Does CEO Power Backﬁre? The Impact of CEO Power on Corporate Strategic Change

: In today’s dynamic economic environment, enterprises must maintain sensitivity and ﬂexibility when responding to the market through continuous strategic change. Anchored in the approach–inhibition theory of power, this study explores the relationship between CEO power and corporate strategic change and examines the moderating effects of company underperformance and product market competition. The study uses data from all A-share listed companies in China during 2006–2017. The results indicate that ﬁrst, there is an inverted U-shaped relationship between CEO power and corporate strategic change. Appropriate centralization of CEO power helps promote corporate strategic change, whereas excessive centralization hinders strategic change. Second, low underperformance strengthens the inverted U-shaped relationship between CEO power and strategic change. Finally, high product market competition strengthens the inverted U-shaped relationship between CEO power and strategic change.


Introduction
Power is the capacity of individuals to exert their will and influence over others [1]; it is derived primarily from control over valuable resources and the ability to punish and reward [2,3]. In management research, CEO power refers to the power that can override objections to influence key decision outcomes within the company [4]. This power can be obtained in formal or informal ways [1,5]. A higher level of CEO power is often associated with the greater possibility of management volatility and extreme corporate performance [4], more serious earnings management [6], and stronger risk-taking desire [7]. Within existing research, corporate strategic change has not received consistent attention, specifically in the exploration of how CEO power could affect this important outcome. Corporate strategic change is "the transformation of the company's configuration mode based on the current status of resources, which is reflected in the change of resource allocation in multiple strategic dimensions of the enterprise" [8,9]. Do powerful CEOs initiate more corporate strategic change because they take more risks and can access and allocate more resources? Or do CEOs bring about less strategic change because resistance to change within the organization grows with their power?
Most previous research has not directly and empirically examined the impact of CEO power on corporate strategic change. However, as a constant adjustment to corporate strategy has become a new normal for today's economy, there is a growing need to explore the relationship. Related studies on the connection between the CEO role and corporate strategic decision-making have provided informative, though inconsistent, insights [10]. Some scholars believe that excessive CEO power leads to risk preference in strategic decision-making [3]. Tang et al. (2011) asserted that corporate strategies under the guidance of a dominant CEO tend to deviate from industry norms, and the probability of performance extremes is high [11]. There is also evidence that power sometimes leads powerholders to make conservative decisions. Maner et al. (2007) found that CEO power may lead to more conservative strategic decisions, especially when power is highly unstable [12]. Despite these seemingly opposite effects, both views signify that CEO power has a tremendous impact on corporate strategic change [11].
In the current study, we argue that CEO power motivates CEOs to engage in more strategic change. However, increased CEO power also elicits resistance from other stakeholders who impact the change process, a result of the CEO's power disinhibition and maintenance. As the two forces push and pull, resistance to change may override initiatives to change at one point, but at another point, increased CEO power may affect strategic change negatively. We depict this relationship between CEO power and corporate strategic change as an inverted U shape based on the approach-inhibition theory of power [3]. According to this theory, power prompts an individual to pay attention to positive arguments for change and encourages a willingness to take risks, but power can also intensify tendencies toward disinhibition and power maintenance, as well as political behaviors [3,12]. CEO power disinhibition and involvement in political behaviors are likely to lead to stakeholder resistance and impede the implementation of strategic change.
Our study makes several contributions to ongoing research on CEO power and corporate strategic change. First, by examining the curvilinear nature of the relationship of that power and change, we provide a novel perspective that integrates individual psychological states with firm strategic implementations; our work explores the individual impact on firm-level activities. Investigating the dynamics between CEO power and strategic change, we focus on CEO power as an antecedent of corporate strategic change and employ the interactionist perspective to incorporate social environment resistance that results from growing CEO power. This interactionist approach helps advance understanding of the complex interplay between person and environment, and of how that interplay affects firm strategy. The curvilinear perspective also helps reconcile previous studies' inconsistent, even contradictory, results about the relationship between CEO power and corporate change.
Second, our study highlights the importance of the understudied psychological processes of key players (CEOs) affecting firm-level strategic outcomes and extends the approach-inhibition theory of power. Power, a force that influences the attitudes and behaviors of others, impacts CEO emotion, cognition, and behavior significantly [13,14]. We propose that the influence of power on CEO attitudes and behavioral outcomes is double-edged. Power can help powerholders focus on and achieve their goals and consolidate their power [7]. It can also prompt a series of negative behaviors due to power disinhibition [12], including self-serving behaviors and violations of norms; these negative behaviors lead to resistance to the powerholder's objectives by others and may even lead to loss of power. We apply and extend the approach-inhibition theory of power within strategic management research, adding a new theoretical framework to explain strategic change at firm level.
Third, our study identifies boundary conditions that moderate the relationship between CEO power and corporate strategic change. Specifically, in view of the psychological pressure caused by underperformance and product market competition possibly affecting the cognitive mechanism of powerful CEOs, we argue that these factors would moderate the curvilinear effects of CEO power on corporate strategic change. We believe that when an organization faces internal pressure, such as underperformance, the inverted U-shaped relationship between CEO power and corporate strategic change is strengthened. This moderating effect expands the internal mechanism of corporate strategic change. At the same time, external challenges faced by organizations, such as product market competition, also strengthen the inverted U-shaped relationship. Our investigation expands external contingent factor research on corporate strategic change.  Keltner et al. (2003) developed the approach-inhibition theory of power and suggested dual effects of power: approach and inhibition [3,15]. The approach effect refers to individuals engaging in behaviors that approach a goal when motivated by positive stimuli; the inhibition effect refers to individuals engaging in actions that aim to avoid risks when motivated by negative incentives. When an individual's power increases, their behavioral approach system (BAS) is activated for two reasons. First, the individual with power can obtain some social resources and material rewards easily and maybe be appreciated and respected by many others. Second, when the individual has more power, they will have more control of the pursuit of goals, be less likely to be obstructed by others, and more smoothly acquire information related to rewards. Access to resources and autonomy in goal pursuit encourages the individual with the power to behave proactively and seek changes [16]. Conversely, the reduction of power activates the individual's behavioral inhibition system (BIS). When individuals have relatively less power, they are subjected to more constraints, resulting in a lack of social resources and information. People with greater power tend to anticipate potential rewards, successes, and information and are more likely to demonstrate risky behavior [7]. People with less power tend to expect potential punishment, threats, failures, and less information; they focus on avoiding negative consequences [3].

Theoretical Background and Hypothesis
Power not only encourages individuals to focus on positive outcomes and take risks, but it also produces disinhibited behavior and power maintenance activities [13]. First, individuals with power may tend to act in ways that violate ethical norms and overlook the influence of their actions on others to pursue personal goals [3,[17][18][19]. Moreover, following enhanced freedom and increased disinhibition to act consistently to pursue personal goals, powerholders exhibit greater self-cognition consistency when interacting with their lowpower counterparts [12,20,21]. Second, the greater the power, the more likely powerholders engage in political activities to maintain their power, such as resolving potential threats in the environment and expanding control over resources [13]. This involvement is likely to increase conflict and escalate friction between powerholders and other stakeholders, both obstacles to powerholders' initiatives [22]. These two adverse effects may damage the interests of other stakeholders in the organization and pose a potential threat to their power and existing status. To cope with adverse situations, stakeholders may deliberately impede initiatives driven by a powerful CEO. Once resistance is greater than the CEO's change initiative, the change process will be impaired.

The Effect of CEO Power on Corporate Strategic Change
Corporate strategic change is the transformation of a company's configuration mode based on the current resource status. The transformation is reflected in changes to resource allocation in multiple strategic dimensions of the enterprise [8]. With a change in resourceallocation mode, the organization's structure and operations will be adjusted. Corporate strategic change usually involves a complicated decision-making process with multiple parties, including CEOs, board members, and other stakeholders. Power is a critical element of the process of formulating and implementing strategic decisions throughout the enterprise [23]. Powerful CEOs can prevail on board members or other stakeholders in their strategic decisions [24]. CEOs with increasing power become the initial drivers to trigger strategic change because they are resourceful and are willing to take risks. However, and at the same time, the self-serving tendencies of powerful CEOs to violate ethical norms and engage in political activities backfire and enflame resistance from other stakeholders, who are also involved with the implementation of a corporate strategic change process. In summary, both forces (from powerful CEOs and other influential stakeholders) interplay in the process of corporate strategic change.
According to the approach-inhibition theory of power, when CEO power increases from baseline to an intermediate level, it is likely to activate the behavioral approach system of CEOs for several reasons. First, a CEO with considerable power can perceive the potential significance and positive influence of corporate strategic change on the development of the organization, regardless of threats within their environments [3,12,14]. High power can reduce individuals' negative expectations of loss, thus reducing their aversion of failure [25]. When the CEO regards power as a CEO responsibility [26], the CEO can ponder the meaning of organizational change, gain insight into external opportunities, assume more responsibilities, seek rational change and the development of the organization, and promote the development of business strategy decisively. In this situation, the CEO's willingness to take risks will grow [7,27,28], helping them make long-term decisions from the perspective of business interest and organizational development [7].
Second, as CEO power increases, the CEOs have access to a wider range of resources and are more likely to receive assistance and resource support from various stakeholders in the organization [3]. The CEO's chance of realizing organizational change through resource allocation increases. Finally, when a CEO is more powerful, they control the company's major decisions [23] and enjoy confidence in the choices and decisions they make [29]. When other executives are deterred, these benefits aid the fight against external interference, questioning, opposition, or obstruction and help them to achieve their own goals [30]. Conversely, when a CEO's power is low, it is easier for them to perceive negative factors such as threats, risks, and pressures behind the corporate strategic change. These perceptions lead to insufficient determination to make an organizational change, and the CEO will become more vulnerable to other executives [31].
According to the approach-inhibition theory of power, increased power can direct an individual's attention to reward, opportunity, and success. When the power reaches an excessive level, however, it also leads powerholders becoming blindly optimistic [7,32], acting on their desires in unethical ways [18,33], and being more disinhibited in behavior [3,[34][35][36]. In this scenario, the powerholder is likely to sacrifice the interests of the group; engage in more self-serving, cheating, and opportunistic behaviors [19,37]; and even violate norms at all costs [38,39], paying less attention to the consequences of actions [12,40]. In fact, high levels of power help increase a CEO's insensitivity to social norms and disapproval [17,41,42], which likely backfires, attracting resistance from, even the rebellion of, other stakeholders in terms of the corporate strategic change advocated by powerful CEOs.
Excessive power also motivates CEOs to focus on strengthening and maintaining their advantageous position and to engage frequently in political activities that help maintain status and position [22,43]. To maintain power (reduce the loss of power), the CEO is likely to extend full control over information and resources, changing economic and social environments to consolidate or establish absolute control over firms, and even promote politics among top management teams (TMTs) [44][45][46].
In summary, as CEO power grows, CEO willingness and enthusiasm for promoting strategic change in the organization also grows. Simultaneously, resistance from other stakeholders may increase, as a powerful CEO may pursue personal goals and, possibly, violate ethical rules. Once CEO power exceeds a threshold, disinhibited tendencies lead to self-interested and opportunistic behavior, resulting in increased resistance to the CEO's initiatives for corporate strategic change. These factors inhibit the initiation and implementation of strategic change [13]. Based on this discussion, we propose Hypothesis 1:

Hypothesis 1 (H1):
There will be an inverted U-shaped relationship between CEO power and corporate strategic change where CEO power is positively related to the corporate strategic change to a point, after which it becomes negative.

The Moderating Role of Underperformance
Underperformance reflects the gap between current performance and past performance or average performance in the industry [47]. Underperformance has a direct effect on a CEO's salary and incentives; therefore, it produces performance pressure and psychological poverty [48,49], in turn affecting decision-making behaviors. We argue that underperformance strengthens the curvilinear relationship between CEO power and corporate strategic change so that both halves of the curve become steeper with increased underperformance.
When CEO power is within a low-to-moderate range, corporate underperformance will strengthen the positive relationship between CEO power and corporate strategic change. First, performance pressure resulting from underperformance will increase the CEO's willingness to take risks. Prospect theory, proposed by Kahneman and Tversky (1979), suggests that people are more willing to take risks facing losses and more inclined to avoid risks facing gains [50]. Thus, once the status quo falls short of expectations, CEOs will take more risky actions, and corporate strategic change will be encouraged under this scenario [51]. Second, performance stress caused by underperformance helps CEOs search for and locate problems [52]. Underperformance suggests that the existing business model cannot efficiently utilize internal and external resources to achieve performance goals. This imbalance strengthens the CEO's problem-solving intentions [53], stimulating the CEO's motivation for strategic change to facilitate long-term performance improvement. Therefore, we suggest that underperformance will enhance the effect of CEO power on corporate strategic change and reflects a steeper curve.
Meanwhile, underperformance also accelerates the increase in stakeholders' resistance to change. Underperformance may endanger stakeholders' financial benefits, leading to a sharp increase in resistance to internal change. Vested interests in the company are often sensitive to short-term performance. Once the performance is not in line with expectations, stakeholders quickly raise doubts about ongoing strategic change and further advocate against change. With the gradual transformation of performance pressure into individual psychological pressure, a powerful CEO will urgently establish a self-defense mechanism to safeguard their vested interests and try to protect the self-interest resulting from power disinhibition [54]. Besides, powerful CEOs tend to seek quick victories, to resolve the psychological distress caused by performance decline, via unethical behaviors; these attempts to earn quick victories may provoke additional objections from other stakeholders. Therefore, we argue the following Hypothesis 2: Hypothesis 2 (H2): Underperformance moderates the inverted U-shaped relationship between CEO power and corporate strategic change such that at a low level of underperformance, the inverted U-shaped relationship is steeper than at a high level of underperformance.

The Moderating Role of Product Market Competition
Product market competition, "the competition aimed at market share and sales return of different products," reflects external industry competition [55]. It has a significant impact on managerial incentives and psychological pressure as well. We argue that product market competition will also strengthen the curvilinear relationship between CEO power and corporate strategic change so that both halves of the curve become steep.
When CEO power is within a low-to-moderate range, product market competition can strengthen the positive relationship between CEO power and corporate strategic change. First, product market competition affects corporate strategic change decision-making by influencing a CEO's problem identification and positioning [10,56]. Intense product market competition often increases the predatory risks of companies in the market. These risks include loss of market share, loss of users, and tight cash flow, all of which reflect an imbalance between enterprise strategic management and the environment. Proposing market and competition strategies to adapt to the environment and to alleviate the pressure of market competition is a critical reaction. Second, pressure from product market competi-tion will further encourage the promotion and execution of strategic change by enhancing the CEO's willingness to take risks [57]. In fierce external product market competition, competitors from all sides often force enterprises to face operational crises such as shrinking market share, loss of users, profit bottlenecks, price wars, and cash-flow tension created by irrational subsidies, all of which may increase the probability of enterprise bankruptcy and liquidation and result in a strong atmosphere of crisis and psychological pressure [55,58]. These crises will draw close attention from management, shareholders, and creditors; performance-appraisal pressure and employment risk will increase, strengthening the CEO's sense of responsibility and risk preference. In this milieu, CEOs are more likely to think about and promote change from an organizational standpoint so that their companies can gain sustained advantages from product market competition.
However, when the CEO has too much power, product market competition will also strengthen the negative relationship between CEO power and corporate strategic change. First, the interests of the company's shareholders and creditors are affected by product market competition. Faced with external competitive pressure, these stakeholders may require management to reduce high-risk investment behavior, increase the negotiation cost of any change, reduce the resources needed for strategic change, and ultimately delay the progress of the company's strategic change. Second, enterprises always risk failure in a fiercely competitive environment; external pressure may reduce a powerful CEO's risk-taking ability by strengthening the CEO's restraint and defensive behavior. Market pressure and internal shareholder pressure will restrict CEO power, forcing the CEO to establish a self-defense mechanism, to safeguard the privilege of power disinhibition, and to protect vested interests.
Third, increased multi-source pressure from shareholders and other stakeholders may threaten the CEO's political power base, forcing the CEO to focus more on power maintenance than risk-taking [11]. CEOs will take actions through information detention and institutional control, leading to extreme fluctuations in performance [44,45]. Therefore, we propose Hypothesis 3:

Hypothesis 3 (H3):
Product market competition moderates the inverted U-shaped relationship between CEO power and corporate strategic change such that the inverted U-shaped relationship is steeper in firms with high product market competition than in firms with low product market competition.

Sample and Dataset
Data for this study were drawn from all of China's A-share companies; relevant data was collected from 2006−2017 annual reports disclosed by all listed companies in the CSMAR (China Stock Market and Accounting Research) database. In the process of the 40 years of Chinese reform and opening-up, Chinese enterprises have experienced rapid development, accompanied by different and rapid enterprise changes, which is inseparable from entrepreneurs and CEOs. Thus, we chose Chinese enterprises as our samples. The original sample was screened according to the following criteria: (1) listed companies that excluded B shares and H shares, (2) listed companies that excluded the financial industry and insurance industry, (3) listed companies that excluded samples of "ST" (special treatment) and "PT" (particular transfer), and (4) listed companies that excluded companies whose asset-liability ratio exceeded 100% (insolvency). Finally, the study formed unbalanced panel data with 18,075 firm-year observations from 3409 A-share listed companies over 12 years.

Model Specification
For hypothesis testing, we formed the following regression equations. For the investigation of two adjustment variables within the relationship between CEO power and corporate strategic change, we adopted group moderating regression analysis [59].
The following regression equations were examined: Change i,t represents the company's corporate strategy change i in the year t. Whether the company experienced strategic change and the strength and intensity is reflected in the fluctuation in the allocation level of essential resources within the enterprise. We used the volatility of strategic resource allocation in an annual interval to measure corporate strategic change, and we used a six-direction dimension to synthesize a comprehensive indicator of strategic change: (1) the ratio of R&D expenditure to sales revenue, (2) the ratio of advertising expenditure to sales revenue, (3) the ratio of non-productive expenditure to sales revenue, (4) the ratio of net fixed assets to total fixed assets, (5) the ratio of inventory to sales revenue, and (6) the financial leverage factor.
Based on the research results of Zhang and Rajagopalan (2010) [9], we first calculated the deviation of the six indicators within one year, then normalized the one-year fluctuation values. Finally, the results of the six indicators after standardization were averaged. The average value (change) reflects the level of corporate strategic change in the year [9]. The larger the value, the higher the intensity and degree of corporate strategic change.
CEOpower i,t-1 refers to the CEO power of company i in year t-1, and CEOpowert-12 is the square of CEO power. According to Hypothesis 1, if there is an inverted U-shaped relationship between CEO power and corporate strategic change, the coefficient β1 of CEOpower i,t-1 would be significantly positive, and the coefficient β2 of CEOpowert-12 would be significantly negative. In measuring CEO power, we drew on Finkelstein's (1992) power model that asserts that CEO power plays an important role in corporate strategic decision-making [1]. Finkelstein divided CEO power into four types: structural power, owner power, expert power, and reputation power; accordingly, we measured those four dimensions of CEO power. We also referenced Chikh and Filbien (2011) to further subdivide the four dimensions into eight virtual variable indicators [60]. The definitions and explanations of each indicator are detailed in Table 1. CEO power was divided into four dimensions and eight indicators. Each indicator had its limitations and different interpretative perspectives, and correlation among the eight indicators was weak. Therefore, a comprehensive indicator was needed to measure CEO power. We used the methods of Chikh and Filbien (2011) in this study to create a variable for these eight indicators, using equal weight mean value with a final value between [0,1].
The control variables of this paper mainly include (1) size of the enterprise: Generally speaking, the larger the enterprise scale, the stronger the resource integration capability and risk-taking ability. This paper used the natural logarithm of the total assets of the enterprise to measure the scale of the enterprise. (2) Total asset profit rate (ROA): This paper measured the return on assets using the ratio of total profit at the end of the year to total assets.  Table 2. To test Hypothesis 2, an internal moderating variable needed to be introduced: underperformance. We compared the firms' current performance with its historical performance, and defined the observations below the historical performance level as underperformance [61]. In this paper, we used the average performance of the three previous years minus the performance of current year to form the index. The higher the index, the bigger the gap between the current performance and the historical performance of the company. Moreover, the gap between reality and expectation is one of the important factors that influences the cognition, attitude, and behavior of managers [62,63]. Facing the effects of different levels of underperformance, CEOs of enterprises might have different risk preferences towards strategic change. Next, we sorted the historical underperformance of all companies in each year from low to high and divided observations into two intervals, low underperformance and high underperformance, using 50% as the standard, according to other scholars' methods [64,65]. Thus, the last 50% of the observations were defined as high underperformance. Based on relevant research, we used the total return on assets (ROA) to measure business performance.
To test Hypothesis 3, an external adjustment variable needed to be introduced: product market competition (COMP i,t-1 ). We used market concentration (CR4) to reflect the degree of fierceness of the product market competition [66]. CR4 refers to the sum of the market share of the first four major enterprises in the industry. The smaller the index, the larger the COMP i,t-1 (1-CR4) and the fiercer the external product market competition. As with underperformance, the product market competition degree of all industries in each year was sorted in order from low to high, and observed values were divided into two intervals, low product market competition and high product market competition, using the 50% standard.

Descriptive Analysis
Descriptive statistics of the primary variables in the model are shown in Table 2. The sample includes 3,409 listed companies with 18,075 firm-year observations covering the period 2006−2017. Considering the time-lagged effect of CEO power on corporate strategic change, we used a one-year lag approach [47,67].
Considering the time difference between CEO power and corporate strategic change, this paper used a one-year lag approach for the related variables [47,68]. Specific to each variable, the CEO power score ranged from 0 to 1, with a standard deviation (SD) of 0.186, indicating that the variance of CEO power in different companies is salient. The score of corporate strategic change ranged from −0.433 to 13.38 (mean = −0.002, SD = 0.361). The underperformance (HI) ranged from 0 to 0.272 (SD = 0.049), and product market competition (Comp) ranged from 0 to 0.875 (SD = 0.208), which indicates that there is a huge difference in product market competition intensity in different industries and enterprises are facing varied levels of intensity of product market competition. Table 3 shows the correlation coefficient matrix between the main variables. The results of Pearson correlation coefficient show that the correlation coefficients between the variables were below 0.6. On this basis, it can be preliminarily judged that there were no multiple collinearities among the variables.

Empirical Results
The empirical results were divided into three sections and six models, as shown in Table 4. The primary effect of the interrelationship of CEO power and corporate strategic change is shown in Model (2). The grouping moderating effect of underperformance on the relationship between CEO power and corporate strategic change is shown in Models (3) and (4). The grouping moderating effect of product market competition on the relationship between CEO power and corporate strategic change is shown in Models (5) and (6). Model (1) reports results using only the control variables.   The results indicate that the regression coefficient β1 of CEO power (CEO power) was positive, and the P-value was 0.007, significant at the 1% confidence level. The regression coefficient of CEO power squared (CEOpower2) was negative, with a P-value of 0.020. These results show an inverted U-shaped relationship between CEO power and corporate strategic change; Hypothesis 1 is supported. Figure 1 presents the inverted U-shaped relationship of the primary effect, demonstrating that there is an apparent threshold for CEO power. Before the threshold is reached, CEO power promotes corporate strategic change. Once the threshold is exceeded, excessive centralization of CEO power hinders corporate strategic change. As the Wald test has general sample deviation and the Suest test cannot be applied to panel data, we chose the Chow test to test the coefficient differences between groups for the adjustment variables [59,69]. The Chow test result is the p-value of the inter-group coefficient difference in Table 4. Models (3) and (4) focus on the moderating effect of historical underperformance (HI i,t-1 ). We divided the data into two sample groups: low underperformance and high underperformance. The results in Table 4 show that the Chow test resulted in a coefficient difference between the two groups with a P-value of 0.001. This result indicates significant differences between the inverted U-shaped relationship of CEO power and corporate strategic change in the two groups, high underperformance and low underperformance. The regression coefficients of the second and first terms changed from insignificant to significant, which shows that low underperformance strengthens the inverted U-shaped relationship. Therefore, Hypothesis 2 is also supported. The results show that under low underperformance, a CEO can reduce the imbalance between environment and organization through enterprise strategic change to reverse the decline in performance, and the inverted U-shaped relationship between CEO power and strategic change will be strengthened; however, if the performance deviation is too large (high underperformance), the CEO might get entangled with psychological poverty and a nervous attitude, and focus more on organizational survival in decision-making [62]. At this time, the CEO would prefer cost control and resource retention, rather than risk-taking change [49].
Models (5) and (6) focus on the moderating effect of product market competition (COMP i,t-1 ). The study divided the data into two sample groups: low product market competition and high product market competition. The results are shown in Table 4. The Chow test resulted in a p-value of 0.034, indicating that the inverted U-shaped relationship between CEO power and corporate strategic change demonstrated a significant difference between the two groups, low product market competition and high product market compe-tition. The regression coefficients of the first and second terms changed from insignificant to significant, indicating that high product market competition strengthens the relationship between CEO power and corporate strategic change. This is consistent with Hypothesis 3.

Robust Test
In order to increase the reliability and stability of the conclusions, we also conducted a robustness of regression by changing the measurement indicators of the dependent variables and changing the time range of the research samples. First, "Change" can also be measured by three-year fluctuation values of the six indicators given in Section 3.2. Based on the new indicator for Change, we conducted regression models once again for the three hypotheses, and the regression results in Table 5 still support these hypotheses. Finally, in order to eliminate the impact of the financial crisis in 2008, we reduced the sample time range to 2009−2017, and then regressed again. The regression results are shown in Table 6, and the results are still robust.    Notes: (1) *, **, and *** indicate significance at the 10%, 5%, and 1% levels, respectively. (2) The value below the regression coefficient is the standard error.

Discussion and Conclusions
In today's economy, the variability, uncertainty, complexity, and ambiguity of the market environment have become more prominent [70]. In such a highly uncertain environment, corporate strategic change has become one of the most critical themes of organizational development [71]. Based on a data sample of 3409 Chinese A-share listed companies during 2006-2017 and anchored in the approach-inhibition theory of power, this study tested the curvilinear effect of CEO power on corporate strategic change empirically and identified moderating effects of internal underperformance and external product market competition. Our findings suggest several conclusions. First, there is an inverted U-shaped relationship between CEO power and corporate strategic change. Second, low underperformance strengthens the inverted U-shaped relationship. Third, high product market competition also tends to intensify the inverted U-shaped relationship between CEO power and strategic change.

Theoretical Contributions
The study has several key theoretical contributions to management research. First, we enrich the theoretical and empirical studies of the approach-inhibition theory of power by exploring the curvilinear relationship between CEO power and enterprise strategic change. The current research concerning the approach-inhibition theory either emphasizes that elevated power promotes approach tendencies or suggests that impaired power activates inhibition behaviors [3], resulting in conflicts of conclusions. Moreover, the empirical studies based on this theory fail to explain the nonlinear relationship between power and behaviors of individuals or organizations [14,15]. We expand the approachinhibition theory of power and reconcile previously inconsistent findings by documenting the curvilinear relationship between CEO power and strategic change of firms. That is, with increased power, the CEO's behavior-approach system is activated effectively [3]. The CEO's willingness to risk strategic change is higher and the CEO's estimation of the positive role of strategic change is more optimistic. In other words, CEOs are inclined to pay more attention to potential benefits than to risks in this situation, and they are more confident and persistent in their personal abilities and opinions [32]. In addition, increased power often brings more resources and enhances the reputation, support, and decision-making voice of powerholders. This means that CEOs can achieve their strategic decision-making goals effectively, facing the pressure of opposition. Our study also shows that excessive CEO power inhibits the implementation of strategic change due to resistance from multiple stakeholders created by power maintenance and political behavior. Ultimately, the push and pull of CEO power and resistance from other stakeholders lead to an inverted Ushaped relationship between CEO power and corporate strategic change. This finding extends and integrates the "approach effect" and the "disinhibition effect" of the approachinhibition theory of power, reconciling previously alleged contradictory effects of power on the corporate change process.
Second, we shed light on the mechanism of power influencing the corporate change process by illuminating the psychological and behavioral mechanism of CEOs in strategic decision-making and further developing the approach-inhibition theory of power. Sociocognitive research indicates that power usually leads to pronounced changes in one's emotions, attitude, cognition, and behavior, especially for key players such as CEOs [14]. We hypothesize that the effects of power on the attitudes and behaviors of CEOs is doubleedged. Generally, appropriate power enables CEOs to focus on goals of strategic change and take more risks [7]. However, at an extreme point of power concentration, dominant power induces negative behaviors due to power disinhibition intendencies and power maintenance, as well as political activities [12,13]. This is likely to lead to resistance and obstacles from stakeholders. We shed light on the inverted U-shaped relationship between CEO power and strategic change outcomes to explain the complicated cognition of powerful CEOs and enrich the approach-inhibition theory of power by incorporating the cognitive and psychological view.
Third, we also attempt to investigate the boundary conditions of the inverted Ushaped relationship between CEO power and strategic change. As noted, corporate strategic change is always constrained by a variety of internal and external factors [11]. However, few previous studies have considered underperformance and product market competition as contingent variables. Our study shows that underperformance and product market competition tend to strengthen the curvilinear relationship between CEO power and corporate strategic change due to the psychological pressure imposed on CEOs' cognition [31,72,73]. The results give us further insights into the internal and external challenges faced by corporate strategic change.

Managerial Implications
Our research also provides practical implications for corporate shareholders, founders, and policymakers. First, to promote corporate strategic change, shareholders and founders must empower CEOs appropriately to carry out the strategic change. The degree of power centralization has an essential impact on a CEO's psychology, cognition, and behavior, and affects preferences surrounding strategic change [13,14]. Appropriate centralization of power helps to promote strategic change, but once CEO power exceeds a threshold, it may result in "the dislocation of the role of the professional manager". CEOs may begin to pursue comprehensive leadership and control, which likely hinders the process of corporate strategic change [68]. Therefore, CEO power must be allocated rationally and supervised to an optimal degree to facilitate strategic change. In addition, the psychological deviation caused by excessive power should also be carefully monitored to reduce the likelihood of the adverse effects created by the expansion of CEO power. We need to caution against dominant CEOs despite the existing advantages or benefits brought by power, unless there is evidence that some mechanisms can effectively reduce the negative behaviors of power disinhibition and political maintenance.
Second, CEOs facing internal pressure caused by underperformance should adjust psychological expectations and avoid the negative impact of excessive pressure. Companies should also be prepared for the negative moderating effect of underperformance. That is, under circumstances of performance decline, powerful CEOs may seek quick solutions to mitigate the psychological distress and protect financial outcomes, even via unethical behaviors [54]. These behaviors might further provoke objections from related stakeholders and thus impede the strategic change process.
Third, considering the improvement of internal efficiency brought on by product market competition [74], the organization itself should maintain a certain degree of external competition from outside. Awareness of the threat of excessive external pressure on a CEO's power and interests is necessary. That threat may force the CEO to abandon longterm strategy and disregard group interests. At the same time, the results also remind firms that indispensable measures to alleviate market pressure are needed to maintain the risk-taking capacity of CEOs. Otherwise, CEOs might seek to focus more on their position and power, instead of corporate strategic change.

Limitations and Directions for Future Research
Several limitations of this study deserve some attention. First, we tested our model based on secondhand archival data, using those data as agencies for CEO power. Although we adopted this methodology from previously validated studies, we believe future research should use multiple data sources, such as questionnaires, focus groups, or interviews, to calibrate our operationalization of the construct [1]. It is meaningful for us to integrate different sources of data to investigate the relationship between CEO power and corporate strategic change.
Second, our research pays close attention to CEO power instead of the exercise of power. To deeply understand the relationship between CEO power and corporate strategic change, we should also illuminate the mechanisms through which CEO power affects strategic change [11]. It might be a good way to comprehend a series of questions concerning both. For example, what roles would political tactics and skills play in mediating the relationship between CEO power and strategic change or other important organizational outcomes? How does power interact with some characteristics of CEOs, such as hubris, narcissism, and charisma?
Third, this study analyzed only the effects of CEO power on the willingness for and possibility of corporate strategic change; it did not involve a series of other key issues related to strategic change, such as ways to promote strategic change and speed or the direction of strategic change [10]. The relationship between CEO power and corporate strategic change deserves further in-depth future research. Furthermore, we also suggest that future research would benefit from examining the moderating variables between power and strategic change, such as the interaction between CEOs and the board, the instability of the power hierarchy, and the characteristics of CEOs.

Conclusions
There is an ongoing debate about whether CEOs with great power exert some positive or adverse effects on corporate values, such as strategic change [11,54]. For a rather long time, researchers have hesitated and struggled with contradictory views. This study provides empirical evidence of an inverted U-shaped relationship between CEO power and corporate strategic change in Chinese firms and offers insights into the psychological mechanism of the effects of power on the strategic process. This U-shaped relationship is further strengthened when a CEO is faced with the internal pressure of underperformance and the external pressure of product market competition. As the world becomes even more complicated, CEOs are more likely to promote strategic change through power centralization. There continues to be far-reaching significance to exploring the complicated process of power-driven strategic change.
Author Contributions: Y.Z. (Yu Zhou) was in charge of the work involving the topic selection and content modification of the paper-in particular, he guided us in terms of design and polishing from beginning to end; H.Z. was mainly responsible for data analysis, article writing, and modification; J.Y. was involved in the modification and polishing of our research; Y.Z. (Yunqing Zou) participated in the modification and submission of this paper. All authors have read and agreed to the published version of the manuscript.