The Impact of Economic Policy Uncertainty on Executives’ Self-Interest Behaviors: Evidence from China

Abstract

This paper empirically studies the impact of economic policy uncertainty on executives’ self-interest behaviors, distinguishes explicit self-interest behaviors from implicit ones, and studies the moderating effect of internal control. The results illustrate that rising policy uncertainty will inhibit explicit self-interest behaviors of executives, yet the implicit ones will be encouraged. Internal control can regulate the above effects. Further research proves that the above-mentioned impact is more significant in state-owned enterprises (SOEs). Stable institutional investors and sound market competition can play a certain role in governance. Our paper contributes to the literature on the impact of economic policy uncertainty on corporate governance.

1. Introduction

In recent years, a series of major global events, including the Sino-US trade friction, the outbreak of the COVID-19 pandemic, the Russia-Ukraine conflict, etc., have significantly elevated economic uncertainty, and the growth of many major economies worldwide has been hit hard. Economic development has been restricted by multiple factors, encountering various challenges and subdued growth. The uncertainty in global economic prospects has increased. China is in the midst of an increasingly grave and complicated international landscape, involving the frequent and sporadic pandemic outbreak, the ongoing “new normal” of pandemic control measures, and the boosted downward pressure on the economy. Owing to the global trend of economic integration, the sustainability and stability of China’s growth is restricted by multiple factors, and the uncertainty of domestic economic policies is still high. Against this background, it is important to measure uncertainty of economic policy and study its impact.

Differently from other countries in the world, China uses the term “socialist market economy with Chinese characteristics” to describe its theory of economic policy. Since 1949, the central government in China has played a dominant role in economic life in the country. Furthermore, since 1978, China has enacted reform and started to use market economy tools to power up its economic development. Overall, while welcoming the free market economy, China still regarded the planned economy as the main component of China’s national economy.

China’s economic theory is relatively similar to the Keynesian economic theory. It advocates that the state adopts expansionary economic policies to promote economic growth by increasing demand; that is, to expand government spending, implement fiscal deficits, stimulate the economy, and maintain prosperity. From this perspective, considering the dependence of the economy on the government, uncertainty in the economic policy of China can exert a significant impact on the economy.

Economic uncertainty is a state in which economic entities cannot have accurate expectations on such issues as “whether to adjust policies”, “how to adjust policies”, and “the outcome and effect of policy execution”. Under rising uncertainty, the government will introduce corresponding macroeconomic policies to stabilize the market and avert growth dilemma. Meanwhile, the adoption of new macro-control measures brings about new uncertainty. The uncertainty of economic policy will be transmitted to or directly affect enterprises, exerting impact on their behaviors. The changes in economic policies will significantly impact enterprises. For enterprise executives, in order to cope with the possible impact of economic policy uncertainty on enterprises and themselves, they will adjust strategic direction and behaviors in a timely manner in response to the external environment, so as to achieve stable development. In this process, facing the adverse effect of uncertainty, due to agency problems, executives may sacrifice the interests of shareholders and enterprises to protect their own interests.

The manner by which to alleviate the principal-agent conflict between managers and shareholders has always been the focus of corporate governance, and its causes and preventive measures have always been highlighted by the academic community. The self-interest behavior of executives is a typical embodiment of the principal-agent problem. According to principal-agent theory, executives adopt self-interest behaviors by sacrificing the interests of shareholders and enterprises to obtain personal benefits. As the utility functions of management and shareholders are different, and the separation of ownership and control leads to information asymmetry and incentive problems, management often seeks benefits for itself through various channels in order to maximize its own interests, which may damage the interests of shareholders and the company, negatively affecting enterprise performance. In the long run, the development of capital market becomes constrained and the healthy development of economy and society endangered [1].

Exploring the causes of executives’ self-interest behavior is helpful to raise people’s awareness to executives’ irrationality and lay the foundation for taking corresponding countermeasures to executives’ self-interest behavior. The existing literature studies the causes of such self-interest behaviors, most of which take the perspectives of corporate governance structure [2], government intervention [3], etc. The uncertainty of external economic policies, a factor often overlooked, also affects the behavior of executives. Existing research has found that the occurrence of executives’ self-interest behavior is closely related to the external environment. The cognition and behavioral choices of executives are often influenced by the macro environment. The impact of rising uncertainty of external economic policy on executives’ self-interest behavior is still unknown. Based on the theory of myopic behavior, for the sake of maximizing personal interests, higher risks will induce executives to pursue short-term interests, which can be quickly materialized, at the price of long-term ones. Based on prospect theory, with the risk of external uncertainty, executives are subject to higher dismission risk and tougher employment conditions. As a result, their decisions and behaviors will be more conservative, thus inhibiting self-interest behaviors. In conclusion, from the theoretical point of view, the jury is still out on whether executives’ self-interest behavior is influenced by the uncertainty of external economic policy.

Based on the existing research, this paper studies the influence of economic policy uncertainty on executives’ self-interest behavior, and introduces the moderating effect of internal control, so as to examine the changes of such influence under different levels of corporate internal control. Selecting A-share listed companies in China from 2010 to 2021 as research samples, this paper empirically discovers that the uncertainty of economic policy will inhibit the explicit self-interest behavior of executives, namely excessive executive compensation. However, the uncertainty of economic policy encourages, rather than restrains, the implicit self-interest behavior of executives, namely excessive on-the-job consumption. Further research illustrates that, under the uncertainty of economic policy, effective internal control is conducive to restraining executives’ implicit self-interest behavior, as well as its explicit counterpart. Finally, we found that the above effect is more significant in SOEs. In addition, strong market competition and stable institutional investors are conducive to restraining executives’ implicit self-interest.

2. Literature Review

The causes of policy uncertainty are failure of the government to clarify the direction and intensity of economic policy expectation, policy execution, and changes in policy position [4]. Economic uncertainty is difficult for enterprises to predict, which will enhance the external risks of business operations, and affects executives’ cognition and behavioral decisions.

The existing literature measures the uncertainty of economic policy from the following three perspectives. The first type of literature measures the uncertainty of economic policy through the occurrence of external events, including major political events and changes in bureaucrats [5]. Since the occurrence of these events is not ongoing, their usage is subject to certain limitations. The second category uses the implied volatility of the stock market as the proxy variable of economic policy uncertainty. However, fluctuation in the financial market, which often lags behind, cannot reflect external economic policy uncertainty in real time. The third category of literature adopts the economic policy uncertainty index developed by Baker et al., based on the text analysis of newspapers. This method is authoritative, real-time and continuous, based on the reports of authoritative newspapers around the world. In this paper, we adopt the third type of measurement method.

At present, the research on the uncertainty of economic policy mostly focuses on corporate behavior, including corporate investment and financing, enterprise innovation, cash holdings [6,7,8,9], etc. The research related to economic policy uncertainty and executives mostly adopts the traits of executives as intermediaries or moderating variables to further study the impact of economic policy uncertainty on corporate behavior, for instance, their career records, including political background and financial experience, overconfidence, management governance [7,10,11,12,13,14,15,16], etc. Few scholars focus on the impact of economic policy uncertainty on executives’ cognition and behavioral decisions. Based on the perspective of internal risk hedging, Rao [17] found that when the external uncertainty is high, the probability of executive turnover decreases. Wu [18] found that the uncertainty of economic policy is negatively correlated with executives’ overconfidence. From the perspective of prospect theory, the uncertainty of economic policy comprises the uncertainties from policy change, policy execution, and the alterations of government positions. Facing external uncertainty, executives have uncertain expectations for the future of enterprises and themselves, which will lead to irrational bias in cognition and behavior choice, affecting their self-interest behaviors. At present, existing research seldom studies the relationship between economic policy uncertainty and executives’ self-interest behavior, and related research needs to be further expanded.

According to the principal-agent theory, executives’ self-interest behavior refers to the exploitation of their own rights and information asymmetry [19], in order to maximize their interests at the cost of shareholders and the enterprise, as illustrated by surging on-the-job consumption [20], seeking excess compensation [21], myopic behavior, cash flow manipulation [3], etc. In order to reflect the common behaviors of executives in listed companies, and present the explicit manifestations of their self-interest behavior, most scholars differentiate explicit self-interest behaviors from the implicit ones for quantitative research [1]. For explicit self-interest behavior, scholars commonly adopt excessive compensation [22,23], etc. With regard to implicit self-interest behavior, scholars use variables including excessive on-the-job consumption, management expenses, turnover of total assets [2,24,25,26], etc. Enterprises often take measures, such as supervision, reward, punishment, and incentive, to restrain the self-interest behavior of executives, and scholars mostly conduct research through the above related measures, including internal and external supervision, salary mechanism [27,28], etc. Change in the external macro-environment also affects executives’ cognition and behavior choices. Executives will be passively restrained from self-interest behavior, or they may actively reduce such behaviors in exchange for stable income. The influence of economic uncertainty on executives’ self-interest behavior needs to be further explored.

According to both the domestic and foreign literature, internal control reduces the asymmetry of internal and external information, strengthens internal supervision, plays an important role in reducing executive fraud [29] and preventing insider trading by executives [30], and exerts certain governance effects on both the implicit and explicit corruption of executives [31]. The above facts also lay a theoretical basis for choosing internal control as the moderating variable in this paper. With regard to the measurement of internal control quality, some scholars made a comprehensive evaluation through the five elements of internal control [32,33]. Scholars often adopt the internal control index of the DIB database to measure the quality of internal control [34,35].

By reviewing the existing literature, we can find that the research on the influencing factors of executives’ self-interest behavior mostly focuses on corporate institutions, the strength of executives’ power, etc., and seldom considers the influence of the external economic policy environment. Based on existing theories, negative cognitive theory induces executives to boost self-interest behaviors and gain immediate benefits under uncertain economic policies; however, based on the potential cost and risk, it is believed that due to the superposition effect of risks, executives are urged to restrain self-interest and tend to “protect themselves” against the risk of dismissal, thus reducing their self-interest behavior. However, the research on the uncertainty of economic policy focuses more on its impact on macroeconomic trends, corporate investment and financing, innovation and R&D, and financial market fluctuation. Macroeconomic policy and the fluctuation in economic fundamentals are the important foundations of enterprise behavior on the micro-level, and the uncertainty of economic policy will profoundly affect and change corporate behavior and decision-making. Under different degrees of policy uncertainty, executives’ self-interest behavior will be affected differently. In this paper, the macro-economic landscape is brought into the analysis framework of influencing factors of executives’ self-interest behavior, and their self-interest behavior is refined into explicit and implicit categories. These two categories are quantitatively analyzed by adopting excessive compensation and on-the-job consumption, and the uncertainty of economic policy is adopted to analyze how the external environment affects executives’ self-interest behavior, enriching the existing research results.

3. Theoretical Analysis and Research Hypothesis

This paper categorizes executive self-interest behaviors into explicit and implicit ones, and analyzes the influence of economic policy uncertainty on such behaviors. The explicit self-interest behavior of executives is exemplified by excessive compensation, and its implicit counterpart is measured by excessive on-the-job consumption.

On the one hand, at the corporate level, rising economic policy uncertainty amplifies the business risks of enterprises and exerts a negative impact on enterprise performance [36]. Changes in the economic environment will affect the survival and growth of enterprises. Rising economic policy uncertainty will make it difficult for management to predict the prospects of the market, as well as corporate financing and cash flow status in the future, making it more difficult to make decisions. In order to avert risks, C-suite often delays investment and lowers investment volume to alleviate the uncertainty, thus adversely affecting the performance of enterprises. Meanwhile, the financing of enterprises will also be affected by macroeconomic policies. Rising economic uncertainty will amplify instability in the future cash flow of enterprises and also aggravate the external financing challenges of enterprises. From the perspective of creditors such as banks, when economic policy is uncertain, the risk of loan default will rise, since it will be difficult to evaluate the credit of enterprises. Therefore, creditors will adopt a tight credit policy, raise the loan threshold, and reduce the loan amount, which pushes up the financing costs of enterprises and further undermines their development. Facing both internal and external difficulties, in order to maintain stable operation and ease operating pressure, enterprises will start to reduce expenses and compensation. On 2 June 2022, the State-owned Assets Supervision and Administration Commission (SASAC) issued the Notice on Further Implementing the Tasks of Boosting Income and Reducing Expenditure of SOEs, which stipulated: “SOEs should strictly implement the linkage mechanism between total wages and profits, and the growth rate of total compensation shall not exceed that of total profit. SOEs with declining profits must cut total compensation.” Therefore, when the uncertainty of economic policy is high, enterprises will face greater business risks, which will reduce C-suite enumeration, inhibiting excessive compensation of executives and their explicit self-interest behavior.

On the other hand, higher uncertainty of economic policy will lead to tougher employment. With rising external uncertainty, the turnover of executives will be more frequent, boosting the risk of dismissal [37]. Based on prospect theory, people tend to choose “safe returns” instead of “taking a chance”. Even though self-interest behavior may bring some excess returns, if such behavior can be uncovered easily, the risk of turnover will also surge. Therefore, executives will suppress their explicit self-interest behavior to obtain “safe returns”. Meanwhile, some executives choose to fulfill their social responsibility by taking the initiative to reduce salaries before ordinary employees. According to Vanke’s 2021 annual report, due to the sharp decline in corporate performance, Yu Liang, chairman of the board of directors, voluntarily waived all bonuses in 2021 and reduced his compensation by nearly 90%. Eight directors, supervisors, and senior managers, including Yu Liang, reduced their salaries by half, totaling 24 million yuan. These executives not only fulfill their social responsibility, but also build good reputation and lay a solid foundation for their future career development. Therefore, when the uncertainty of economic policy is high, executives themselves will also choose to reduce their explicit self-interest behavior and excess remuneration.

However, executives’ implicit self-interest, namely excessive on-the-job consumption, is affected differently by the uncertainty of economic policies. On-the-job consumption, also known as perquisite consumption or perk consumption, is the nonmonetary consumption spent by the managers, incurred when the manager is performing duties. On-the-job consumption is generally included in administrative expenses, which can be further divided into office allowance, travel expenses, correspondence expenses, etc. Although the “optimal” level of on-the-job consumption may motivate managers to perform well, excess on-the-job consumption may hurt firm value [38].

Managers increase their nonmonetary compensation to achieve self-utility maximization when the government supervision or other restrictions are weak [28]. With weak legal constraints and corporate governance mechanisms, on-the-job consumption has become an important part of management compensation in China, far more even than monetary compensation. Furthermore, on-the-job consumption is implicit, has no contractual constraints, and is largely determined by the top executives; executives are likely to deviate from the objective of maximizing stakeholder value and use on-the-job consumption to gain self-interest [39].

Higher uncertainty of economic policy leads to greater information asymmetry in enterprises; thus, the difficulties of internal and external supervision are also elevated, which facilitates executives to make use of their authority for on-the-job consumption and implicit self-interest behavior. Furthermore, as analyzed above, whether voluntary or forced, when the external uncertainty is high, the explicit remuneration of executives will decline. Therefore, executives are more inclined to use their authority to earn implicit benefits through on-the-job consumption to make up for salary loss. In addition, rising uncertainty of external economic policies often brings about greater financing difficulties for enterprises. In order to prevent capital chain rupture, enterprises tend to reduce investment while holding excess cash to cope with financial risks, which also provides opportunities for executives’ invisible self-interest behavior.

In conclusion, concerning the impact of economic policy uncertainty on executives’ self-interest behavior, this paper puts forward the following assumptions:

Hypothesis 1 (H1a).

Rising economic policy uncertainty will inhibit the explicit self-interest behavior of executives.

Hypothesis 1 (H1b).

Rising economic policy uncertainty will encourage the implicit self-interest behavior of executives.

Executives’ responses to the fluctuation of external economic policy uncertainty are also influenced by the internal organizational structure and supervision system of their enterprises. Facing the uncertainty of the external macro-environment, enterprises need to improve internal control to better manage production and business activities and counter unknown risks. An effective internal control system can restrain executive opportunism, reduce information asymmetry, and alleviate the principal-agent problem between shareholders and C-suite, which is conducive to the long-term development of enterprises. “Internal environment”, one of the five elements of internal control, requires enterprises to set up corresponding supervision mechanisms, establish sound organizational mechanisms, realize internal power checks and balances, avert “tyranny of the minority” and “insider control issue”, and supervise the behavioral decisions of executives. Among the five elements of internal control, “Risk Assessment” requires enterprises to assess the possible risks in the future and formulate corresponding policies to prevent executives from speculating on the grounds of external risks. “Control activities” require enterprises to improve various management systems, better control the actual income and cost of each production and operation procedure, fix the loopholes of which C-suite may take advantage, and prevent their self-interest behaviors. “Information communication” encourages enterprises to reduce internal and external information asymmetry and improve the quality of information disclosure, which is conducive to the supervision of agents by stakeholders.

To sum up, this paper puts forward the following assumptions:

Hypothesis 2 (H2a).

Under economic policy uncertainty, effective internal control can enhance the inhibition of executives’ explicit self-interest behavior.

Hypothesis 2 (H2b).

Under economic policy uncertainty, effective internal control can weaken the facilitation of executives’ explicit self-interest behavior.

4. Research Design

4.1. Data Sources and Sample Selection

This paper adopts the data of A-share-listed companies in the Shanghai and Shenzhen stock exchanges from 2006 to 2021 as the research object, and has taken the following screening measures: (1) excluding the listed companies that were classified as ST and PT at the end of 2012 to 2021; (2) excluding listed companies in the financial sector; (3) excluding the listed companies that were forcibly delisted during the sample period; (4) eliminating the samples with missing observation values in the selected period; meanwhile, the continuous variables are winsorized at the top and bottom 1%. A total of 24484 observation samples were obtained after screening. The related data of explanatory and control variables in this paper were extracted from CSMAR, the economic policy uncertainty index of explanatory variables was sourced from the http://www.policyuncertainty.com/ website (accessed on 3 November 2022), and the internal control index of moderating variables was generated from the DIB database.

4.2. Definition of Variables

4.2.1. Explanatory Variable

In this paper, the self-interest behavior of executives is regarded as the explanatory variable, which is divided into explicit and implicit self-interest behavior, measured by C-suite excessive compensation (OverPay) and excessive on-the-job consumption (UnPerks), respectively.

The detailed measurement of explicit self-interest behavior of executives-excessive compensation (OverPay) is as follows. In this paper, we adopt “the total compensation of top 3 executives” disclosed in the annual reports of A-share-listed companies in the Shanghai and Shenzhen stock exchanges as the data of executives’ salaries, and the variables of C-suite compensation were obtained using a taking logarithm. Next, we drew from the practices of Core et al. [40], Luo et al. [41], etc. to measure the excessive compensation of executives. Firstly, Model (1) is regressed by years and industries using sample data to obtain the regression coefficient. Then, the estimated regression coefficient is multiplied by the factors that affect C-suite compensation, such as corporate scale and performance, to estimate the expected level of executive compensation; finally, according to Model (2), the actual C-suite compensation is subtracted from the expected C-suite compensation, and the difference is excessive compensation. The definitions of variables used in Model (1) are presented in

Table 1. Definition table of overpay variables of executives.

Variable	Name	Definition
Pay	Executive absolute compensation	The natural logarithm of the total remuneration of the Top 3 executives of listed companies
Size	Corporate size	The natural logarithm of the total income of the listed company in the same year
ROA	Accounting performance	Ratio of net profit to total assets
IA	Intangible assets ratio	Ratio of intangible assets to total assets
Zone	Region dummy variable	If corporate place of registration is coastal area, the value is 0; otherwise, the value is 1
Industry	Industry dummy variable
Year	Year dummy variable

.

$$Pa{y}_{i,t}={\alpha }_0+{\alpha }_{1}Siz{e}_{i,t}+{\alpha }_{2}RO{A}_{i,t}+{\alpha }_{3}I{A}_{i,t}+{\alpha }_{4}Zon{e}_{i,t}+{\displaystyle \sum }Industry+{\displaystyle \sum }Year+{\epsilon }_{i,t}$$

(1)

$$OverPay=Pa{y}_{i,t}-ExpectedPa{y}_{i,t}$$

(2)

The detailed measurement of implicit self-interest behavior of executives-excessive on-the-job consumption (UnPerks) is as follows. We draw from the research models of Quan et al. [42], Luo et al. [43], etc. and predict the normal on-the-job consumption of executives through the relevant corporate conditions of Model (3). The difference between actual and normal on-the-job consumption is abnormal on-the-job consumption. Among them, the C-suite on-the-job consumption is the balance of management expenses after deducting the total annual compensation, amortization of intangible assets, total provision for bad debts, and total provision for inventory depreciation. The definitions of variables used in Model (3) are presented in

Table 2. Definition table of excessive on-the-job consumption variables of executives.

Variable	Name	Definition
Perks	On-the-job consumption of executives	Management expenses-total annual compensation of the management-amortization of intangible assets-total bad debt reserve-total inventory depreciation reserve
Asseti,t−1	Ending total assets	Total assets at the end of last period
$\Delta sal{e}_{i,t-1}{}_{-}$	Changes in main business income	Change of main business income in the current period
PPEi,t	Fixed assets	Net value of fixed assets in the current period
Inventoryi,t	Stock	Total inventory in current period
LnEmployeei,t	Number of employees	The natural logarithm of the total number of employees of the enterprise

.

$$\frac{Perk{s}_{i,t}}{Asse{t}_{i,t-1}}={\beta }_0+{\beta }_1\frac{1}{Asse{t}_{i,t-1}}+{\beta }_2\frac{\Delta sal{e}_{i,t}}{Asse{t}_{i,t-1}}+{\beta }_3\frac{PP{E}_{i,t}}{Asse{t}_{i,t-1}}+{\beta }_4\frac{Inventor{y}_{i,t}}{Asse{t}_{i,t-1}}+{\beta }_{5}LnEmploye{e}_{i,t}+{\epsilon }_{i,t}$$

(3)

$$UnPerks=Perk{s}_{i,t}-ExpectedPerk{s}_{i,t}$$

(4)

4.2.2. Explanatory Variables

The Economic Policy Uncertainty Index (EPU) of China is drawn from the website of http://www.policyuncertainty.com/ (accessed on 3 November 2022). From the newspapers People’s Daily and Guangming Daily, Davis, Liu, and Sheng adopt the method of text analysis, counting articles related to economic policy uncertainty through scaling frequency, and formulating EPU after standardization treatment. At present, scholars at home and abroad use EPU when conducting research on economic policy uncertainty. This paper draws on the practice of Gu et al. [8]: the monthly data are arithmetically averaged, and then the logarithm is taken to obtain the explanatory variable EPU. Formula (5) provides more details.

$$EPU=LN\left(\frac{\sum Economic policy uncertainty monthly}{12}\right)$$

(5)

4.2.3. Moderating Variables

In this paper, moderating variables are selected from the DIB internal control index, which is based on the five elements of internal control and granularly measures the internal control quality of listed companies. When the index is high, corporate internal control management activities are effective and the internal control system is sound. In this paper, we refer to the practices of Zhou et al. [44], etc., to establish an internal control moderating index after taking the logarithm of the internal control index.

4.2.4. Control Variables

In this paper, the following indicators are selected as control variables, and the descriptions of each variable are illustrated in

Table 3. Variable definition table.

Variable	Name	Meaning	Measurement Method
Explanatory variable	OverPay	Executive overpayment	Modeling and calculation
	UnPerks	On-the-job consumption of executives	Modeling and calculation
Explanatory variable	EPU	Economic policy uncertainty index	Calculation
Moderating variable	IC	Internal control index	Ln (internal control index score)
Control variable	Size	Corporate size	Ln (total assets)
	Lev	Corporate capital structure	Total liabilities/total assets
	ROA	Return on total assets	Total return on assets adjusted by industry median
	Cashflow	Cash flow ratio	Net cash flow from operating activities divided by total assets
	SOE	Nature of the property right	If the company is a state-owned enterprise, the value is 1; otherwise, it is 0
	Indep	Proportion of independent directors	Number of independent directors/number of board members
	Dual	Combination of two powers	If the chairman and CEO are the same person, the value is 1; otherwise, it is 0
	Top1	Equity concentration	Shareholding ratio of the largest shareholder
	Balance	Equity balances and checks	Shareholding ratio of the second largest shareholder/shareholding ratio of the largest shareholder
	Big4	Audit quality	If the auditor is one of “Big 4” accounting firms, the value is 1; otherwise, it is 0
	INST	Share holding ratio of institutional investors	Total shares held by institutional investors divided by circulating equity capital
	Industry	Industry	Industry dummy variable to control industrial influence
	Year	Year	Year dummy variable to control the influence of the year

.

4.3. Model Setting

4.3.1. Economic Policy Uncertainty and Executives’ Self-Interest Behavior

In order to validate the relationship between economic policy uncertainty and executives’ self-interest behavior, this paper constructs the following econometric regression models. Model (6) is used to test the influence of economic policy uncertainty on executives’ explicit self-interest behavior, so as to verify hypothesis H1a in this paper. According to H1a, ${\alpha }_1$ of Model (6) should be significantly negative. Model (7) is used to test the influence of economic policy uncertainty on executives’ implicit self-interest behavior, so as to verify hypothesis H1b in this paper. According to H1b, ${\beta }_1$ of Model (7) should be significantly positive.

$$OverPay={\alpha }_0+{\alpha }_{1}EPU+{\alpha }_n\sum Controls+{\epsilon }_{i,t}$$

(6)

$$Perk={\beta }_0+{\beta }_{1}EPU+{\beta }_n{\displaystyle \sum }Controls+{\epsilon }_{i,t}$$

(7)

4.3.2. Moderating Role of Internal Control

Models (8) and (9) are used to test hypotheses H2a and H2b in this paper, respectively. By adding an interaction term including economic policy uncertainty and the internal control index to the regression equation, we can discuss the moderating function of internal control. Among them, in order to avoid the collinearity problem, in the interaction term of economic policy uncertainty and internal control index, the decentralized values of the two factors are multiplied.

$$OverPay={\alpha }_0+{\alpha }_{1}EPU+{\alpha }_{2}IC+{\alpha }_{3}EPU\times IC+{\alpha }_n\sum Controls+{\epsilon }_{i,t}$$

(8)

$$Perk={\beta }_0+{\beta }_{1}EPU+{\beta }_{2}IC+{\beta }_{3}EPU\times IC+{\beta }_n{\displaystyle \sum }Controls+{\epsilon }_{i,t}$$

(9)

5. Empirical Results and Analysis

5.1. Descriptive Statistical Analysis

Table 4. Descriptive statistical analysis.

	Ccount	Mean	S.D.	Min	P50	Max
OverPay	24,484	−0.0065	0.5603	−1.5678	−0.0181	1.5047
UnPerks	24,484	−0.0003	0.0235	−0.1168	−0.0007	0.0935
EPU	24,484	5.2291	0.5374	4.5230	5.3310	5.9671
IC	24,484	6.4722	0.1557	5.1402	6.4999	6.8512
Perk	24,484	374.1991	2270.3217	−14,741.5908	106.8724	83,383.8516
Pay	24,484	2.5679	289.1930	0.00005	1.852210	79.9235
Size	24,484	22.3851	1.2867	19.5245	22.2129	26.4297
Lev	24,484	0.4503	0.2009	0.0274	0.4481	0.9246
ROA_m	24,484	−0.0053	0.0630	−0.4595	−0.0052	0.2299
SOE	24,484	0.3989	0.4897	0.0000	0.0000	1.0000
Cashflow	24,484	0.0462	0.0680	−0.2244	0.0449	0.2568
Indep	24,484	0.3756	0.0543	0.2727	0.3636	0.6000
Dual	24,484	0.2485	0.4321	0.0000	0.0000	1.0000
Top1	24,484	0.3358	0.1472	0.0813	0.3114	0.7584
Balance	24,484	0.3473	0.2857	0.0062	0.2590	1.0000
Big4	24,484	0.0614	0.2401	0.0000	0.0000	1.0000
INST	24,484	0.4202	0.2298	0.0000	0.4324	0.8867

illustrates the descriptive statistical results of related variables. The average value of OverPay is −0.006, which indicates that C-suite excessive compensation has been restrained to a certain extent in recent years. The median and average values of OverPay are both negative, indicating that most executives relinquish overpayment, yet the maximum value is 1.505, which is still high. The standard deviation of UnPerks is 0.023, which indicates that C-suite excessive on-the-job consumption in different enterprises is not divergent, and some executives still have high on-the-job consumption. The mean value of EPU is 5.229, with the standard deviation of 0.537, showing that the uncertainty of China’s economic policy is still high, and the economic fundamentals are still volatile. The average value of Perks is 374.1991, which is much larger than the average value of Pay. The standard deviations of these two variables are relatively larger than other variables presented in the table, which indicates that the salary and consumption of managers varies greatly across different companies. The above results are in line with the fact that the managers in China, particularly managers in SOEs, are more likely to receive lower monetary compensation. With weak monitor mechanisms, managers in China are more likely to use on-the-job consumption to gain self-interest. Meanwhile, the standard deviation of corporate scale is 1.287, indicating quite different enterprise sizes.

5.2. Correlation Analysis

The results of the Pearson correlation coefficient among variables involved in this model are illustrated in

Table 5. Correlation analysis of variables.

	OverPay	UnPerks	EPU	IC	Size	Lev	ROA	SOE	Cashflow	Indep	Dual	Top1
OverPay	1.000
UnPerks	0.089 ***	1.000
EPU	−0.023 ***	0.015 **	1.000
IC	0.019 ***	0.022 ***	−0.120 ***	1.000
Size	0.041 ***	−0.103 ***	0.117 ***	0.187 ***	1.000
Lev	−0.067 ***	−0.112 ***	−0.054 ***	−0.017 ***	0.471 ***	1.000
ROA	0.021 ***	0.113 ***	−0.076 ***	0.354 ***	0.096 ***	−0.268 ***	1.000
SOE	−0.091 ***	−0.065 ***	−0.113 ***	0.086 ***	0.295 ***	0.245 ***	−0.021 ***	1.000
Cashflow	0.045 ***	0.108 ***	0.083 ***	0.122 ***	0.062 ***	−0.178 ***	0.365 ***	−0.013 **	1.000
Indep	−0.019 ***	−0.023 ***	0.055 ***	−0.007	0.016 **	−0.009	−0.029 ***	−0.057 ***	−0.014 **	1.000
Dual	0.052 ***	0.017 ***	0.061 ***	−0.030 ***	−0.131 ***	−0.098 ***	−0.008	−0.283 ***	−0.009	0.116 ***	1.000
Top1	−0.102 ***	0.004	−0.083 ***	0.140 ***	0.237 ***	0.102 ***	0.141 ***	0.276 ***	0.087 ***	0.034 ***	−0.077 ***	1.000
Balance	0.109 ***	0.005	0.106 ***	−0.052 ***	−0.035 ***	−0.085 ***	−0.029 ***	−0.213 ***	0.006	−0.016 **	0.048 ***	−0.589 ***
Big4	0.102 ***	−0.062 ***	0.001	0.124 ***	0.353 ***	0.109 ***	0.060 ***	0.128 ***	0.079 ***	0.043 ***	−0.051 ***	0.152 ***
INST	0.037 ***	−0.012 *	−0.045 ***	0.167 ***	0.411 ***	0.166 ***	0.173 ***	0.363 ***	0.128 ***	−0.048 ***	−0.170 ***	0.493 ***

* p < 0.1, ** p < 0.05, *** p < 0.01.

. It can be seen that the absolute values of the correlation coefficients between the main variables are all less than 0.5, demonstrating that the collinearity problem is not serious among the selected correlation indexes. As for the main explanatory variables, the correlation coefficient between economic policy uncertainty (EPU) and OverPay is −0.023 and significant, which indicates that rising economic policy uncertainty will inhibit executive excessive compensation, validating hypothesis H1a. The correlation coefficient between economic policy uncertainty (EPU) and UnPerks is 0.015 and significant, indicating that economic policy uncertainty will encourage C-suite excessive on-the-job consumption, validating hypothesis H1b. From the correlation between control variables and OverPay, corporate indebtedness, the shareholding ratio of the largest shareholder, and the number of independent directors are negatively correlated with OverPay. Meanwhile, the correlation coefficient between corporate operating cash flow and UnPerks is 0.108, which is significant. When the corporate operating cash flow rises, C-suite excessive on-the-job consumption will be boosted.

5.3. Collinearity Analysis

Table 6. Multiple collinearity analysis.

Variable	VIF	1/VIF
Top1	2.17	0.4607
Size	1.85	0.5407
INST	1.71	0.5851
Balance	1.70	0.5885
Lev	1.56	0.6415
ROA	1.46	0.6854
SOE	1.34	0.7460
IC	1.20	0.8332
Cashflow	1.19	0.8373
Big4	1.18	0.8484
EPU	1.11	0.8989
Dual	1.11	0.9020
Indep	1.03	0.9735

illustrates the results of the variance inflation factor (VIF) of related variables, which is used to test whether there is multicollinearity among explanatory variables selected in this model. As can be seen from Table 6, the VIF values of all relevant variables selected in this model are all less than 5, proving that no multicollinearity problem exists in this research model. Further analysis can be performed according to the design model.

5.4. Regression Analysis

Table 7. Analysis of principal regression results.

	Panel Regression 1		Panel Regression 2
	OverPay	UnPerks	Overpay	UnPerks
EPU	−0.172 *** (−6.526)	0.004 *** (2.805)	−0.101 *** (−6.054)	0.002 * (1.903)
Size	0.055 *** (3.735)	−0.003 *** (−3.668)	0.044 *** (4.768)	−0.002 *** (−5.610)
Lev	−0.164 *** (−3.469)	−0.003 (−1.349)	−0.359 *** (−7.290)	−0.004 ** (−2.022)
ROA	−1.155 *** (−15.890)	0.014 *** (4.043)	−0.460 *** (−4.441)	0.030 *** (7.224)
SOE	−0.030 (−0.874)	−0.002 (−1.328)	−0.114 *** (−5.430)	−0.002 ** (−2.217)
Cashflow	0.053 (1.025)	0.005 * (1.804)	0.333 *** (3.789)	0.029 *** (7.792)
Indep	−0.259 ** (−2.422)	0.004 (0.763)	−0.240 * (−1.833)	−0.009 * (−1.788)
Dual	0.038 *** (2.967)	−0.000 (−0.717)	0.061 *** (3.640)	−0.000 (−0.180)
Top1	−0.092 (−0.974)	0.008 * (1.862)	−0.555 *** (−7.144)	0.003 (1.100)
Balance	0.008 (0.234)	0.002 (0.882)	0.031 (0.906)	0.001 (0.458)
Big4	0.162 *** (2.991)	−0.006 ** (−2.339)	0.217 *** (4.989)	−0.004 *** (−2.646)
INST	0.102 *** (2.943)	0.001 (0.696)	0.268 *** (6.810)	0.002 (1.188)
_cons	−0.190 (−0.711)	0.038 *** (2.830)	−0.069 (−0.346)	0.039 *** (4.488)
Industry	——	——	Yes	Yes
Year	Yes	Yes	Yes	Yes
N	24,484	24,484	24,484	24,484
adj. R2	0.037	0.010	0.058	0.038
F	15.044	3.584	7.488	7.573

* p < 0.1, ** p < 0.05, *** p < 0.01.

illustrates the test results of Model (6) and (7). The fixed-effects panel regression model of control year and corporate individual (columns 2 and 3), as well as of control year and industry, are used in regression analysis, and corporate-level cluster-robust standard error is adopted to alleviate possible sequence-related problems (columns 4 and 5). The regression results demonstrate that the regression coefficient between economic policy uncertainty (EPU) and OverPay is always negative, which is significant at the level of 1%, indicating that there is a significant negative correlation between economic policy uncertainty and C-suite overpayment; that is, higher economic policy uncertainty will inhibit the C-suite OverPay policy. Therefore, Hypothesis H1a in this paper is supported. The regression coefficient between economic policy uncertainty (EPU) and UnPerks is always positive and significant, which indicates that there is a significant positive correlation between economic policy uncertainty and C-suite on-the-job consumption. Therefore, higher uncertainty of economic policy will encourage the implicit self-interest behavior of executives. Hypothesis H1b in this paper is thus supported.

The economic significance of the regression in Table 7 is presented as follows. For panel regression 1, OverPay increases 1422% when EPU increases by one standard deviation and UnPerks decreases 717% when EPU increases by one standard deviation. For panel regression 2, OverPay increases 835% when EPU increases by one standard deviation and UnPerks decreases 358% when EPU increases by one standard deviation.

The value for economic significance is very large, this may result from the following two reasons: (1) the values for OverPay and UnPerks are relatively small compared to the real amount of managers compensation and on-the-job consumption and (2) change in OverPay and UnPerks can lead to a reasonable economic significance change for the Pay and Perks.

At the level of the control variables, if CEO also serves as Chairman, he/she will be more incentivized to seek excessive compensation. The reason is that “tyranny of the minority” is more likely to occur when the two powers are concentrated, and C-suite will pursue self-interest by excessive compensation. There is a significant positive correlation between the cash flow generated by corporate business activities and on-the-job consumption of executives. When an enterprise generates larger amounts of cash flow, on-the-job consumption of executives will be “easier” to realize and thus be promoted. Moreover, the control variable Big4 is negatively correlated with the excess on-the-job consumption of executives, which is also significant. Therefore, when an enterprise chooses “Big 4” accounting firms to perform audits, the excessive consumption of executives will be constrained, which is consistent with the previous research conclusions.

5.5. Moderating Role of Internal Control

In order to test Hypothesis H2, this paper constructs Models (7) and (8) and adds the internal control index as a moderating variable to study the role of internal control on the relationship between economic policy uncertainty and executives’ self-interest behavior. As illustrated in

Table 8. Moderating effect regression.

	Panel Regression 1		Panel Regression 2
	OverPay	UnPerks	Overpay	UnPerks
EPU	−0.173 *** (−6.537)	0.004 *** (2.868)	−0.092 *** (−5.355)	0.002 *** (2.588)
IC	−0.025 (−1.120)	0.000 (0.309)	0.038 (1.039)	0.004 *** (2.750)
EPU*IC	−0.173 *** (−4.535)	−0.007 *** (−3.635)	−0.254 *** (−4.514)	−0.013 *** (−5.407)
Size	0.059 *** (4.016)	−0.003 *** (−3.481)	0.043 *** (4.633)	−0.002 *** (−5.973)
Lev	−0.168 *** (−3.571)	−0.003 (−1.410)	−0.355 *** (−7.205)	−0.004 * (−1.891)
ROA	−1.112 *** (−15.354)	0.015 *** (4.207)	−0.439 *** (−4.144)	0.030 *** (7.004)
SOE	−0.030 (−0.870)	−0.002 (−1.312)	−0.113 *** (−5.356)	−0.002 ** (−2.134)
Cashflow	0.046 (0.896)	0.005 * (1.693)	0.319 *** (3.629)	0.029 *** (7.636)
Indep	−0.255 ** (−2.387)	0.004 (0.791)	−0.242 * (−1.841)	−0.009 * (−1.814)
Dual	0.038 *** (2.978)	−0.000 (−0.708)	0.061 *** (3.642)	−0.000 (−0.198)
Top1	−0.092 (−0.985)	0.008 * (1.847)	−0.555 *** (−7.140)	0.003 (1.083)
Balance	0.005 (0.134)	0.001 (0.798)	0.030 (0.859)	0.001 (0.402)
Big4	0.163 *** (3.000)	−0.006 ** (−2.333)	0.217 *** (4.977)	−0.004 *** (−2.696)
INST	0.101 *** (2.936)	0.001 (0.657)	0.266 *** (6.771)	0.002 (1.129)
_cons	−0.115 (−0.377)	0.032 ** (1.983)	−0.337 (−1.051)	0.013 (1.025)
Industry	——	——	Yes	Yes
Year	Yes	Yes	Yes	Yes
N	24,484	24,484	24,484	24,484
adj. R2	0.039	0.011	0.060	0.040
F	14.769	4.014	7.833	8.070

* p < 0.1, ** p < 0.05, *** p < 0.01.

, if α3 and β3, the coefficients of the interaction term EPU*IC, are significantly negative, the effective internal control can strengthen the restraining effect of economic policy uncertainty on OverPay, which is conducive to suppressing executives’ explicit self-interest behavior. Meanwhile, effective internal control can also weaken the role of economic policy uncertainty in promoting C-suite on-the-job consumption, and help to restrain executives’ invisible self-interest behavior. To summarize, effective internal control is conducive to restraining the self-interest behavior of executives. Therefore, Hypotheses H2a and H2b are verified.

6. Robustness Test and Further Analysis

6.1. Robustness Test

In order to ensure the reliability of the research conclusion, we conducted the following robustness tests.

6.1.1. Changing the Measurement Method of Executives’ Excessive Compensation and On-The-Job Consumption

Referring to the practices of Luo et al. [43], etc., this paper adopts the total monetary compensation of the Top 3 directors, supervisors, and executives as the absolute compensation of corporate executives, then uses its logarithm as the executive compensation variable (Pay_r) and calculates the replacement variable (OverPay_r) of the executives through the model. Drawing from the practices of Luo Jinhui, etc., we adjust the management expenses through the main business income to obtain the on-the-job consumption (Perk_r) of executives. See

Table 9. Variables of executives’ self-interest behavior under different methods.

	OverPay_r	Overpay_r	Perks	Perks
EPU	−0.183 *** (−7.163)	−0.107 *** (−6.525)	0.033 *** (8.501)	0.009 *** (4.039)
Size	0.065 *** (4.617)	0.051 *** (5.465)	−0.008 *** (−3.629)	−0.004 *** (−4.183)
Lev	−0.166 *** (−3.667)	−0.357 *** (−7.195)	−0.039 *** (−5.193)	−0.064 *** (−11.693)
ROA	−1.129 *** (−16.419)	−0.439 *** (−4.288)	−0.051 *** (−4.222)	−0.070 *** (−5.774)
SOE	−0.065 * (−1.906)	−0.176 *** (−8.407)	−0.012 *** (−2.706)	−0.005 *** (−3.031)
Cashflow	0.064 (1.258)	0.362 *** (4.187)	−0.035 *** (−4.276)	0.007 (0.753)
Indep	−0.254 ** (−2.522)	−0.284 ** (−2.207)	0.012 (0.957)	0.015 (1.255)
Dual	−0.022 * (−1.796)	−0.012 (−0.719)	−0.001 (−0.844)	0.001 (0.799)
Top1	−0.091 (−1.011)	−0.586 *** (−7.572)	0.050 *** (4.452)	−0.006 (−0.778)
Balance	0.002 (0.056)	0.012 (0.359)	0.008 * (1.835)	0.003 (0.803)
Big4	0.156 *** (2.918)	0.202 *** (4.721)	−0.009 ** (−2.166)	−0.006 ** (−2.056)
INST	0.092 *** (2.769)	0.250 *** (6.449)	0.004 (0.934)	0.003 (0.811)
_cons	−0.320 (−1.247)	−0.097 (−0.492)	0.041 (1.119)	0.091 *** (3.966)
Industry	——	Yes	——	Yes
Year	Yes	Yes	Yes	Yes
N	24,484	24,484	24,484	24,484
adj. R2	0.038	0.067	0.139	0.191
F	15.888	8.310	62.835	49.331

* p < 0.05, ** p < 0.01, *** p < 0.001.

for the regression results. The test coefficients of the main variables in the regression results have not changed and are still significant, which demonstrates the robustness of the conclusion.

6.1.2. Changing the Measurement Method of Economic Policy Uncertainty Index

When the above regression analysis was carried out in this paper, the method of processing the economic policy uncertainty index was to take the logarithm of its average. In the robustness test, the index is divided by 100 to obtain EPU2, the replacement variable of the economic policy uncertainty index, and then a regression analysis is carried out once again. The regression results are illustrated in

Table 10. Variables of economic policy uncertainty under different methods.

	OverPay	UnPerks	Overpay	UnPerks
EPU2	−0.094 *** (−6.526)	0.002 *** (2.805)	−0.045 *** (−6.054)	0.001 * (1.903)
Size	0.055 *** (3.735)	−0.003 *** (−3.668)	0.044 *** (4.768)	−0.002 *** (−5.610)
Lev	−0.164 *** (−3.469)	−0.003 (−1.349)	−0.359 *** (−7.290)	−0.004 ** (−2.022)
ROA	−1.155 *** (−15.890)	0.014 *** (4.043)	−0.460 *** (−4.441)	0.030 *** (7.224)
SOE	−0.030 (−0.874)	−0.002 (−1.328)	−0.114 *** (−5.430)	−0.002 ** (−2.217)
Cashflow	0.053 (1.025)	0.005 * (1.804)	0.333 *** (3.789)	0.029 *** (7.792)
Indep	−0.259 ** (−2.422)	0.004 (0.763)	−0.240 * (−1.833)	−0.009 * (−1.788)
Dual	0.038 *** (2.967)	−0.000 (−0.717)	0.061 *** (3.640)	−0.000 (−0.180)
Top1	−0.092 (−0.974)	0.008 * (1.862)	−0.555 *** (−7.144)	0.003 (1.100)
Balance	0.008 (0.234)	0.002 (0.882)	0.031 (0.906)	0.001 (0.458)
Big4	0.162 *** (2.991)	−0.006 ** (−2.339)	0.217 *** (4.989)	−0.004 *** (−2.646)
INST	0.102 *** (2.943)	0.001 (0.696)	0.268 *** (6.810)	0.002 (1.188)
_cons	−0.894 *** (−2.861)	0.053 *** (3.458)	−0.495 ** (−2.426)	0.045 *** (5.667)
Industry	——	——	Yes	Yes
Year	Yes	Yes	Yes	Yes
N	24,484	24,484	24,484	24,484
adj. R2	0.037	0.010	0.058	0.038
F	15.044	3.584	7.488	7.573

* p < 0.05, ** p < 0.01, *** p < 0.001.

. As can be seen from Table 10, the uncertainty of economic policy has a negative correlation with OverPay and is still significant, while the uncertainty of economic policy has a positive correlation with UnPerks. The main test coefficient in the regression results has not changed, so the conclusion of this paper is robust.

6.1.3. Instrumental Variable Regression

In this paper, instrumental variables are used for regression to reduce the interference of endogeneity problem on the regression conclusion. drawing on the research of existing literature [45], the uncertainty of global economic policy will affect the uncertainty of China’s economic policy, which accords with the correlation characteristics of the instrumental variables. However, the uncertainty of global economic policy will not directly affect the choice of C-suite self-interest behavior in China, which accords with the exogenous characteristics of the instrumental variables. Therefore, this paper uses the global economic policy uncertainty index as an instrumental variable to alleviate the endogeneity problems that may exist in the original regression model. Its calculation method is the same as the measurement method of China’s economic policy uncertainty in the main regression, obtained by calculating the logarithm of the annual arithmetic average.

Table 11. Instrumental variables.

	OverPay	UnPerks	Overpay	UnPerks
GEPU	−0.262 *** (−6.526)	0.006 *** (2.805)	−0.127 *** (−6.054)	0.002 * (1.903)
Size	0.055 *** (3.735)	−0.003 *** (−3.668)	0.044 *** (4.768)	−0.002 *** (−5.610)
Lev	−0.164 *** (−3.469)	−0.003 (−1.349)	−0.359 *** (−7.290)	−0.004 ** (−2.022)
ROA	−1.155 *** (−15.890)	0.014 *** (4.043)	−0.460 *** (−4.441)	0.030 *** (7.224)
SOE	−0.030 (−0.874)	−0.002 (−1.328)	−0.114 *** (−5.430)	−0.002 ** (−2.217)
Cashflow	0.053 (1.025)	0.005 * (1.804)	0.333 *** (3.789)	0.029 *** (7.792)
Indep	−0.259 ** (−2.422)	0.004 (0.763)	−0.240 * (−1.833)	−0.009 * (−1.788)
Dual	0.038 *** (2.967)	−0.000 (−0.717)	0.061 *** (3.640)	−0.000 (−0.180)
Top1	−0.092 (−0.974)	0.008 * (1.862)	−0.555 *** (−7.144)	0.003 (1.100)
Balance	0.008 (0.234)	0.002 (0.882)	0.031 (0.906)	0.001 (0.458)
Big4	0.162 *** (2.991)	−0.006 ** (−2.339)	0.217 *** (4.989)	−0.004 *** (−2.646)
INST	0.102 *** (2.943)	0.001 (0.696)	0.268 *** (6.810)	0.002 (1.188)
_cons	0.257 (0.998)	0.028 ** (2.137)	0.066 (0.322)	0.037 *** (4.033)
Industry	——	——	Yes	Yes
Year	Yes	Yes	Yes	Yes
N	24,484	24,484	24,484	24,484
adj. R2	0.037	0.010	0.058	0.038
F	15.044	3.584	7.488	7.573

* p < 0.05, ** p < 0.01, *** p < 0.001.

illustrates the regression results for the instrumental variables. From the results of Table 11, we can find that the regression coefficients of economic policy uncertainty and executive excessive compensation are always negative and significant, while the regression coefficients of economic policy uncertainty and executive excessive on-the-job consumption are always positive and significant, which is consistent with the original regression results. Therefore, the model in this paper can still be used to draw consistent research conclusions after using instrumental variables to control endogeneity problems.

6.1.4. Adding Macro-Level Control Variables

The conclusions of this paper may be subject to endogeneity problems due to the lack of macro-level variables. To alleviate this issue and prevent endogenous errors caused by changes in macro-economic factors, this paper refers to the research of Li and Yang [46], Gulen [4], etc., and adds the year-on-year GDP growth variable to the control variables of the original regression model. The regression results are illustrated in the

Table 12. Adding macro-control variables.

	OverPay	UnPerks	Overpay	UnPerks
EPU	−0.188 *** (−6.472)	0.004 *** (2.907)	−0.149 *** (−5.835)	0.004 *** (3.157)
Size	0.055 *** (3.735)	−0.003 *** (−3.668)	0.044 *** (4.768)	−0.002 *** (−5.610)
Lev	−0.164 *** (−3.469)	−0.003 (−1.349)	−0.359 *** (−7.290)	−0.004 ** (−2.022)
ROA	−1.155 *** (−15.890)	0.014 *** (4.043)	−0.460 *** (−4.441)	0.030 *** (7.224)
SOE	−0.030 (−0.874)	−0.002 (−1.328)	−0.114 *** (−5.430)	−0.002 ** (−2.217)
Cashflow	0.053 (1.025)	0.005 * (1.804)	0.333 *** (3.789)	0.029 *** (7.792)
Indep	−0.259 ** (−2.422)	0.004 (0.763)	−0.240 * (−1.833)	−0.009 * (−1.788)
Dual	0.038 *** (2.967)	−0.000 (−0.717)	0.061 *** (3.640)	−0.000 (−0.180)
Top1	−0.092 (−0.974)	0.008 * (1.862)	−0.555 *** (−7.144)	0.003 (1.100)
Balance	0.008 (0.234)	0.002 (0.882)	0.031 (0.906)	0.001 (0.458)
Big4	0.162 *** (2.991)	−0.006 ** (−2.339)	0.217 *** (4.989)	−0.004 *** (−2.646)
INST	0.102 *** (2.943)	0.001 (0.696)	0.268 *** (6.810)	0.002 (1.188)
GDP	−0.055 *** (−5.303)	0.002 *** (3.597)	−0.039 *** (−4.075)	0.002 *** (4.522)
_cons	0.017 (0.067)	0.032 ** (2.402)	0.249 (1.178)	0.024 ** (2.460)
Industry	——	——	Yes	Yes
Year	Yes	Yes	Yes	Yes
N	24,484	24,484	24,484	24,484
adj. R2	0.037	0.010	0.058	0.038
F	15.044	3.584	7.488	7.573

* p < 0.05, ** p < 0.01, *** p < 0.001.

. After controlling for the macro-factors, the regression results are still stable, alleviating the endogenous factors caused by the lack of macro-level variables.

6.1.5. Controlling the Fixed Effects of Provinces

Since the uncertainty levels of different regions can be different, the influence of provincial factors can only be overlooked if controlling for the influence from industries. Therefore, the fixed effects of provinces were added to the robustness test, so as to reduce endogeneity problems. As can be seen from

Table 13. Controlling provinces as fixed effects.

	OverPay	UnPerks
EPU	−0.1067 *** (−6.5579)	0.0015 * (1.8796)
Size	0.0452 *** (5.1008)	−0.0021 *** (−5.7559)
Lev	−0.3052 *** (−6.6304)	−0.0029 (−1.4902)
ROA	−0.4247 *** (−4.2440)	0.0305 *** (7.3725)
SOE	−0.1095 *** (−5.2031)	−0.0016 ** (−2.0246)
Cashflow	0.3448 *** (4.1071)	0.0293 *** (7.9098)
Indep	−0.3589 *** (−2.8264)	−0.0102 ** (−2.0465)
Dual	0.0402 ** (2.4097)	−0.0005 (−0.8059)
Top1	−0.5854 *** (−7.8029)	0.0023 (0.7215)
Balance	0.0139 (0.4120)	0.0004 (0.3091)
Big4	0.1659 *** (3.8376)	−0.0045 *** (−3.1353)
INST	0.2727 *** (7.1380)	0.0018 (1.0879)
_cons	0.0875 (0.4450)	0.0453 *** (5.2376)
Industry	Yes	Yes
Province	Yes	Yes
Year	Yes	Yes
N	24,484	24,484
adj. R2	0.116	0.050

* p < 0.05, ** p < 0.01, *** p < 0.001.

, the regression coefficient between economic policy uncertainty and OverPay remains negative and significant at the level of 1%. The regression coefficient between economic policy uncertainty and executives’ excessive on-the-job consumption is still positive and significant, which is consistent with the main regression results.

6.2. Further Analysis

6.2.1. Heterogeneity Analysis of Enterprise Property Rights

Considering China’s political system and actual national conditions, enterprises with different property rights are divergent in business models, corporate organizational frameworks, etc. Therefore, a comparative analysis of enterprises with different property rights is an essential step in the study of corporate governance. In this paper, sample enterprises are divided into “SOE” and “non-SOE” groups according to the nature of property rights, and the relationship between the uncertainty of economic policy and excessive compensation and on-the-job consumption of executives is verified in groups. The verification results are illustrated in

Table 14. Heterogeneity analysis of property lefts of enterprises.

	OverPay		UnPerks
	SOEs	Non-SOEs	SOEs	Non-SOEs
EPU	−0.277 *** (−8.016)	−0.050 (−1.244)	0.004 * (1.741)	0.001 (0.665)
Size	0.037 (1.611)	0.049 *** (2.656)	−0.002 (−1.271)	−0.003 *** (−3.150)
Lev	−0.191 ** (−2.477)	−0.134 ** (−2.356)	−0.013 *** (−2.720)	0.003 (0.994)
ROA	−0.195 (−1.237)	−1.375 *** (−17.428)	0.027 *** (2.935)	0.010 *** (2.709)
Cashflow	−0.187 *** (−2.665)	0.155 ** (2.246)	−0.000 (−0.030)	0.007 ** (2.075)
Indep	−0.349 ** (−2.322)	−0.070 (−0.470)	0.004 (0.553)	0.001 (0.222)
Dual	0.033 (1.487)	0.040 *** (2.597)	−0.001 (−0.607)	0.000 (0.565)
Top1	−0.269 * (−1.848)	0.177 (1.472)	0.008 (1.114)	0.010 * (1.753)
Balance1	−0.051 (−0.923)	0.055 (1.261)	−0.000 (−0.137)	0.003 (1.296)
Big4	0.076 (1.207)	0.235 *** (2.736)	−0.002 (−0.657)	−0.006 ** (−2.012)
INST	0.093 ** (2.107)	0.132 *** (2.806)	0.002 (0.665)	0.001 (0.265)
_cons	0.818 ** (2.014)	−0.877 ** (−2.501)	0.019 (0.848)	0.047 *** (2.911)
Year	Yes	Yes	Yes	Yes
N	9766	14,718	9766	14,718
adj. R2	0.056	0.060	0.013	0.010
F	7.634	18.239	2.884	2.586

* p < 0.05, ** p < 0.01, *** p < 0.001.

. From the results of the grouping regression, we can see that the impact of economic policy uncertainty on executive OverPay is still negative and significant in SOEs, but negative and not significant in non-SOEs. The reason is that, compared with non-SOEs, SOEs feature stricter control over the compensation of executives, exerting a strong restraining effect. Meanwhile, uncertain economic policy has a more significant effect on the promotion of C-suite excessive on-the-job consumption in SOEs. When executives’ excessive compensation is restrained, they will actively make up for the shortfall through alternative methods, which is also in line with the hypotheses of this paper. Meanwhile, the nature of property rights of SOEs leads to multiple agency problems. The shareholders of SOEs do not actively play the role of supervisors, and thus a series of issues, including excessive on-the-job consumption of executives, have not been effectively addressed.

6.2.2. Heterogeneity Analysis of Institutional Investors

Different types of institutional investors play different roles in corporate governance. Therefore, this paper further analyzes the model by distinguishing the types of institutional investors. Drawing on the practices of Niu et al. [47], etc., this paper uses Formula 10 to study the heterogeneity of institutional investors through the dimensions of time and industry. The definitions of the variables used in Formula (10) are presented in

Table 15. Definition table of heterogeneity variables of institutional investors.

Variable	Definition
INVHit	Shareholding of institutional investors in Company i at Year t
STD(INVHit-3, INVHit-2, INVHit-1)	Standard deviation of institutional investors’ shareholding of Company i in the first three years
SDit	The ratio of shareholding of institutional investors in Company i at Year t to the standard deviation of institutional investors’ shareholding in the past three years
MEDIANtj (SDtj)	Median of Industry j in Year t
STABLEit	Dummy variables of institutional investor stability

.

$$\{\begin{array}{c}S{D}_{it}=\frac{INV{H}_{it}}{STD\left(INV{H}_{it-3},INV{H}_{it-2},INV{H}_{it-1}\right)}\\ STABL{E}_{it}=\{\begin{array}{c}1,S{D}_{it}\ge MEDIA{N}_{tj}\left(S{D}_{tj}\right)\\ 0,others\end{array}\end{array}$$

(10)

The results of the grouping regression for different types of institutional investors are illustrated in

Table 16. Heterogeneity grouping regression results of institutional investors.

	OverPay		UnPerks
	Stable	Transactional	Stable	Transactional
EPU	−0.204 *** (−5.790)	−0.128 *** (−3.248)	0.001 (0.578)	0.006 *** (2.992)
Size	0.045 ** (2.489)	0.061 *** (2.996)	−0.003 ** (−2.512)	−0.002 ** (−2.339)
Lev	−0.129 ** (−1.997)	−0.150 ** (−2.524)	−0.008 ** (−2.217)	0.002 (0.757)
ROA_m	−0.904 *** (−8.003)	−1.340 *** (−15.376)	0.017 *** (2.855)	0.012 *** (2.786)
SOE	−0.036 (−0.758)	0.006 (0.131)	−0.001 (−0.489)	−0.004 ** (−2.003)
Cashflow	−0.058 (−0.780)	0.088 (1.165)	0.007 (1.513)	0.003 (0.762)
Indep	−0.355 ** (−2.529)	−0.254 (−1.580)	0.002 (0.310)	0.005 (0.776)
Dual	0.032 * (1.781)	0.041 ** (2.379)	0.000 (0.084)	−0.001 (−1.411)
Top1	−0.072 (−0.583)	−0.035 (−0.268)	0.008 (1.210)	0.009 * (1.738)
Balance1	0.037 (0.800)	0.008 (0.177)	0.001 (0.573)	0.002 (1.020)
Big4	0.186 *** (2.978)	0.162 * (1.853)	−0.005 * (−1.880)	−0.006 (−1.562)
INST	0.095 * (1.740)	0.104 ** (2.351)	0.004 (1.250)	−0.001 (−0.639)
_cons	0.207 (0.640)	−0.597 (−1.556)	0.052 *** (2.761)	0.013 (0.734)
Year	Yes	Yes	Yes	Yes
N	12,415	12,069	12,415	12,069
adj. R2	0.030	0.056	0.013	0.008
F	6.363	13.201	2.987	1.774

* p < 0.1, ** p < 0.05, *** p < 0.01.

. The impact of economic policy uncertainty on OverPay is negative for both stable and transactional institutional investors, which is significant at the level of 1%. Yet the impact of economic policy on UnPerks is only significant for transactional institutional investors. The reason for this finding is that institutional investors of different natures are not uniformly enthusiastic about enterprise supervision. Due to long-term shareholding, stable institutional investors boast deeper understanding of the enterprise and are more active in corporate governance and supervision. Moreover, stable institutional investors highlight the long-term profits of enterprises, and thus their supervision of C-suite is more effective. Therefore, stable institutional investors wield a more prominent inhibitory effect on executives’ excessive on-the-job consumption.

6.2.3. Heterogeneity Analysis of Competitive Position of Enterprises

Market competition is an important bridge between macro-economy and corporate development. Drawing on the research of Yang and Yin [48], etc., this paper adopts the Lerner Index to measure the competitive position of enterprises, calculating the median of the Lerner Index by year and industry. If an enterprise’s competitive index exceeds the median, the value of “1” is assigned, representing a strong competitive position. The competitive index of enterprises below the median value is assigned as “0”, representing a weak competitive position. The results of the grouping regression according to competition positions are illustrated in

Table 17. Heterogeneity group results of competitive position of enterprises.

	OverPay		UnPerks
	Competitive	Non-Competitive	Competitive	Non-Competitive
EPU	−0.188 *** (−4.433)	−0.138 *** (−4.209)	−0.004 * (−1.917)	0.010 *** (5.331)
Size	0.046 ** (2.146)	0.044 ** (2.370)	−0.002 ** (−1.988)	−0.002 ** (−2.207)
Lev	−0.049 (−0.737)	−0.206 *** (−3.165)	0.000 (0.042)	−0.004 (−1.168)
ROA	−1.024 *** (−7.756)	−1.276 *** (−14.216)	0.036 *** (6.546)	0.011 ** (2.293)
SOE	0.014 (0.299)	−0.032 (−0.684)	−0.000 (−0.021)	−0.003 * (−1.798)
Cashflow	0.099 (1.201)	−0.111 * (−1.719)	0.001 (0.271)	0.005 (1.313)
Indep	−0.444 *** (−2.918)	−0.301 ** (−2.341)	0.007 (1.026)	−0.002 (−0.349)
Dual	0.029 * (1.673)	0.061 *** (3.297)	−0.000 (−0.460)	0.000 (0.188)
Top1	−0.086 (−0.648)	−0.002 (−0.019)	0.002 (0.400)	0.009 (1.396)
Balance	0.016 (0.331)	0.008 (0.178)	−0.001 (−0.373)	0.002 (0.777)
Big4	0.154 ** (2.093)	0.161 * (1.959)	−0.005 ** (−2.022)	−0.002 (−0.550)
INST	0.119 *** (2.782)	0.080 (1.595)	0.000 (0.136)	−0.001 (−0.253)
_cons	0.157 (0.423)	−0.220 (−0.617)	0.067 *** (3.625)	−0.001 (−0.057)
Year	Yes	Yes	Yes	Yes
N	12.297	12,187	12,297	12,187
adj. R2	0.030	0.046	0.032	0.015
F	5.630	11.984	5.612	3.379

* p < 0.1, ** p < 0.05, *** p < 0.01.

. The influence of economic policy on OverPay is negative in enterprises with both strong and weak competitive positions, which is significant at the level of 1%, indicating that the uncertainty of economic policy will inhibit executive excessive compensation. However, in the enterprises with a stronger competitive position, higher uncertainty of economic policy will restrain C-suite excessive on-the-job consumption. On the contrary, in the enterprises with a weaker competitive position, rising policy uncertainty will encourage such consumption. The conclusion is that market competition can play a certain governance role and alleviate the principal-agent problem [49]. Market competition provides opportunities for corporate performance comparison, which empowers investors and analysts with more motives and channels to obtain relevant information of enterprises, influences the decision-making of enterprise executives through various ways such as information effect and reputation mechanism, and restricts the speculation and self-interest behavior of executives in more competitive enterprises.

7. Conclusions and Discussion

At present, with the normalization of pandemic control in China and the turbulent international landscape, it is important to highlight the impact of economic policy uncertainty. In addition to the impact of the policy itself on industries and enterprises, the uncertainties caused by policy changes and execution will also exert an adverse effect on enterprises. When the prospects of an enterprise are unclear, it will implement contractionary policies through reduce investment, which is not conducive to the growth of the enterprise and market as well as employment expansion, exerting a negative impact on the overall macro-economy. Therefore, in the period of sound economic operation, the government should give full control to the self-regulation of the market and enhance the capability of enterprises to predict and avert risks. If we are determined to adjust the economic policy, government organizations should establish consultation platforms. Before new policies are rolled out, we should thoroughly investigate and analyze the market trends and corporate development and reassure the market with stable expectations. After the introduction of policies, we should also manage to drive away people’s concerns through policy interpretation. The policy should also be fully executed to avoid frequent and unnecessary changes, so as to ensure its long-term consistency.

Furthermore, facing the rising uncertainty of external macroeconomic policies, enterprises should improve risk identification and response mechanisms, stay alert to changes in economic policies, tap into their advantages, and improve the efficiency of resource allocation. Meanwhile, they should strengthen internal control and power restriction. The C-suite supervision should not only target explicit self-interest behavior, but also loopholes in the underlying implicit self-interest behavior, so as to reduce such behaviors, including excessive on-the-job consumption. Furthermore, when determining C-suite compensation, enterprises should comprehensively consider the traits of executives, including capability, career goals, expected income, etc., and be aware that executives will not obtain perks only from one single channel. When suppressing their compensation, executives will seek alternative benefits through on-the-job consumption, etc. While allowing executives to have bargaining power in compensation, enterprises should also anticipate and guard against alternative channels of self-interest.

Finally, through the further analysis of this paper, we can find that SOEs should focus on the promotion effect of economic policy uncertainty on executives’ implicit self-interest behavior. Stable institutional investors exert certain governance effects on C-suite self-interest behavior. Enterprises with unstable institutional investors should strengthen the quality of information disclosure, reduce information asymmetry, enhance internal control and supervision, and accept external monitoring from media and analysts, so as to better restrain the self-interest behavior of executives. Fair and orderly market competition can also deliver certain governance effects. The government should do its utmost to create a sound environment of market competition for the development of enterprises and give full control to the self-regulation of the market.