The Effect of Corporate Ethical Level and Ethical Efforts on Corporate Performance: Evidence of a Corporate Moral Licensing Phenomenon

Abstract

This study investigates the complex relationship between corporate ethics and financial performance in the context of sustainable management. By subdividing corporate social responsibility (CSR) into corporate ethical levels and ethical efforts, it analyzes how these factors, along with their interaction, influence corporate financial performance. The analysis employs ordinary least squares (OLS) regression with year fixed effects (year dummy variables), using twelve years (2012–2023) of ESG evaluation data from 384 publicly listed Korean manufacturing companies. The study empirically tests whether highly ethical firms may fall into a corporate moral licensing phenomenon, whereby past ethical achievements are used to justify socially irresponsible decisions in pursuit of short-term profit. The findings reveal that while higher corporate ethical levels generally improve financial performance, sustained ethical efforts at a high level can paradoxically reduce short-term profitability, thereby exposing firms to risks that undermine sustainability. Taken together, these results highlight the dual nature of CSR as both an enabler and a potential constraint for sustainable business practices. Overall, the study contributes to sustainability research by offering novel evidence of moral licensing at the corporate level and providing actionable insights for managers and policymakers seeking to balance ethical commitments with long-term sustainable performance.

1. Introduction

Amid the emphasis on corporate social responsibility (CSR), the fundamentals of stakeholder theory have been widely adopted in business and academia, in which CSR activities have a net positive effect across all stakeholders, including maximizing shareholder value. For example, business round table (BRT), a group of the 200 largest corporations in the United States, replaced the phrase “maximizing shareholder value” as a corporate goal in August 2019. Instead, BRT presented five new goals—“Value to customers,” “Invest in employees,” “Treat suppliers fairly and ethically,” “Support local communities,” and “Create long-term value for shareholders” [1]. Although the focus is still on shareholder value creation, the BRT officially declaring that various stakeholders should be considered indicates that the fundamentals of stakeholder theory are being accepted in business practice as well as academia. Further evidence of widespread acceptance of this view can be observed in Larry Fink, CEO of BlackRock, advocating for other CEOs to adopt stakeholder capitalism in his 2022 Annual Letter to CEOs, in which he argues that serving the interests of all stakeholders is essential to the long-term success of any business [2].

While CSR is accepted as an obvious and important component of sustainable management, paradoxically, there is a possibility that overemphasizing it can lead to a moral licensing phenomenon. Moral licensing theory has garnered much attention in the field of psychology since Monin and Miller [3] demonstrated an individual’s inconsistent moral attitude. Subsequent studies have shown that moral licensing can also manifest among leaders and organizations. In particular, Ormiston and Wong [4] found that CEOs with strong moral identities were paradoxically more likely to engage their firms in socially irresponsible behaviors, indicating that moral licensing mechanisms can operate at the executive and corporate level. Similarly, Effron and Conway [5] emphasized that past moral actions can psychologically justify later unethical conduct, even within organizational settings. These findings collectively suggest that moral licensing is not confined to individual cognition but can also influence managerial and organizational decision-making. Building on these insights, the present study extends the scope of moral licensing theory from individual and leadership levels to the corporate level, exploring how a firm’s accumulated ethical standing and ongoing ethical efforts interact to shape corporate behavior and performance.

Moral licensing is a phenomenon in which individuals justify subsequent immoral behavior based on prior moral acts [6]. It reflects a psychological mechanism whereby past ethical conduct creates a perceived license to engage in subsequent unethical or irresponsible actions while maintaining a sense of moral self-approval, a dynamic that can extend beyond individuals to organizations. Applying this concept to the corporate context, it is conceivable that decisions made by management could be influenced by a similar phenomenon in which firms that have achieved strong CSR performance in the past might neglect CSR for short-term profits and management may justify their decisions based on past CSR behaviors. This is an important issue that deserves more academic attention, since as to the best knowledge of the authors, there is no extant published research that has demonstrated such a phenomenon. Furthermore, the negative ripple effects of an incident or accident caused by unethical behavior by a company are much more far-reaching and powerful than those of an individual, so it is an important topic to shed light on.

This study aims to empirically demonstrate that the phenomenon of moral licensing, previously discussed at the individual level, can also manifest at the corporate or organizational level. Firstly, drawing on theories of corporate social responsibility and irresponsibility, the study defines corporate ethical level and corporate ethical efforts. Subsequently, hypotheses are formulated concerning the impact of corporate social responsibility on business performance, considering both positive and negative perspectives. Specifically, the corporate ethical level at time t − 1 and the corporate ethical efforts during the period from time t − 1 to time t (hereafter denoted as period_t−1,t), influence business performance during the specified period. Additionally, the study proposes hypotheses concerning the interaction between corporate ethical level at time t − 1 and corporate ethical efforts during period_t−1,t from the standpoint of moral licensing theory, and their collective impact on business performance during the period.

To empirically test the proposed hypotheses, this study distinguishes between cumulative CSR levels, dynamic ethical efforts, and their interaction as independent variables, and examines their combined impact on corporate performance. The analysis employs ordinary least squares (OLS) regression with year fixed effects (year dummy variables), using ESG rating data from the Korea Corporate Governance Service (KCGS) for 384 publicly listed Korean manufacturing companies over the twelve-year period from 2012 to 2023, along with corresponding financial data.

Compared with prior studies on the relationship between corporate social responsibility (CSR) and corporate financial performance (CFP), this research offers several contributions. First, it conceptually distinguishes between the static (corporate ethical level) and dynamic (corporate ethical efforts) dimensions of corporate ethics, addressing the limitations of prior research that treated CSR as a single construct. Second, it empirically examines whether moral licensing can emerge at the corporate level, extending the moral licensing theory from individuals to organizations. Third, by employing longitudinal ESG data from 2012 to 2023, this study provides robust evidence from an Asian context, thereby broadening the geographical and temporal scope of CSR research. Finally, the findings contribute to sustainable management practice by offering managerial and policy implications for balancing ethical behavior with financial performance.

2. Theory and Hypotheses

2.1. Corporate Social Responsibility and Corporate Social Irresponsibility

The concept of corporate social responsibility (CSR) is widely regarded as a core element of sustainable management. Since its early discussion in the 1930s, research on CSR has evolved from a normative question of what businesses ought to do for society to empirical investigations of how socially responsible behavior affects corporate performance. Bowen [7] defined CSR as the duty of businesspeople to pursue policies and make decisions desirable in terms of society’s values. In contrast, Friedman [8] viewed CSR narrowly as engaging in profit-oriented activities within legal and social norms. Over time, however, scholars have reached a broader consensus that corporations bear not only economic and legal responsibilities but also ethical and discretionary ones [9,10]. Although definitions vary, most agree that CSR represents a firm’s voluntary and proactive commitment to create social value while complying with laws and ethical standards.

In contrast, corporate social irresponsibility (CSiR) refers to actions or omissions that harm stakeholders or society. Armstrong [11] characterized CSiR as decision-making that prioritizes private interests at the expense of societal welfare. Lange and Washburn [12] described it as organizational behavior causing social loss by acting irresponsibly in opposition to CSR, and Zhang and Zhang [13] distinguished between doing harm and failing to do good—the latter implying that a firm may not violate laws yet still fall short of societal expectations. Scholars continue to debate whether CSR and CSiR exist on a single ethical continuum or as independent constructs [4,14,15,16,17,18,19]. The present study explicitly adopts the single-dimension perspective, in which CSR and CSiR represent opposite ends of the same continuum of corporate ethical behavior. This stance aligns with the “not doing the right thing” interpretation of CSiR, emphasizing omission or negligence rather than overt illegality.

Building on this view, the next key step is to specify how corporate ethical behavior can be operationalized in measurable terms. To this end, this study differentiates between a firm’s static ethical standing and dynamic ethical efforts, which are further conceptualized in the following section.

2.2. Corporate Ethical Level, Ethical Efforts, and Corporate Social Responsibility

Following this conceptual foundation, this study builds upon stakeholder and agency theories to define two complementary dimensions of CSR—Corporate Ethical Level (CEL) and Corporate Ethical Efforts (CEE)—representing its static and dynamic aspects, respectively.

Corporate Ethical Level (CEL) refers to the cumulative ethical standing of a firm at a given point in time, reflecting its long-term compliance with environmental, social, and governance standards. It represents a firm’s stock of ethical capital accumulated through consistent socially responsible practices. From the perspective of stakeholder theory, a high CEL signals to stakeholders that the firm is trustworthy, credible, and aligned with broader societal interests, thereby enhancing its access to resources, reputation, and ultimately its financial performance.

In contrast, Corporate Ethical Efforts (CEE) capture the change-oriented dimension of corporate ethics—the extent to which a firm strengthens or weakens its ethical practices over a specific period. Unlike CEL, which is static, CEE reflects managerial commitment and behavioral adjustment in response to new stakeholder demands, institutional pressures, or internal strategic choices. Drawing on agency theory, however, such efforts may entail additional costs, resource reallocations, and short-term inefficiencies that could temporarily depress financial performance.

Accordingly, CSR can be viewed as an umbrella construct encompassing both CEL (the level of ethical capital) and CEE (the change in ethical behavior). CEL corresponds to the “stock” of accumulated ethical value, while CEE represents the “flow” of ongoing ethical engagement. This dual conceptualization—summarized as CSR = f(CEL, CEE)—clarifies that the two dimensions may have distinct or even opposing effects on corporate outcomes.

2.3. Theoretical Linkages to Corporate Financial Performance (CFP)

From an instrumental stakeholder theory perspective, higher Corporate Ethical Levels (CELs) enhance performance by strengthening legitimacy and cooperative stakeholder relationships, leading to cost reductions and improved competitiveness. This logic leads to Hypothesis 1, predicting a positive relationship between CEL and corporate financial performance (CFP).

Hypothesis 1:

Higher corporate ethical level at time t − 1 positively influences corporate performance during period_t−1,t.

However, based on agency and resource-allocation logic, ethical efforts (CEE) require new investments and managerial attention, which may divert resources from short-term profit-maximizing activities. Consequently, while CEL contributes positively to long-term performance, CEE can negatively affect short-term profitability due to the costs of ethical transformation. This reasoning supports Hypothesis 2, predicting a negative association between CEE and CFP within the same period.

Hypothesis 2:

Strengthening ethical efforts during period_t−1,t negatively influences corporate performance during period_t−1,t.

Finally, to integrate the above theoretical perspectives, this study draws on moral licensing theory to explain how the joint effect of CEL and CEE may vary depending on a firm’s prior ethical level. Specifically, the interaction between a company’s existing ethical standing and its ongoing ethical efforts is expected to shape financial outcomes differently, as firms at higher ethical levels may experience different behavioral motivations than those at lower levels. This logic leads to Hypothesis 3, which posits that the effect of ethical efforts on financial performance is moderated by a firm’s prior ethical level.

2.4. Moral Licensing and Moral Cleansing

Theoretical support for Hypothesis 3 is provided by the moral licensing and moral cleansing framework, which explains how prior moral behaviors shape subsequent ethical or unethical decisions at both individual and organizational levels.

The moral licensing theory has garnered much attention in the field of psychology since Monin and Miller [3] demonstrated an individual’s inconsistent moral attitude through experiments. Effron and Conway [5] further elaborated that moral self-licensing operates through psychological mechanisms of self-justification. Prior moral conduct weakens self-reproach and provides a perceived allowance for later unethical behavior. Previous research also suggests that moral licensing can extend beyond individual cognition to the organizational level, as demonstrated by Ormiston and Wong [4], who found that firms led by morally identified CEOs were more prone to engage in socially irresponsible actions. Moral licensing, therefore, refers to a phenomenon in which, when faced with committing an immoral act, one justifies this act based on prior moral actions [6]. The opposite concept of moral licensing is moral cleansing, which describes the tendency to perform good deeds to restore the moral balance that has been damaged by previous immoral actions [20].

Both moral licensing and moral cleansing operate within a broader moral balancing mechanism, through which individuals—and by extension, organizations—seek to maintain moral equilibrium. In this process, prior good deeds can justify later unethical acts (licensing), while past misconduct can trigger compensatory ethical behavior (cleansing) [21].

Borrowing the essence of this moral-balance mechanism from individual psychology, this study extends it to the corporate context. Attitudes toward corporate ethics can vary depending on a firm’s ethical level. Firms with a high ethical level may display a moral-licensing tendency, relaxing further ethical efforts or justifying questionable decisions based on their accumulated moral record. Conversely, firms with a low ethical level may exhibit a moral-cleansing tendency, intensifying ethical efforts to rebuild legitimacy and trust. Taken together, these theoretical perspectives provide a foundation for examining how a firm’s prior ethical standing conditions the financial consequences of its ongoing ethical efforts. These dynamics suggest that variations in corporate ethics may generate distinct financial outcomes, leading to the following hypotheses.

Hypothesis 3:

The effect of corporate ethical efforts during period_t−1,t on corporate performance is moderated by the firm’s prior ethical level.

Hypothesis 3a:

A higher corporate ethical level at time t − 1 leads to a decline in ethical efforts due to the moral licensing phenomenon, which positively affects corporate performance during period_t−1,t.

Hypothesis 3b:

A lower corporate ethical level at time t − 1 leads to an increase in ethical efforts due to the moral cleansing phenomenon, which negatively affects corporate performance during period_t−1,t.

3. Methods

3.1. Sample and Data

In this study, the ESG rating data and financial performance data of Korean listed companies are used for analysis. Many prior studies utilize US-based KLD data; however, this study uses data from Korean listed companies to extend the generalizability of previous findings and to identify country-specific differences in ethical and sustainability practices. A sample of companies listed on the Korean stock market over a consecutive twelve-year period is used, which allows for a robust sample size over an adequate period of time. However, proper caution is recommended when generalizing the results, as the nature of the sample generally tends toward larger and more mature companies.

First, the data from the Korea Corporate Governance Service (KCGS) is used for the ESG evaluation rating data of listed companies in Korea. KCGS has evaluated the performance of Korean listed companies in three areas annually since 2012: environmental management (E), socially responsible management (S), and governance (G). The KCGS announces individual ratings for each area and the integrated ESG rating annually. Through this, it contributes to establishing Korea’s ESG-related system by encouraging companies to faithfully carry out their responsibilities to stakeholders and by providing stakeholders with information for decision-making.

As shown in

Table 1. Total number of companies published in ESG integrated ratings.

Year	2012	2013	2014	2015	2016	2017	2018	2019	2020	2021	2022	2023
Number of corporations	797	867	881	829	868	852	881	875	908	950	974	987

, the number of companies with published ESG-integrated ratings from KCGS has steadily increased, from 797 in 2012 to 987 in 2023. This study analyses 384 manufacturing companies with integrated ratings that span across a consecutive twelve-year period without missing values. Given that ESG pressures and evaluation criteria vary significantly across industries, restricting the sample to the manufacturing sector enhances internal validity by mitigating heterogeneity across sectors. Manufacturing firms are also the most representative and comparable group in the KCGS dataset, ensuring that sectoral differences do not distort the results. Thus, the focus on manufacturing minimizes heterogeneity rather than introducing bias.

Since the analysis measures the change in evaluation grade from time point t − 1 to time point t, companies with missing values during any year are excluded because missingness may introduce period-specific bias. Therefore, the analysis suitably estimates the effects of ratings via eleven repeated measures for each company across the twelve-year period.

Specifically, the financial performance data of the sampled 384 companies is analyzed through the DataGuide system provided by FnGuide in South Korea. Due to the characteristic of measuring the change between time points t − 1 and t, two years are required per observation, thereby creating a total of eleven periods (e.g., period_2012,2013, period_2013,2014, ···, period_2022,2023). Therefore, a final sample of 384 companies represented by 4224 data points is used for analysis.

3.2. Operational Definition and Measurement of Variables

Corporate ethical levels (CEL) are obtained from integrated ESG ratings provided by the Korea Corporate Governance Service (KCGS) and are calculated via an absolute evaluation method. More specifically, the ESG-integrated rating of KCGS is calculated by summing the scores for each ESG area by adding points for CSR activities and deducting points for CSiR incidents, although the exact point system is not publicly disclosed. Therefore, only the categorical grades for each ESG category and the total ESG-integrated ratings are available to the public. The ESG-integrated grades are used to represent the corporate ethical level. Based on the meaning and definition of the KCGS ratings, as shown in

Table 2. Defining and measuring the corporate ethical level.

Integrated ESG Grade	S	A+	A	B+	B	C	D
Grade definition	Excellent	Very good	Good	Fair	Normal	Bad	Very bad
Corporate ethical levelt−1	High			Mid		Low

, the corporate ethical level at time t − 1 is classified as High, Mid, or Low.

The corporate ethical efforts (CEE) are derived from the change in the integrated ESG ratings between t − 1 and t. A company’s ethical efforts are measured as CSR, CSiR, or Complacent depending on the direction of ESG rating change. Figure 1 presents the measurement standards used for defining corporate ethical efforts. Corporate ethical efforts at time t are considered “CSR” if the ESG-integrated rating of the individual company at time t has increased compared to time t − 1. The efforts are considered “CSiR” if the rating has decreased compared to time t − 1 and are “Complacent” if the rating at time t is the same as time t − 1. Conceptually, if the evaluation grade of an individual company at time t has risen compared to the previously reported time point t − 1, it can be seen as having a relatively strong CSR propensity. Conversely, if the evaluation grade of an individual company at time t has decreased compared to time t − 1, it can be judged that the CSiR propensity has become stronger. Therefore, the corporate ethical efforts (CEE) are considered as “CSR” when the total ESG evaluation grade is upgraded from time t − 1 to time t, “CSiR” when the total ESG evaluation grade is downgraded from time t − 1 to time t, and “Complacent” when the total ESG grade is maintained between time t − 1 and time t. Defining corporate ethical efforts in this manner is summarized in Figure 1.

Due to autocorrelation between a company’s ESG evaluation grade at time t and t − 1, the company’s ethical efforts are measured dynamically over period_t−1,t to reflect relative change. Following Ormiston and Wong [4], who used prior CSR and CSiR ratings as lagged controls in modeling subsequent ratings, this study also considers the previous CSR level to control for temporal dependence in ESG performance.

When measuring CEE, the CEL may affect the impact of CEE on CFP. For example, if a company’s ESG rating rises from D at time t − 1 to C at time t and from A at time t − 1 to A+ at time t, the company’s ethical efforts are defined as CSR in both cases, and the question arises as to whether these cases should be viewed as equivalent. Considering this issue, in this study, the absolute meaning of the ESG evaluation rating at the time point t − 1 was measured as the corporate ethical level; therefore, this difference in corporate ethical level is controlled for. Since the impact on CFP of the company’s ethical efforts at period_t−1,t may vary depending on the company’s ethical level at time t − 1, a 3x3 interaction variable is constructed, as shown in

Table 3. Measurement of interaction variables between the corporate ethical level and ethical efforts.

		Corporate Ethical Levelt−1
		High	Mid	Low
Corporate ethical effortst	CSR	CSR-High	CSR-Mid	CSR-Low
	Complacent	Complacent-High	Complacent-Mid	Complacent-Low
	CSiR	CSiR-High	CSiR-Mid	CSiR-Low

, to analyze the interaction effect.

The operating profit over period_t−1,t is represented in financial statements at time point t, and is used as a dependent variable to measure CFP. Although various indicators, such as Tobin’s Q and return on assets (ROA), are used to measure CFP, operating profit for period t is the most direct measure of the operational performance of a company’s own business activities during that period. Tobin’s Q is a corporate value that reflects market expectations and is often used as an index to measure corporate performance and value from a future-oriented perspective. However, the influence of external stakeholders’ expectations confound the measurement of corporate performance for period_t−1,t when using Tobin’s Q. Alternatively, ROA is used as an accounting performance indicator that considers the size of total assets in net profit for period_t−1,t, but the net profit for period_t−1,t also reflects the profit and loss from the company’s non-operating activities. Thus, external factors are also involved in measuring the original operating performance when using ROA. Therefore, this study selected operating profit for period_t−1,t as a dependent variable to directly measure and analyze corporate operational performance.

To control for external and firm-specific influences on corporate financial performance, several control variables are incorporated into the model design. First, year dummies are included for each of the eleven periods from 2012 to 2023 to absorb macroeconomic and regulatory shocks that affect all firms uniformly. This approach ensures that temporal fluctuations—such as policy reforms, global market volatility, or ESG disclosure mandates—are properly controlled without introducing unnecessary complexity into the model.

Second, the lagged operating profit (Operating Profit_{_t−1}) is incorporated to account for performance persistence and mitigate simultaneity bias between ethical variables and current performance outcomes. By aligning temporal order—CEL measured at _t−1, CEE defined over _t−1,t, and CFP measured at t—the analysis ensures that ethical variables precede financial outcomes, thereby reducing endogeneity concerns.

Third, firm size is controlled using a binary dummy variable that distinguishes between large firms (1) and small or medium-sized enterprises (0). While the inclusion of the lagged dependent variable (operating profit at time t − 1) already accounts for much of the persistent firm-level financial heterogeneity—such as leverage, liquidity, or R&D intensity—the size dummy is retained to capture structural characteristics unique to the Korean economy, where large conglomerates (chaebols) and smaller firms often differ in their governance systems, stakeholder relations, and resource availability. This allows the analysis to more accurately reflect the dual structure of Korean corporate behavior.

Regarding the estimation method, an OLS regression with year dummies (i.e., year fixed effects) is employed rather than a firm fixed-effects model. The theoretical focus of this study lies in identifying how changes in corporate ethical efforts (CEE) interact with prior ethical levels (CEL) to influence financial performance. Applying firm fixed effects would eliminate between-firm variation in CEL—an essential component of this interaction—thus obscuring the primary relationship of interest. The chosen specification preserves both the interpretability of interaction terms and the ability to control for common time shocks across firms.

Equations (1), (2) and (3) list the regression model equations with operating profit of company i at time t as the dependent variable. Equation (1) is a model consisting of only dependent and control variables without independent variables, to serve as the controlled base model for comparing the significance of Equations (2) and (3), which are composed of the main variables of interest. Size represents a binary variable reflective whether the company is large (1) or an SME (0). The eleven periods are controlled for using one-hot encoding for time t of each period denoted as Year_j. Both CEE and CEL are included as categorical variables with CSiR and Low as their respective reference groups.

$${OP}_{it}={\beta }_0+{\beta }_1{OP}_{i\left(t-1\right)}+{\gamma }_1{Size}_{it}+{\sum }_{j=2}^{12}{\gamma }_j{Year}_{jit}+{\epsilon }_{it}$$

(1)

$$\begin{gathered} {OP}_{it}=(1)+{\gamma }_{13}{CEE\_\mathrm{C}\mathrm{o}\mathrm{m}\mathrm{p}\mathrm{l}\mathrm{a}\mathrm{c}\mathrm{e}\mathrm{n}\mathrm{t}}_{it}+{\gamma }_{14}{CEE\_\mathrm{C}\mathrm{S}\mathrm{R}}_{it}+{\gamma }_{15}{CEL\_\mathrm{M}\mathrm{i}\mathrm{d}}_{i(t-1)} \\ +{\gamma }_{16}{CEL\_\mathrm{H}\mathrm{i}\mathrm{g}\mathrm{h}}_{i(t-1)}+{\epsilon }_{it} \end{gathered}$$

(2)

$$\begin{gathered} {OP}_{it}=\left(2\right)+{\delta }_1{CE{E}_{\mathrm{C}\mathrm{o}\mathrm{m}\mathrm{p}\mathrm{l}\mathrm{a}\mathrm{c}\mathrm{e}\mathrm{n}\mathrm{t}}}_{it}\times {CE{L}_{\mathrm{M}\mathrm{i}\mathrm{d}}}_{i\left(t-1\right)}+{\delta }_2{CE{E}_{\mathrm{C}\mathrm{S}\mathrm{R}}}_{it}{\times CE{L}_{\mathrm{M}\mathrm{i}\mathrm{d}}}_{i\left(t-1\right)} \\ +{\delta }_3{CEE\_\mathrm{C}\mathrm{o}\mathrm{m}\mathrm{p}\mathrm{l}\mathrm{a}\mathrm{c}\mathrm{e}\mathrm{n}\mathrm{t}}_{it}\times {CEL\_\mathrm{H}\mathrm{i}\mathrm{g}\mathrm{h}}_{i(t-1)}+{\delta }_4{CEE\_\mathrm{C}\mathrm{S}\mathrm{R}}_{it}\times {CEL\_\mathrm{H}\mathrm{i}\mathrm{g}\mathrm{h}}_{i(t-1)}+{\epsilon }_{it} \end{gathered}$$

(3)

(OP = operating profit; reference groups: CEE = CSiR, CEL = Low; base year__t = 2013.)

This model specification thus captures the temporal, structural, and ethical determinants of corporate performance while maintaining parsimony and theoretical coherence.

4. Results

4.1. Descriptive Statistics

Table 4. Descriptive statistics and correlation matrix.

Variable	Mean	St. Dev.	1	2	3	4	5
Operating Profit_t	137.75	1374.19	1.00
Operating Profit_t−1	146.05	1384.01	0.75 ***	1.00
CEL_t−1			0.33 ***	0.33 ***	1.00
CEE_t−1,t			0.04 **	0.07 ***	−0.29 ***	1.00
Size			0.36 ***	0.36 ***	0.26 ***	0.04 *	1.00

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

shows the mean, standard deviation, and correlation of the main independent, dependent, and control variables used in this study’s model. The correlation coefficient between the operating profit of time t and the operating profit of time t − 1 was high and significant at 0.75. This outcome is expected and reinforces the usage of operating profit at time t − 1 as a control variable in the model. In addition, the correlation coefficients between the ethics level and the operating profit, and the company size and the operating profit were 0.33 and 0.36, respectively, showing a weak positive relationship consistent with existing studies. Analyzing VIF reveals that the highest VIF value does not exceed 1.1, so multicollinearity is not a concern.

4.2. Main Results

Table 5. Hierarchical regression results for the hypotheses.

	Model 1	Model 2	Model 3
CEE_Complacentt−1,t		−46.93	−41.49
CEE_CSRt−1,t		15.13	−41.27
CEL_Midt−1		−0.98	−44.73
CEL_Hight−1		197.84 **	414.09 ***
$\mathbf{CEE}\mathbf{\_}{\mathbf{Complacent}}_{\mathit{t}\mathbf{-}\mathbf{1}\mathbf{,}\mathit{t}} \mathbf{\times }$CEL_Midt−1			26.25
$\mathbf{CEE}\mathbf{\_}{\mathbf{CSR}}_{\mathit{t}\mathbf{-}\mathbf{1}\mathbf{,}\mathit{t}} \mathbf{\times }$CEL_Midt−1			156.70
$\mathbf{CEE}\mathbf{\_}{\mathbf{Complacent}}_{\mathit{t}\mathbf{-}\mathbf{1}\mathbf{,}\mathit{t}} \mathbf{\times }$CEL_Hight−1			−343.33 *
$\mathbf{CEE}\mathbf{\_}{\mathbf{CSR}}_{\mathit{t}\mathbf{-}\mathbf{1}\mathbf{,}\mathit{t}} \mathbf{\times }$CEL_Hight−1			−30.19
Intercept	5.79	43.22	56.02
Operating Profitt−1	0.74 ***	0.73 ***	0.73 ***
Size_Large	34.26	16.68	17.57
Yeart	Included	Included	Included
Adj. R2	0.5596	0.5608	0.5616
Change in R2		+0.0012	+0.0020
Overall F statistic	448.2 ***	338 ***	271.5 ***

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

summarizes the results of the regression analysis to verify the hypotheses. In Model 1, the operating profit of time t, which is the dependent variable, is explained significantly by the operating profit of time t − 1 (β = 0.74, p < 0.01).

The corporate ethical level (CEL) and ethical efforts (CEE) are added to Model 1 to create Model 2. As a regression analysis composed of dummy variables, CSiR became the reference group for corporate ethical efforts, and Low became the reference group for corporate ethical level. The adjusted coefficient of determination (Adj. R²) is 0.5608, which is 0.0012 higher than the corresponding value in Model 1. Thus, it is observed that the independent variables, and specifically the major variables of interest (corporate ethical level and ethical efforts), have statistically significant effects on the operating profit during period_t−1,t. In particular, companies with a high ethical level showed higher operating profit during period_t−1,t than companies with a low ethical level ($\gamma$ = 197.84, p < 0.01), consistent with previous studies. This result supports Hypothesis 1, in that a higher corporate ethical level at time t − 1 has a positive effect on CFP over period_t−1,t. These empirical results are consistent with the results of previous studies that CSR performance positively affects CFP. However, none of the corporate ethical efforts during period_t−1,t are statistically significant, thereby not supporting Hypothesis 2.

In addition to Model 2, Model 3 is an analysis model that considers the interaction between corporate ethical level and ethical efforts. The adjusted coefficient of determination in Model 3, which considers the interaction variables, is 0.5616, which is 0.0020 higher than Model 1, indicating that the interaction variables of corporate ethical level and ethical efforts are statistically significant and improve the model. Model 3′s regression coefficient estimate of a high ethical level is found to positively affect the operating profit in period_t−1,t, as in Model 2, and the absolute value of the coefficient is larger ($\gamma$ = 414.09, p < 0.01). Therefore, Hypothesis 1 is also supported in Model 3. However, Hypothesis 2 is not supported because there is no statistically significant difference between CEE, except for in the interaction terms. Yet, the interaction terms in Model 3 indicate a complex relationship of CEE with CEL and CFP.

The interaction is significant when the ethics level at time point t − 1 is High. More specifically, when the ethical level at time point t − 1 is High, the regression coefficient for Complacent versus the reference group (CSiR) for ethical effort is −343.33. This can be interpreted to mean that companies with a high CSR rating can obtain higher operating profit in the next period by acting socially irresponsibly. Interpreted another way, operating profit may suffer from opportunity cost due to unearned operating profits when companies at a high level of CSR continue to invest in CSR-related initiatives. Therefore, depending on the corporate ethical level at time point t − 1, the impact of its ethical efforts on CFP during period t will be differentiated, supporting Hypothesis 3. In particular, Hypothesis 3a is statistically supported because of the significantly negative coefficients for the interaction between high ethical level and complacent ethical efforts.

Figure 2 shows the analysis results of Model 3, with statistical significance indicated by solid-colored bars and non-significance indicated by diagonal lines. It demonstrates that the corporate ethical efforts during period_t−1,t affect the operating profit differently depending on the corporate ethical level at time t − 1. When the ethical level at time t − 1 is low, maintaining or improving ethical efforts during period_t−1,t from CSiR to CSR does not affect CFP for the same period_t−1,t. Conversely, when the ethical level at time t − 1 is high, maintaining ethical efforts during period_t−1,t, negatively impacts CFP during that period significantly.

These results suggest that it is costly for companies with high ethical levels to maintain them, which negatively impacts financial performance. On the other hand, if ethical management or decision-making is neglected, it can be interpreted that the resulting cost reduction results in a short-term increase in operating profit. When the level of corporate ethics is not high, raising the level of ethics and improving ethical efforts will help increase corporate financial performance. However, at the high level of ethics, continuing ethical efforts do not help improve financial performance. Rather, there may be a temptation to increase the operating profit of the current period (i.e., a short-term result) if ethical management is neglected due to hubris in the ethical performance accumulated so far. At the individual level, the current immoral behavior without guilt based on one’s moral behavior in the past is called the phenomenon of moral licensing [6]. The analysis result of this study related to corporate behavior is very similar to the flow of the moral license generation mechanism discussed at the individual level, but as applied to the corporate level.

4.3. Robustness and Supplementary Analyses

To ensure the reliability and stability of the regression results, several robustness and supplementary analyses were conducted. First, potential outliers in operating profit that could disproportionately influence the coefficients were removed based on studentized residuals exceeding ±3 standard deviations. The regression models were then re-estimated using the same specifications as in the main analysis. As reported in

Table A1. Regression Results after Removing Outliers (|Studentized Residual| > 3).

	Model 1	Model 2	Model 3
CEE_Complacentt−1,t		−10.63	−8.46
CEE_CSRt−1,t		7.02	−7.14
CEL_Midt−1		9.72	−3.32
CEL_Hight−1		71.77 ***	154.52 ***
CEE_Complacentt−1,t × CEL_Midt−1			8.94
CEE_CSRt−1,t × CEL_Midt−1			46.73 *
CEE_Complacentt−1,t × CEL_Hight−1			−122.11 ***
CEE_CSRt−1,t × CEL_Hight−1			−87.18
Intercept	−10.69	−7.18	−4.12
Operating Profitt−1	0.75 ***	0.75 ***	0.75 ***
Size_Large	16.74 *	8.20	8.11
Yeart	Included	Included	Included
Adj. R2	0.8931	0.8939	0.8947
Change in R2		+0.0008	+0.0016
Overall F statistic	2925 ***	2214 ***	1785 ***

*** p < 0.001, * p < 0.05. Note: Outliers were defined as observations with studentized residuals beyond ±3 standard deviations. A total of 22 outliers were removed, resulting in a final sample size of 4202 firm-year observations. The direction and significance of key coefficients remain consistent with the main analysis, confirming robustness to extreme values.

, the results remained qualitatively consistent with the main findings, confirming that the observed relationships—particularly the moderating effect of corporate ethical level on the link between ethical efforts and operating profit—are not driven by extreme observations.

Second, to address potential skewness and heteroskedasticity in the dependent variable, the operating profit was log-transformed and the models were re-estimated. As presented in

Table A2. Regression Results Using Log-transformed Operating Profit.

	Model 1	Model 2	Model 3
CEE_Complacentt−1,t		−0.09	−0.16
CEE_CSRt−1,t		0.00	−0.13
CEL_Midt−1		0.19 ***	0.07
CEL_Hight−1		0.68 ***	0.90 ***
CEE_Complacentt−1,t × CEL_Midt−1			0.12
CEE_CSRt−1,t × CEL_Midt−1			0.28 *
CEE_Complacentt−1,t × CEL_Hight−1			−0.34 *
CEE_CSRt−1,t × CEL_Hight−1			−0.32
Intercept	0.11	0.11	0.19
Operating Profitt−1	0.79 ***	0.76 ***	0.76 ***
Size_Large	0.41 ***	0.37 ***	0.37 ***
Yeart	Included	Included	Included
Adj. R2	0.6706	0.6772	0.6780
Change in R2		+0.0066	+0.0074
Overall F statistic	717.4 ***	554.7 ***	445.6 ***

*** p < 0.001, * p < 0.05. Note: Using the log-transformed dependent variable produces consistent coefficient directions with the main analysis, confirming robustness to functional form changes.

, the results based on log-transformed operating profit were directionally consistent with those of the main analysis, further demonstrating the robustness of the empirical findings. This supplementary analysis also confirms that the moderating effect of ethical level remains statistically significant and theoretically meaningful under alternative variable specifications.

Third, to assess the practical relevance of the explanatory variables beyond statistical significance, both standardized regression coefficients and relative importance (LMG) analyses were conducted. As shown in Appendix B, the standardized coefficients (

Table A3. Standardized Regression Coefficients.

Variable	Standardized β
CEE_Complacent	−0.015
CEE_CSR	−0.011
CEL_Mid	−0.016
CEL_High	0.078
Size_Large	0.005
Operating Profit_t−1	0.738
Year Controls	Controlled
CEE_Complacentt−1,t × CEL_Midt−1	0.009
CEE_CSRt−1,t × CEL_Midt−1	0.027
CEE_Complacentt−1,t × CEL_Hight−1	−0.053
CEE_CSRt−1,t × CEL_Hight−1	−0.001

Note: Standardized coefficients (β) were calculated using the lm.beta package in R. The largest standardized effect was observed for the lagged dependent variable (Operating Profit__t−1 = 0.738), followed by the high corporate ethical level (0.078). Interaction effects exhibited small magnitudes but remained consistent in direction with the main analysis.

) indicate that the lagged operating profit (β = 0.738) exerts the largest standardized effect, followed by the high corporate ethical level (CEL_High; β = 0.078). Interaction terms (CEE × CEL) show small but directionally consistent effects with the main analysis.

The relative importance results (

Table A4. Relative Importance (LMG) of Explanatory Variables.

Variable	Relative Importance (%)
CEE	0.23
CEL	4.27
CEE × CEL	0.46
Year Controls	0.38
Firm Size (Large)	0.09
Operating Profit_t−1	94.58

Note: Relative importance (LMG %) was estimated using the relaimpo package (method = “lmg”). The lagged dependent variable explains approximately 94.6% of the variance in operating profit, reflecting high temporal persistence. Among the explanatory variables of theoretical interest, corporate ethical level (CEL) accounts for 4.3%, followed by the interaction term CEE × CEL (0.5%), confirming that ethical dimensions contribute meaningfully—though modestly—to explaining firm performance variation. All metrics are normalized to sum to 100%.

) further reveal that the lagged operating profit explains approximately 94.6% of the model’s R², while the corporate ethical level (CEL) accounts for about 4.3%, and the interaction term (CEE × CEL) contributes roughly 0.5%. Although the incremental increase in adjusted R² across models is modest, these decompositions confirm that ethical dimensions—especially a high ethical level and its interaction with ethical efforts—contribute meaningfully to the overall explanatory power of the model and thus possess substantive managerial relevance.

Finally, potential endogeneity among CEL, CEE, and corporate performance was considered. The study design incorporates a one-period lag structure—where CEL is measured at time t − 1, CEE is defined over period_t−1,t, and CFP is observed at time t—to ensure temporal precedence of ethical variables. This temporal ordering mitigates simultaneity bias and supports an explanatory interpretation of causal direction.

Nevertheless, as firm-level heterogeneity and unobserved confounders cannot be fully eliminated in non-experimental data, future studies could employ instrumental-variable or dynamic-panel approaches (e.g., system GMM) to further assess causal robustness.

Collectively, these robustness and supplementary analyses confirm that the moderating effect of corporate ethical level remains statistically, theoretically, and economically robust across multiple alternative model specifications.

5. Discussion

CSR has expanded and evolved into related concepts such as ESG and stakeholder capitalism, and is now regarded as a central pillar of sustainable management in both academia and practice. Building on this evolution, this study advances prior research by empirically distinguishing between two interrelated but distinct dimensions of CSR—corporate ethical level (CEL) and corporate ethical effort (CEE)—and by examining their interactive impact on corporate financial performance (CFP). The results of this study shed new light on the intricacies of CSR impacts on CFP. Specifically, it demonstrates that CSR performance level and ethical effort are conceptually and empirically distinct constructs whose interaction produces significant effects on firm outcomes. This nuanced relationship can be more fully understood within the framework of instrumental stakeholder theory, which emphasizes the creation of long-term sustainable value.

Although the models in this study are statistically significant, the explanatory power remains relatively modest. This indicates that while ethical variables have meaningful effects, they explain only part of the variance in corporate financial performance. Other contextual and firm-specific factors such as corporate culture, market dynamics, and industry characteristics are also likely to influence the CSR–CFP relationship. Therefore, the findings should be interpreted as revealing one important dimension of this complex linkage rather than as a comprehensive explanatory model.

In this study, the concept of corporate ethical efforts is expanded from corporate social responsibility (CSR) to corporate social irresponsibility (CSiR) theory, which represents the opposite end of the ethical spectrum. A more in-depth empirical analysis is conducted compared with prior research by subdividing CSR activities and performance into corporate ethical levels and ethical efforts. Based on stakeholder theory and agency theory, hypotheses are established regarding the effects of corporate ethical level and ethical efforts on corporate financial performance (CFP), as well as their potential interaction. To test these hypotheses, the study employs ordinary least squares (OLS) regression with year fixed effects (year dummy variables), using ESG rating data and financial performance data from 384 publicly listed Korean manufacturing companies evaluated by the Korea Corporate Governance Service (KCGS) over the twelve-year period from 2012 to 2023.

Methodologically, this study focuses on between-firm variation rather than within-firm temporal changes. The purpose is to identify structural performance differences across firms that vary in their ethical orientations, rather than short-term shifts within the same firm. Accordingly, OLS with year fixed effects was selected, as it preserves cross-sectional differences central to the theoretical framework. To parsimoniously control for firm-specific persistence without overfitting, the model includes lagged operating profit (OP_t−1), which captures prior performance momentum and absorbs a wide range of firm-level unobserved influences. While firm fixed effects or dynamic panel estimators could further strengthen causal inference, the present design appropriately aligns with the study’s aim to reveal systematic performance differentials among firms with distinct ethical profiles.

As a result of the analysis, it is shown that a higher corporate ethical level positively affects CFP, confirming the validity of instrumental stakeholder theory. However, the interaction between corporate ethical level and ethical effort—newly proposed in this study—reveals a paradoxical short-term effect: when the corporate ethical level is already high, continuing ethical efforts to fulfill CSR can adversely impact short-term financial performance. In other words, neglecting ethical efforts may yield temporary improvements in operating profit.

Importantly, this finding does not imply simple complacency or negligence among highly ethical firms. Rather, maintaining top ESG performance itself requires substantial and ongoing investments of resources and managerial attention. This ongoing cost structure can create pressure to temporarily relax ethical efforts in pursuit of short-term gains. In this sense, moral licensing arises not from passive neglect but as a cost-driven temptation—a strategic rationalization process in which firms justify minor irresponsibility under the burden of sustaining ethical excellence. Accordingly, this study provides empirical evidence that corporate moral licensing can emerge at the organizational level, posing systemic risks to sustainability.

However, the results did not support the hypothesized moral cleansing effect. This finding suggests that firms with lower ethical levels may lack genuine motivation to restore legitimacy after engaging in unethical or irresponsible behavior. Rather than pursuing substantive ethical improvement, these firms may focus on short-term financial recovery or superficial reputation management to mitigate immediate losses.

In addition, firms with weak ethical foundations often face financial and organizational resource constraints that limit their capacity to implement substantive CSR reforms, which further inhibits the occurrence of moral cleansing. This asymmetry between licensing and cleansing aligns with prior behavioral theories suggesting that the psychological and institutional conditions necessary for moral restoration are rarely met in firms driven primarily by short-term survival motives [5,22]. Consequently, moral cleansing behavior appears unlikely in such firms, which tend to prioritize self-interested strategic considerations over genuine ethical restoration. Future research could further explore how such motivational asymmetry between moral licensing and moral cleansing develops under different institutional or cultural settings.

The results based on Korean corporate data are consistent with global debates, despite cultural and regulatory differences from Western contexts. This alignment strengthens the external validity of the findings and enhances the generalizability of stakeholder theory across diverse institutional environments. Importantly, the discovery of a moderating effect consistent with moral licensing theory adds a new dimension to CSR research, illustrating how ethical achievements can paradoxically generate behavioral complacency. This insight highlights that CSR is not a static commitment but a dynamic process in which ethical orientation must be continuously renewed to sustain its positive impact. From a practical perspective, recognizing these structural risks is essential for managers and policymakers to design preventive measures that safeguard both short-term performance and long-term sustainability.

Specifically, the emergence of corporate moral licensing may foster complacency after strong ESG achievements, weakening firms’ commitment to responsible resource use, ethical supply chain management, and transparent governance. Such post-achievement complacency can erode the genuine ethical vigilance that underpins sustainable management. As a result, these behavioral tendencies threaten key Sustainable Development Goals (SDGs)—particularly SDG 12 (Responsible Consumption and Production) and SDG 16 (Peace, Justice, and Strong Institutions)—by undermining responsible production practices and institutional trust. Managing these behavioral risks is crucial to ensure that CSR reinforces, rather than weakens, progress toward global sustainability goals.

Nevertheless, prior empirical studies in the corporate ethics and ESG domains highlight that CSR initiatives can enhance competitiveness and reduce financial risk when firms maintain consistent ethical engagement and credible governance structures [23,24]. This complementary evidence suggests that the potential downsides of CSR arise not from ethical engagement itself but from inconsistent or performative practices. Taken together, these insights emphasize that ethical performance and ethical effort must be carefully balanced so that CSR continues to generate sustainable value rather than devolving into moral complacency.

Overall, these findings deepen our understanding of how moral licensing can compromise sustainable development goals and provide a foundation for future research examining the dynamic interplay between moral motivation, ethical behavior, and corporate performance across diverse institutional contexts.

6. Implications

6.1. Theoretical Implications

This study makes major academic contributions in two main aspects. First, from the methodological perspective of the conceptual measurement for empirical analysis, this is the first study to use dynamic changes in CSR activities and performance as a concept of corporate ethical efforts through measurement with long-term data from a credible evaluation agency. Specifically, this study defines the dynamic change in the corporate social responsibility performance level over time as a company’s ethical efforts and analyzes it as a key variable of interest. Empirical studies that have measured and analyzed corporate social responsibility have either analyzed the level of corporate social responsibility performance at a point in time or indirectly analyzed changes in performance through panel data analysis. However, this study went further and measured changes in CSR performance levels over time, defined as a company’s ethical effort, and reflected this in the analysis. Through this methodology, the changes are analyzed regarding corporate attitudes toward ethics over time, and interactions between ethical levels, which had not previously been analyzed, thereby deriving richer implications.

Second, this research supports instrumental stakeholder theory related to CSR, which has been widely accepted throughout academia and practice and contributes to revitalizing related research by presenting empirical evidence that deepens understanding. In this study, in addition to the level of corporate ethics, which has been measured in previous studies [25,26,27,28,29], the dynamic change in corporate ethics is also measured through defining corporate ethical efforts. This study also considers the interaction between the two and empirically analyzes their complex impact on corporate financial performance. Instrumental stakeholder theory, which has shown a weak positive relationship between CSR and CFP, is accepted as the mainstream opinion amidst contradictory results in the extant literature, wherein both positive and negative outcomes regarding the relationship between CSR activities and CFP have been observed [30]. Due to the treatment of corporate ethical level and corporate ethical efforts not being differentiated among the body of extant literature, it is possible that the ethical efforts and ethical level have confounded the measurement of CSR, which have been shown in the current study to be distinctively different concepts. This empirical evidence raises the understanding of related research in that it presents a theoretical background that can be interpreted by integrating the conflicting research results, and demonstrates that corporate ethical level and corporate ethical efforts over time are unique concepts that need to be taken into account. From these theoretical insights emerge several practical implications, which are discussed below.

6.2. Managerial Implications

This study informs managers and other stakeholders of a systematic moral licensing phenomenon occurring at the firm level, which can undermine a firm’s efforts toward sustainability. As a result of the analysis, it is shown that a higher level of corporate ethics at time t − 1 has a positive effect on CFP over the period_t−1,t, which is in line with the existing instrumental stakeholder theory [10,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47]. However, the interaction between the ethical level at time t − 1 and the ethical effort of the period_t−1,t is found to have a negative effect on CFP during period_t−1,t. This is a novel approach in CSR research, indicating that continuously requiring ethical efforts from companies with already high ethical levels can have a negative impact on CFP in the short term. Consequently, companies with high ethical levels may be more likely to succumb to the temptation of acting socially irresponsibly in pursuit of short-term profit—a behavioral pattern referred to as corporate moral licensing (CML). Paradoxically, neglecting ethical efforts can lead to improved short-term CFP for companies that already possess high CSR ratings.

These paradoxical findings imply that there is a systematic and predictable risk of a corporate moral license phenomenon in which highly ethical companies excuse themselves based on their moral behavior, when combined with poor decision-making biased toward improving short-term CFP. Up until now, the validity of the moral licensing theory has been demonstrated and discussed through experiments at the individual level [48,49]. Findings show that the continuous pursuit of good management negatively impacts the short-term performance of operating profit for a firm with a high ethical level due to various constraints and cost increase factors, relative to poorer ethical practices. Conversely, if ethical management is neglected, there is the possibility that operating profit for the current period will improve due to various cost reductions. These findings imply that fundamentally, for-profit organizations facing strong short-term performance pressures may have structural incentives to justify socially irresponsible behavior through self-excusing moral reasoning.

To mitigate this risk, managers should design governance systems that balance ethical engagement with financial discipline. Incorporating long-term sustainability metrics into performance evaluations and linking executive incentives to consistent ethical conduct can help prevent moral licensing. Encouraging transparent reporting and stakeholder dialogue can further strengthen ethical accountability. Ultimately, recognizing and addressing these moral risks enables firms to maintain both ethical integrity and sustainable financial performance.

7. Limitations

This study has the following limitations due to the nature of empirical research, and supplementing and improving them should be the subject of follow-up research. First, in this study, the dependent variable of corporate performance is selected as the current operating profit. However, other indicators, such as Tobin’s Q, ROA, and stock return rate, also measure corporate performance. As each indicator has a slightly different concept and meaning to its measurement, in this study, operating profit for the current period has been used as the dependent variable to measure the more direct effect that the independent variable would have on the dependent variable in the short term. However, considering a time lag, it would be a meaningful research topic to examine how CSR activities affect the long-term financial performance of companies, in which case other dependent variables might better encompass the concept of long-term financial performance. In the case of analyzing the long-term financial impacts of CSR, the authors suggest that other dependent variables be considered for such measurements. Previous studies claiming a positive relationship between CSR activities and CFP insist that the effect appears in the long run, so empirical analysis of this will help expand the foundation for related research.

On the other hand, in this study, the relationship between CSR and CSiR is analyzed, assuming they are opposing concepts of the same dimension. However, there are many discussions that the two concepts exist in different dimensions [4,14,16,17,18]. In the process of this empirical analysis using the ESG-integrated evaluation rating data of Korean publicly listed companies, due to the nature of the data, it is not possible to separately measure or analyze CSR and CSiR independently of one another. Since KLD’s Social Performance Ratings Data, mainly used in existing studies, has separate items for measuring CSR and CSiR, it is possible to conduct empirical analysis from the standpoint of claiming that the two concepts exist independently. However, using both concepts as opposites in the same dimension has precedent in other research and is still a sufficiently valid viewpoint supported by other scholars [15,19]. In particular, there is no logical problem in the assumption because it results in “ethical or unethical” as alternatives rather than as compatible in the final comprehensive evaluation of the ethical disposition of an individual or organization. The authors suggest that separating these concepts into different dimensions for future research may result in a more nuanced view of their interactions.

Moreover, within the framework of a moral balancing mechanism, which is the concept that moral and immoral actions have an offset effect on each other’s actions [21], it is shown to occur for corporations at a high ethical level behaving unethically in pursuit of short-term profit (i.e., “corporate moral licensing”), but the opposite is not observed where corporations at a low ethical level tend to put more efforts into becoming more socially responsible (i.e., “corporate moral cleansing”). The lack of support for this corporate moral cleansing makes it difficult to discuss it comprehensively from the viewpoint of the moral balancing mechanism. Corporate moral licensing was statistically significant, but corporate moral cleansing was not. More research should be done to better understand why this is the case. One possible explanation is that firms with low ethical levels may face severe financial and organizational resource constraints, which limit their capacity to engage in genuine moral restoration or long-term CSR investment, thereby suppressing observable moral cleansing effects within the analyzed period.

Another limitation concerns potential endogeneity among corporate ethical level (CEL), corporate ethical efforts (CEE), and corporate financial performance (CFP). While this study applies pooled OLS with year fixed effects and includes lagged operating profit to control for prior firm performance, such an approach cannot fully eliminate potential reverse causality or omitted-variable bias. The dataset used in this research does not allow for the construction of valid instrumental variables (IV), which limits the ability to address endogeneity through econometric estimation. Future research should therefore consider applying IV or generalized method of moments (GMM) approaches to strengthen causal inference and validate the robustness of the relationships identified in this study.

Lastly, this study utilizes data from Korean publicly listed manufacturing firms, which may reflect the unique institutional and cultural characteristics of Korea’s business environment. In particular, large conglomerates (chaebols) often dominate the market and shape CSR activities through hierarchical decision-making and compliance-driven motivations rather than voluntary stakeholder engagement. Such structural and cultural specificities may restrict the generalizability of the findings to other contexts where CSR initiatives are shaped by different governance systems and stakeholder pressures. Therefore, future research should pursue cross-cultural or comparative analyses to determine whether the mechanisms of corporate moral licensing and moral cleansing identified in this study persist across diverse institutional and cultural settings.

8. Conclusions

Incidents driven by the illusion of moral licensing can lead to severe societal and environmental consequences. For example, in 2008, Tony Hayward, CEO of British Petroleum (BP), took pride in promoting their safety record, education, and a culture that values safety. However, two years later, the Deepwater Horizon oil spill, hailed as the worst in history, occurred when management overlooked key safety warning signs [50]. This contradiction illustrates how past moral confidence can foster complacency, echoing the mechanisms identified in this study.

This study demonstrates that moral licensing is not only an individual phenomenon but also a systematic risk at the organizational level, with direct implications for sustainability. Firms with high ethical standards may paradoxically reduce their ethical efforts. While this may yield short-term financial gains, it ultimately undermines sustainable value creation. Such dynamics create serious risks for society and the environment, escalating into large-scale sustainability crises.

For stakeholders, policymakers, and managers, these findings underscore the urgent need to ensure ethical consistency, strengthen ESG governance, and design accountability mechanisms that align short-term incentives with long-term sustainability goals. By empirically revealing the risks of moral licensing within a sustainability framework, this study advances sustainable management theory and provides actionable guidance for achieving the Sustainable Development Goals (SDGs) through responsible corporate practice. In summary, sustaining ethical vigilance is not only a moral imperative but also a strategic necessity for long-term corporate resilience.