when using neighboring cities’ GDP manipulation degrees as an instrumental variable. There should be no concerns for over-identification, because there is just one instrumental variable and one instrumented variable.

In the first stage of the regression analysis, we employ GDP manipulation by the government under study as the dependent variable and utilize the instrumental variable as the independent variable. By performing a regression analysis using Equation (1), we can calculate the predicted value of GDP manipulation. In the second phase, we substitute the existing GDP manipulation values with the forecasted values obtained from the first stage and then re-evaluate Equation (1).

The results are shown in Column 2 of Table 3. The regression analysis reveals a coefficient of 707.614 for the variable GDP_Manip, which is statistically significant at the 10% level. This indicates a positive correlation between a firm’s degree of greenwashing and its local government’s degree of GDP manipulation. This finding supports our first hypothesis, which posits that the manipulative actions of local governments’ GDP encourage companies to partake in a greater amount of greenwashing. Our first hypothesis remains supported even after addressing issues related to endogeneity.

4.4. Robustness Tests

Aside from the two-stage regression, we performed two further tests to assess the robustness of our findings. First, we normalized our dependent variable. The normalized dependent variable is obtained by subtracting the mean of its raw value from the raw value itself, and then dividing the result by the standard deviation of the raw value. The normalized value is utilized as the new dependent variable in Equation (1). Subsequently, the regression is re-estimated, and the results are presented in Column 1 of

Table 4. Robustness tests.

	(1)	(2)
	Normalized GW	Time_Trend
City-level Variable
GDP_Manip	2.698 *	711.153 *
	(1.634)	(428.539)
Firm-level Variable
SIZE	−0.028	−1.502
	(0.081)	(21.413)
Growth	0.003	−1.084
	(0.048)	(12.426)
LEV	0.546 *	92.595
	(0.283)	(70.262)
OWN1	−0.002	−0.722
	(0.007)	(1.663)
OWN2_10	−0.001	−0.713
	(0.004)	(1.020)
ROA	0.31	68.83
	(0.570)	(149.228)
CASH	0.800 **	197.383 **
	(0.349)	(89.563)
CFO	0.327	89.249
	(0.328)	(84.664)
AGE	0.122	−22.954
	(0.430)	(124.379)
INDEP	−0.029	−19.137
	(0.507)	(128.586)
DUAL	0.078	17.591
	(0.060)	(16.519)
SIZE_2007		0.000 *
		(0.000)
LEV_2007		−26.590 *
		(13.693)
ROA_2007		44.641
		(51.332)
Constant	−0.096	16,449.355
	(2.171)	(16,522.038)
Firm fixed effect	Included	Included
Year fixed effect	Included	Included
Cluster	City	City
No. of Obs.	2401	2401
R-Squared	0.054	0.062

Notes: * p < 0.1; ** p < 0.05; standard errors in parentheses.

. The coefficient on GDP_Manip remains positive and significant at the 10% level, providing support for hypothesis 1.

Second, we incorporate a time trend adjustment at the start of the sample period for the three variables, SIZE, LEV, and ROA, in the model. The test variable may have an impact on these three control variables as time progresses, and incorporating these three variables’ initial values of the sample period into the regression analysis assists in addressing this concern. The test findings are documented in Column 2 of Table 4. A strong and significant association remains between firms engaging in greenwashing and the manipulation of local governments’ GDP. Thus, Hypothesis 1 is still supported. In conclusion, the findings of the baseline regression analysis are robust.

4.5. Channel Test

We are interested in comprehending the correlation between corporate greenwashing and government manipulation of GDP, as well as the specific channels through which this manipulation affects firms’ greenwashing practices. As mentioned in the Section 2.3, there are at least two channels that influence firms’ greenwashing behaviors: financial constraints faced by firms and external monitoring. We conducted analyses on these channels and report the findings in

Table 5. Channel tests.

	(1)	(2)
	OLS	2SLS
City-level Variable
GDP_Manip	9852.524 *	1542.454 **
	(5583.700)	(684.829)
Firm-level Variable
SA	−69.789
	(161.387)
GDP_Manip × SA	2485.120 *
	(1470.070)
Analysts		3.513 *
		(1.817)
GDP_Manip × Analysts		−37.267 *
		(22.239)
SIZE	−14.293	−14.746
	(21.567)	(26.555)
Growth	6.077	5.967
	(12.465)	(13.009)
LEV	106.169	71.093
	(77.056)	(85.238)
OWN1	−0.34	0.76
	(1.754)	(1.592)
OWN2_10	−0.534	0.201
	(1.043)	(1.077)
ROA	60.424	−122.697
	(157.827)	(153.240)
CASH	171.370 *	264.441 ***
	(92.496)	(96.978)
CFO	93.135	124.64
	(86.418)	(90.811)
AGE	87.802	89.719
	(110.434)	(120.050)
INDEP	21.211	75.04
	(133.438)	(140.001)
DUAL	20.757	16.843
	(15.923)	(17.853)
Constant	−379.048	−243.374
	(885.242)	(639.143)
Firm fixed effect	Included	Included
Year fixed effect	Included	Included
Cluster	City	City
No. of Obs.	2401	2087
R-Squared	0.044	0.048

Notes: * p < 0.1; ** p < 0.05; *** p < 0.01; standard errors in parentheses.

.

In China, the majority of the largest banks are state-owned and government-controlled. The existing literature has confirmed that political connections with governments can help firms obtain loans with greater ease and at a reduced cost. When firms become aware of the government’s emphasis on GDP growth, they may decrease their investment in environmental initiatives in order to align with the government’s focus on economic development and instead invest in expanding production. In return, companies may request governmental assistance in the form of loans to relieve their financial constraints. Compared to firms without financial constraints, firms experiencing greater financial distress are particularly inclined to collaborate with governments to meet their GDP requirements. Consequently, they may resort to engaging in more greenwashing practices to fulfill shareholders’ demands for ESG initiatives without actually making substantial investments in such activities. If our hypothesis about the financial constraint channel is correct, we would expect to observe a greater effect of GDP manipulation on firm greenwashing when financial constraints are greater.

We employ the size-age (SA) index developed by Hadlock and Pierce [50] to assess the financial constraints faced by a company. A higher SA index signifies an elevated degree of financial constraint. We introduce the variable SA index (SA) and its interaction with GDP manipulation into Equation (1), with the aim of determining whether financial constraints play a role in the relationship between firm greenwashing and government GDP manipulation. The findings are presented in Column 1 of Table 5. Upon introducing these two additional variables, our test variable for GDP manipulation continues to exhibit a positive and statistically significant relationship. More importantly, the coefficient of the interaction between GDP manipulation and the SA index is positive and significant. This finding indicates that as financial constraints increase, the influence of manipulating GDP on the practice of firm greenwashing becomes stronger. This confirms our claim that manipulating GDP has an impact on how aggressively firms engage in greenwashing, which is influenced by the extent of their financial constraints.

Another channel we suggest is external monitoring. When firms are closely monitored, they may be reluctant to engage in greenwashing, even if they are aware that the local government expects them to contribute to GDP growth and invest in GDP-related activities. The likelihood of shareholders uncovering their greenwashing practices would increase. Should a company’s greenwashing practices be exposed, it could lead to significant repercussions, including a decline in stock prices and a loss of trust from both clients and suppliers. If our conjecture regarding the external monitoring channel is accurate, we would anticipate a diminished impact of GDP manipulation on the practice of firm greenwashing in firms subjected to closer monitoring.

We use the number of analysts as a proxy to represent the level of external monitoring that a firm encounters. Once again, we introduce two variables into Equation (1): the number of analysts (Analysts) and their interactions with GDP manipulation (GDP_Manip × Analysts). The test results are displayed in Column 2 of Table 5. The test we conducted on the variable of interest, which is the interaction between analysts and GDP manipulation, shows a negative and statistically significant result at the 10% level. This suggests that analyst monitoring has a negative effect on the relationship between firm greenwashing and government GDP manipulation. Put simply, when financial analysts closely monitor firms, the effect of government manipulation of GDP on firm greenwashing is reduced. Furthermore, our regression model confirms hypothesis 1, as we observe a significant and positive association between GW and GDP_Manip even after including the two new variables.

4.6. Heterogeneity Analysis

In this section, we analyze the circumstances under which the influence of government manipulation of GDP on firms’ greenwashing becomes more or less pronounced. The existing literature demonstrates that the level of institutional development, such as the advancement of laws and shareholder protection, influences the extent to which firms depend on governments. Greater institutional development in a region reduces firms’ dependence on governments. In contrast, in regions with underdeveloped institutional environments, firms must depend heavily on governments to secure loans and government contracts. To this end, we first select two moderators: the marketization development index in China and the anti-corruption campaign in China. Next, we select two additional moderators based on the distinctive attributes of the firms. The ownership structure of firms plays a crucial role in influencing various behaviors, such as earnings management and fraud. Thus, we employ the state ownership status of a firm as the third moderator. The business cycle of firms has been observed to have an impact on many decisions of firms, and it is tested as the fourth moderator.

We first investigate whether the development of institutions influences the relationship between a company engaging in greenwashing and the manipulation of government GDP. In line with previous research, we employ the marketization development index created by Fan et al. [51] as a metric for assessing institutional development in China. By calculating the average value of the marketization index for all provinces, we categorize our sample into two subgroups. The first subgroup comprises firm-year observations in provinces with a marketization index above the average. We then use this subgroup to estimate Equation (1) and present the findings in Column 1 of

Table 6. Heterogeneity test—marketization development.

	(1)	(2)
	High MKT	Low MKT
City-level Variable
GDP_Manip	−624.901	1251.716 **
	(1426.613)	(496.633)
Firm-level Variable
SIZE	−37.011	25.55
	(40.857)	(25.496)
Growth	21.67	−20.368
	(20.831)	(15.761)
LEV	187.138	139.546
	(118.657)	(100.021)
OWN1	−2.249	0.886
	(3.581)	(1.850)
OWN2_10	−1.431	0.353
	(1.183)	(1.304)
ROA	6.202	153.282
	(306.412)	(162.447)
CASH	184.215	231.646 *
	(119.290)	(133.122)
CFO	122.932	47.609
	(135.328)	(104.709)
AGE	−17.745	59.166
	(250.535)	(137.317)
INDEP	15.053	−19.699
	(272.299)	(160.261)
DUAL	41.084 *	0.66
	(22.919)	(22.000)
Constant	775.431	−1099.536 *
	(1102.635)	(638.521)
Firm fixed effect	Included	Included
Year fixed effect	Included	Included
Cluster	City	City
No. of Obs.	1058	1343
R-Squared	0.073	0.031

Notes: * p < 0.1; ** p < 0.05; standard errors in parentheses.

. The second group consists of firm-year observations situated in provinces with a marketization index below the mean. We present the findings obtained using this subgroup to re-estimate Equation (1) in Column 2 of Table 6. In the subgroup with weak institutional development (Column 2), the coefficient on GDP_Manip is positive and significant, whereas in the subgroup with strong institutional development (Column 1), the coefficient on GDP_Manip is not statistically significant. This aligns with our hypothesis that in a region where firms rely more heavily on governments, the influence of government manipulation of GDP on the practice of firm greenwashing would be more pronounced. Conversely, if firms have a low dependence on the government, the effect of government manipulation of GDP on the practice of firm greenwashing would be minimal.

China’s anti-corruption campaign, initiated in 2012, serves as the second moderator. The implementation of the anti-corruption campaign has been found to enhance the overall quality of China’s institutional environment. After the anti-corruption campaign, we anticipate that the connection between firm greenwashing and government GDP manipulation will become more pronounced. This is because an improved institutional environment decreases firms’ dependence on governments. The sample is divided into two segments: the period before the anti-corruption campaign (2007–2011) and the period after (2012–2019).

Table 7. Heterogeneity test—anti-corruption campaign.

	(1)	(2)
	Pre-Anti-Corruption	Post-Anti-Corruption
City-level Variable
GDP_Manip	2004.466 *	302.045
	(1135.430)	(552.987)
Firm-level Variable
SIZE	−88.37	−23.767
	(75.434)	(25.346)
Growth	−54.784	12.461
	(34.698)	(13.855)
LEV	−289.717	219.063 **
	(203.384)	(89.946)
OWN1	3.739	−0.469
	(2.654)	(2.001)
OWN2_10	0.011	−0.197
	(3.130)	(1.156)
ROA	107.633	154.046
	(348.291)	(164.907)
CASH	58.986	272.631 **
	(209.976)	(108.847)
CFO	−13.84	119.274
	(189.293)	(98.192)
AGE	94.177	111.823
	(402.094)	(173.461)
INDEP	−162.794	−42.742
	(340.864)	(129.321)
DUAL	−105.728 **	26.536
	(52.299)	(18.599)
Constant	1578.276	−114.474
	(2001.950)	(884.376)
Firm fixed effect	Included	Included
Year fixed effect	Included	Included
Cluster	City	City
No. of Obs.	361	2040
R-Squared	0.121	0.027

Notes: * p < 0.1; ** p < 0.05; standard errors in parentheses.

displays the regression results for the pre-anti-corruption campaign sample in Column 1 and the post-anti-corruption campaign results in Column 2. When the institutional environment is characterized by weakness and firms heavily depend on governments, there exists a positive and significant association between the test variable GDP_Manip and the dependent variable GW in Column 1. Nevertheless, this positive association diminished with decreasing firm dependence on governments and the launch of the anti-corruption campaign as evidenced by the regression findings in Column 2. The results align with the findings presented in Table 6, indicating that our baseline results are more pronounced in situations where the institutional environment lacks strength.

The third moderator in our group is the concept of state ownership. Research has shown that state-owned enterprises (SOEs) tend to be more amenable to supporting the objectives of governments. SOEs are more likely to assist when they recognize that governments prioritize increasing GDP. During this process, greenwashing behaviors can be influenced by the allocation of resources. For instance, allocating resources toward expanding production can lead to an increase in GDP while simultaneously reducing the availability of resources for ESG initiatives. Compared with SOEs, non-state-owned enterprises (non-SOEs) are less susceptible to government intervention aimed at boosting GDP. We once again split our sample into two separate subsamples: SOEs are companies that are ultimately under the control of governments, as opposed to non-SOEs. The ultimate controller data are from the CSMAR database. The regression results of employing SOEs to estimate Equation (1) in Column 1 of

Table 8. Heterogeneity test—state ownership.

	(1)	(2)
	SOEs	Non-SOEs
City-level Variable
GDP_Manip	774.766 *	280.971
	(461.794)	(1031.094)
Firm-level Variable
SIZE	−11.493	−15.375
	(26.997)	(29.725)
Growth	8.752	−5.829
	(16.826)	(22.015)
LEV	194.881 **	39.403
	(99.269)	(105.482)
OWN1	−2.481	1.585
	(2.606)	(1.413)
OWN2_10	−1.188	0.66
	(1.543)	(1.378)
ROA	31.37	105.217
	(195.276)	(226.424)
CASH	285.717 **	72.99
	(132.828)	(154.932)
CFO	7.918	250.581
	(93.677)	(155.516)
AGE	−71.469	219.468
	(132.434)	(223.401)
INDEP	−15.237	−20.222
	(166.936)	(224.528)
DUAL	18.034	31.403
	(20.221)	(26.888)
Constant	241.826	−474.66
	(752.596)	(866.049)
Firm fixed effect	Included	Included
Year fixed effect	Included	Included
Cluster	City	City
No. of Obs.	1490	911
R-Squared	0.073	0.050

Notes: * p < 0.1; ** p < 0.05; standard errors in parentheses.

reveal a positive and statistically significant association between corporate greenwashing and the government manipulation of GDP. For Column 2 of Table 8, we employed the non-SOEs subsample to calculate the estimation of Equation (1). However, the coefficient on GDP_Manip is no longer statistically significant. These findings suggest that firm ownership influences our baseline results.

The last moderator of our discussion is the business life cycle. The conventional business life cycle encompasses the stages of startup and growth, maturity, and decline. The literature extensively documents that firms display distinct behaviors at various stages of their life cycle. We anticipate that the manipulation of government GDP figures will have a more pronounced impact on the practice of greenwashing among firms that are in the mature stage. This is due to the fact that these companies encounter more constraints in acquiring funding for additional expansion, and their already significant scale facilitates their ability to address external oversight. On the other hand, companies in the early stages of development and expansion are relatively small and encounter fewer limitations in acquiring funding, although they have limited ability to manage monitoring. Hence, the influence of manipulating GDP on the practice of greenwashing may not hold much significance for this specific cohort of companies. In the decline stage, firms have minimal financing requirements. Consequently, the influence of government manipulation of GDP on greenwashing may not have a substantial effect on the firms that are going through a period of economic decline.

The sample is categorized into three groups: firm-year observations during the startup and growth phase, samples during the maturity phase, and samples during the decline phase. We performed the baseline regression for each subgroup and present the results in

Table 9. Heterogeneity test—business life cycle.

	(1)	(2)	(3)
	Startup and Growth	Maturity	Transition and Succession
City-level Variable
GDP_Manip	1000.479	1653.557 **	411.262
	(637.101)	(683.797)	(1558.390)
Firm-level Variable
SIZE	−31.352	32.093	30.619
	(27.591)	(42.410)	(84.195)
Growth	−4.366	0.724	−22.527
	(18.784)	(27.506)	(39.130)
LEV	155.56	36.699	236.181
	(103.571)	(120.357)	(291.494)
OWN1	−3.757 *	3.000	1.585
	(2.119)	(2.047)	(4.423)
OWN2_10	−2.37	2.65	0.574
	(1.636)	(1.819)	(3.299)
ROA	307.248 *	−93.307	−507.926
	(167.734)	(254.871)	(528.738)
CASH	272.849 *	−16.285	148.76
	(140.605)	(157.695)	(291.587)
CFO	113.602	183.432	−83.111
	(107.463)	(185.387)	(274.188)
AGE	−70.212	−161.079	−1467.328
	(170.369)	(128.480)	(1141.548)
INDEP	−28.67	−99.062	640.655 **
	(171.795)	(256.993)	(302.582)
DUAL	22.529	−1.702	−16.89
	(19.933)	(36.166)	(48.363)
Constant	746.892	−681.415	3532.668
	(707.900)	(1172.349)	(3966.459)
Firm fixed effect	Included	Included	Included
Year fixed effect	Included	Included	Included
Cluster	City	City	City
No. of Obs.	1186	601	229
R-Squared	0.085	0.044	0.209

Notes: * p < 0.1; ** p < 0.05; standard errors in parentheses.

. Our analysis reveals that the coefficient on GDP_Manip is statistically significant only in the mature subgroup (Column 2), while it is not statistically significant in the other two subgroups (Columns 1 and 3). Thus, our prediction is supported.

In summary, our findings indicate that our baseline result (i.e., the positive correlation between firms’ greenwashing and local governments’ manipulation of GDP) is particularly pronounced in regions with less developed institutional frameworks, prior to China’s anti-corruption campaign, among state-owned enterprises, and in firms in the mature stage of the business life cycle.

5. Discussion and Conclusions

The prevalence of greenwashing scandals has increased in recent years. According to a report published by Standard and Poor, over 44% of investors prioritize addressing the issue of greenwashing when making ESG investments. Inspired by this phenomenon, our objective is to investigate the factors that impact companies’ greenwashing practices [1]. The existing research on the determinants of greenwashing is relatively scarce compared to the extensive studies on the consequences of greenwashing. In particular, the previous literature has largely overlooked the role of government in this area. Therefore, we investigated the correlation between firms’ greenwashing and local governments’ GDP manipulation.

We found a positive and statistically significant correlation between the level of greenwashing practiced by firms and the level of GDP manipulation carried out by the local governments. Further empirical findings indicate that the financial constraints of companies and external oversight are the means by which governments exert influence on companies’ greenwashing practices. Moreover, the baseline results are more pronounced in regions with a lower level of marketization, during periods before China’s anti-corruption campaign, within state-owned companies, and among firms in the mature stage of their business life cycle. The results of our study remained robust to various tests, including a two-stage regression analysis to account for potential endogeneity problems.

Our study contributes to the existing literature by deepening our understanding of the factors that influence greenwashing. In addition, our study presents a novel approach to quantify greenwashing. Instead of conducting a comparison between two ESG rankings provided by third-party agencies, we evaluated the environmental investment input and the resulting ranking given by the third-party agency to proxy greenwashing. We argue that this measure could more effectively indicate when firms are exaggerating their ESG performance without actually investing in the necessary work. According to agency theory, managers tend to prioritize their own interests over shareholder returns, which may contribute to greenwashing. In this sense, our findings are applicable to similar contexts, even though our results are based on data from China.

Nevertheless, it is imperative to acknowledge the limitations of our new greenwashing measure. Our measure primarily emphasizes the “E” aspect, neglecting to account for the investment in the “S” and “G” aspects. This is because few firms disclose the expenses or investments for “S” and “G” separately. Instead, most combine them with other expenses and report them under the general and administrative expenses category. We advocate for future research to enhance our metric by integrating these two factors into the measurement.