Ethical Dilemmas in Performance-Oriented Management: A Dual-Path Systems Model

Abstract

Background: High-performance work systems (HPWSs), while designed to boost corporate performance, can inadvertently create a core organizational paradox, triggering a negative feedback loop. Specifically, their intense focus on performance outcomes can create a climate conducive to unethical pro-organizational behavior (UPB), as employees navigate the pressures and perceived obligations, ultimately undermining the organization’s long-term sustainability and viability. While prior research has identified important singular pathways, the mechanisms through which HPWSs simultaneously generate both perceived obligations and performance pressures remain ambiguous. Methods: Drawing on the Job Demands–Resources (JD-R) model, we propose and test a moderated dual-mediation framework. Using survey data from 473 employees, we examine psychological contract fulfillment and bottom-line mentality as parallel mediators, with moral identity as a moderator, in the HPWS-UPB relationship. Results: The analysis demonstrated that HPWSs influence UPB through two distinct and paradoxical pathways: a pressure-driven path via an increased bottom-line mentality, and a reciprocity-driven path via enhanced psychological contract fulfillment. Moral identity emerged as a crucial, albeit asymmetrical, buffer, with its buffering role being particularly consequential for the pressure-driven pathway, as moral identity also significantly weakened the indirect effect of HPWSs on UPB channeled through bottom-line mentality. Conclusions: These findings offer a holistic, systems-based understanding of the performance-ethics paradox. The validation of a dual-pathway model provides a new blueprint for how a single management system produces contradictory outcomes through competing mechanisms. The identification of key intervention points (e.g., fostering moral identity) offers practical strategies for managers to foster systems that support both high productivity and a sustainable ethical climate.

1. Introduction

From a systems perspective, the relentless pursuit of corporate performance has led modern enterprises to adopt sophisticated Human Resource Management (HRM) frameworks. High-performance work systems (HPWSs), a strategic HRM approach, are celebrated for their potential to improve both employee and organizational performance [1]. In the context of China’s high-quality economic transition, these systems often manifest as a hybrid, combining commitment-based practices (e.g., development opportunities) with stringent control-based approaches (e.g., strict performance standards) [2,3]. For instance, the e-commerce giant JD.com implements both generous incentives and rigorous ranking systems [4].

Adopting a more holistic view is crucial, however, as emerging evidence suggests that the benefits of HPWSs are not unequivocal. While the benefits of HPWSs are well-documented [5], their control-oriented aspects trigger significant unintended consequences [6], giving rise to a persistent organizational paradox. This paradox is defined as “contradictory yet interrelated elements that exist simultaneously and persist over time” [7]. The intense performance pressures from practices like “last-place elimination” can increase employee stress and burnout [8,9,10]. Indeed, this tension is not unique to the Chinese context. A growing body of international research highlights a significant dark side to HPWSs, where the very practices designed to boost performance paradoxically harm employee well-being and increase work-related stress [11,12]. This global scholarly conversation reflects a growing consensus that the performance-ethics paradox is a widespread organizational challenge, a notion supported by recent research in Western contexts demonstrating how these systems intensify hindrance stressors for employees [13]. We argue that these pressures create a deeper, more systemic problem: they erode an organization’s most vital hidden asset—its psychological safety. This study examines how this erosion fosters unethical pro-organizational behavior (UPB), defined as actions that benefit the organization while violating ethical standards [14]. This creates a critical paradox: despite potential short-term advantages, the resulting UPB ultimately damages an organization’s reputation and long-term sustainability [15].

To date, however, the processes by which the competing demands and resources within HPWSs systematically generate these paradoxical outcomes remain ambiguous. Addressing this gap, our study moves beyond a static analysis by framing HPWSs as a dynamic, iterative management cycle. Driven by institutional pressures for legitimacy and efficiency [16], this cycle becomes deeply embedded in an organization’s daily life, solidifying over time into a set of powerful, taken-for-granted routines [17]. It is within these very routines—the “actual culture” revealed through recurring practices like performance evaluation [18]—that the paradoxical tension arises. Accordingly, we formally frame this tension as an organizational paradox—a persistent contradiction between interdependent elements [7]—aiming to model the mechanisms through which these entrenched routines simultaneously foster both commitment and pressure.

Current UPB research follows two divergent paths. Social exchange theory emphasizes pro-organizational motives [15,19], while trait activation theory focuses on self-serving drivers like Machiavellianism [20,21,22,23]. A notable exception is the foundational work by Xu and Lv (2018), which compellingly demonstrated that HPWSs can foster UPB through the single mediating mechanism of psychological ownership [24]. Inspired by their pioneering research, our study delves deeper into the organizational paradox they brought to light, examining the internal contradictions of how HPWSs function as both a source of commitment and a source of pressure. To conduct this more nuanced examination, the Job Demands–Resources (JD-R) model is particularly well-suited to address this limitation. As a unifying framework capable of deconstructing this organizational paradox, a perspective consistent with a burgeoning research stream that utilizes the JD-R model to unpack competing tensions in organizations [25,26], the model allows us to posit that HPWSs simultaneously provide resources (potentially distorting pro-organizational intent) and impose demands (triggering a bottom-line mentality) [20,27]. It allows us to propose a complementary set of mechanisms, offering an integrated explanation that builds upon the foundational single-pathway model [24].

Accordingly, this study advances a moderated dual-pathway systems model to disentangle these interwoven dynamics. HPWSs are hypothesized to shape UPB via two countervailing channels: (1) a reciprocity-driven path mediated by psychological contract fulfillment and (2) a pressure-driven path mediated by the cultivation of a bottom-line mentality, wherein performance imperatives overshadow all other considerations [28]. Moreover, moral identity emerges as a pivotal boundary condition capable of mitigating the adverse systemic pressures [29]. This study contributes by: (1) applying a systems thinking lens to reveal the unintended consequences of HPWSs [30], thereby constructing a holistic representation of the interplay between performance imperatives and ethical equilibria within organizational systems; and (2) advancing UPB literature by providing a comprehensive framework that integrates and reconciles its dual motivational nature [31]. Crucially, the dual-pathway approach seeks to complement and extend the foundational insights of prior work, particularly the compelling link established through the mechanism of psychological ownership [24]. Rather than viewing this model as an alternative to the important role of ownership, it is intended to illuminate two additional, parallel pathways—one rooted in distorted reciprocity and the other in performance pressure—that operate concurrently. This research, therefore, aims to provide a more holistic and nuanced understanding of the complex ethical dilemmas that managers routinely confront.

2. Theoretical Foundation and Hypotheses

As established in the introduction, the dynamic, iterative cycle of HPWSs solidifies into powerful routines that simultaneously provide employees with valuable resources while imposing intense performance demands [7]. The core paradox, best deconstructed through the JD-R model, creates the conditions for two competing psychological mechanisms to emerge. This study conceptualizes these mechanisms as a dual-pathway model, which we use to explain how HPWSs can inadvertently foster UPB. The following sections will develop hypotheses for each of these pathways: a reciprocity-driven path, rooted in the system’s provision of resources, and a pressure-driven path, stemming from its imposition of demands. The dual-pathway approach aims to build upon foundational research by providing a more comprehensive, systemic explanation for the HPWS-UPB link.

2.1. The Mediating Role of Psychological Contract Fulfillment

The reciprocity-driven path of the HPWSs paradox begins with the system’s role in shaping the psychological contract. HPWSs represent a coherent architecture of human resource practices designed to signal an organization’s long-term commitment to its employees. By providing a rich array of job resources [5], such systems act as tangible evidence of the organization’s intent to invest in and support its workforce, thereby fostering a sense of reciprocity [32] central to the formation and maintenance of a healthy psychological contract. The important role of this social exchange dynamic was powerfully illustrated in the HPWS-UPB context, identifying psychological ownership as a key conduit [24].

The psychological contract is predicated on employees’ beliefs regarding mutual obligations, and its fulfillment hinges on the perception that the organization has honored its promised obligations [33]. HPWSs are instrumental in solidifying employee perceptions of contract fulfillment. Functioning as a “resource pool” [34], HPWSs institutionalize organizational support through specific practices. For instance, comprehensive training, clear promotion pathways, and participatory decision-making are not merely operational tools but rather potent signals of the organization’s commitment to employee development, fairness, and valuing employee contributions [35].

When employees experience the tangible benefits of such structured investments—gaining new skills, observing equitable advancement opportunities, and feeling their voice is heard—they are likely to conclude that the organization is upholding its end of the employment bargain [36,37]. Such a conclusion, wherein the organization is seen as fulfilling its promises of career growth, fair treatment, and recognition, directly constitutes psychological contract fulfillment [38]. The systematic and resource-intensive nature of HPWSs therefore sends a clear message of organizational obligation and support, fostering positive employee evaluations of the employment relationship.

Accordingly, we propose:

H1a: 

HPWSs positively influence psychological contract fulfillment.

Although psychological contract fulfillment generally yields beneficial outcomes, it is precisely at this juncture that the core tension of the paradox inherent in HPWSs manifests. The theoretical significance of psychological contract fulfillment as a mediator is rooted in its capacity to precisely test the social exchange mechanism. By directly operationalizing the fulfillment of specific organizational promises [33], it moves beyond broader constructs of belonging. Such precision is crucial as it casts the core organizational paradox in sharp relief, demonstrating how a fundamentally positive action—keeping promises—can be perversely co-opted to motivate unethical behavior [7]. The intense performance-driven pressures, representing the system’s “control” pole, paradoxically distort the motivational impact of this “enablement” pole. Conservation of Resources (COR) theory posits that individuals endeavor to accumulate and safeguard valued resources, frequently deploying current assets to yield prospective returns [39]. Within this paradigm, employees construe unethical pro-organizational behavior (UPB) as a tactical investment of effort—a deliberate organizational input aimed at procuring assured incentives and preserving positional security [40].This instrumental mindset is especially pronounced when fulfillment manifests through a transactional prism in high-pressure, low-safety organizational environments [41], wherein obligation discharge signals not authentic concern but an unequivocal emphasis on measurable achievements as the employment bond’s foundation [42].

This transformation of pro-organizational motives is principally explained through two complementary theoretical lenses. From a social exchange perspective, psychological contract fulfillment can foster a potent but distorted sense of obligation, wherein employees feel compelled to reciprocate the organization’s “benevolence” by any means necessary, including unethical acts [43], leading to “defensive reciprocity”—strategic maneuvers to secure ongoing resource flows rather than an expression of sincere allegiance [44]. Complementing this account, a social identity perspective suggests that strong organizational identification can lead employees to prioritize in-group interests to such an extent that they relax their moral standards to “protect” the collective [43]. This process of moral disengagement, a foundational concept in psychosocial functioning [45], has been empirically confirmed as a key psychological mechanism through which organizational identification fosters UPB [46]. By subordinating ethical tenets to benefit accrual, they rationalize practices like data alteration or hazard obfuscation as indispensable for contractual goal realization [40].

This theoretical framing, which depicts a paradoxical “dark side” to positive organizational phenomena [43], finds robust empirical support. Indeed, meta-analytic evidence confirms that favorable antecedents such as organizational identification, high-quality leader-member exchange, and transformational leadership are positively and robustly associated with UPB [47]. Ultimately, organizational promise fulfillment under such conditions engenders a powerful, albeit erroneous, compulsion for employees to reciprocate via attainment, unhindered by ethical considerations.

Based on this analysis, we propose:

H1b: 

Psychological contract fulfillment positively influences employee UPB.

Psychological contract fulfillment serves as a pivotal mediating mechanism, converting the positive resource signals emanating from HPWSs into motivational drivers for UPB. By supplying abundant job resources, HPWSs cultivate employees’ perceptions of contract fulfillment [48], which, per COR theory, affirms the organization as a dependable resource provider and initiates a resource gain spiral, prompting reinvestment of efforts for amplified returns [39].

Such affirmation, however, induces a distortion in motivational dynamics, wherein reciprocity evolves into an instrumental exchange aimed at maximizing personal gains rather than embodying authentic loyalty [49]. Employees view the pursuit of organizational objectives as the optimal route to sustained resource accrual—encompassing rewards like bonuses, advancements, or prestige—thereby subordinating ethical considerations to end-oriented priorities, a pattern conducive to UPB [44,50].

In this instrumental context, UPB manifests as an ostensibly judicious approach [23], facilitated by neutralization techniques that recast unethical acts as organizational benefactions, alleviating guilt [45]. Paradoxically, contract fulfillment engenders justification, misconstruing organizational fairness as license for any-means-necessary protection of interests, amplified by robust identification that favors collective gains over ethical universals [43]. Consequently, the affirmative organizational cues morph into tacit endorsements of expedient, dubious tactics, particularly amid bottom-line-dominated environments [51].

Accordingly, psychological contract fulfillment channels the resource-endowing attributes of HPWSs into a potent yet aberrant impetus for UPB, thereby mediating the relationship between HPWSs and employee UPB.

H1c: 

Psychological contract fulfillment mediates the relationship between HPWSs and employee UPB.

2.2. The Mediating Role of Bottom-Line Mentality

While reciprocity-based mechanisms offer crucial insights [24], they present an incomplete picture. The theoretical significance of bottom-line mentality lies in its capacity to fill this gap by illuminating the overlooked “demands” side of the HPWS paradox. The pressure-driven path of the HPWSs paradox, operating in direct tension with the reciprocity-driven path, centers on the system’s inherent performance demands. While HPWSs are designed to enhance job resources (e.g., developmental opportunities, organizational support), their inherent performance demands simultaneously cultivate a high-pressure environment that acts as a significant drain on employees’ finite personal resources (e.g., cognitive focus, emotional energy). This distinction is critical: the very practices that build up organizational-level resources can deplete individual-level ones. The intense pressure to meet stringent targets, coupled with high-stakes evaluations, fosters a persistent threat of personal resource loss [27,34,52], pushing employees into a state of vigilance and interpersonal fear [53,54]. It is this perceived scarcity of personal resources—rather than a lack of job resources—that propels employees along the pressure-driven path, activating a defensive cognitive response that fosters a bottom-line mentality.

In line with COR theory, operating within such a threatening environment is psychologically taxing [55,56], compelling individuals to shift from proactive goal pursuit to a defensive, resource-preservation mode. This defensive shift precipitates a cognitive narrowing, where employees’ focus contracts from complex, long-term organizational goals to the most immediate and unambiguous metrics of individual survival: the bottom line [28].

A bottom-line mentality represents a utilitarian coping strategy adopted by employees under conditions of resource scarcity and work pressure [57]. The mindset prioritizes short-term, directly measurable actions over long-term, complex solutions whose value is harder to quantify [58]. While this focus may be adaptive for the individual in the short term, it systematically sidelines critical but less tangible organizational priorities like ethical conduct, innovation, and employee well-being. Consequently, the high-pressure climate engendered by HPWSs provides a direct impetus for this narrow, results-obsessed cognitive framework.

Based on this reasoning, we propose that the performance pressures inherent in HPWSs trigger a defensive cognitive response that crystallizes as a bottom-line mentality:

H2a: 

HPWSs positively influence a bottom-line mentality.

Employees with a bottom-line mentality elevate performance targets above all, systematically subordinating ethical considerations. This prioritization intensifies in low psychological safety climates, fostering conditions where UPB emerges as a seemingly rational choice. From a resource conservation viewpoint, individuals feel entrapped: they must expend substantial resources to meet goals but perceive seeking help or admitting errors—key to ethical resolution—as punishable weaknesses [59].

When legitimate methods prove inadequate for stringent targets, UPB becomes a risk-averse shortcut, masking ethical violations under pro-organizational guises to avoid punishment. For those with a bottom-line mentality, moral norms weaken; low psychological safety erodes remaining social barriers, heightening UPB likelihood [58]. Ultimately, UPB serves not true organizational loyalty but self-protection in resource-scarce, threatening environments, reframing unethical acts as advantageous to shield from sanctions while advancing performance [59].

Based on this reasoning, we propose:

H2b: 

A bottom-line mentality positively influences employee unethical pro-organizational behavior.

HPWSs impose substantial job demands on employees through intensive goal setting, stringent performance evaluations, and competitive incentive structures, thereby engendering perceptions of work pressure and resource depletion threats as outlined in the resource loss pathway of the JD-R model. COR theory elucidates that sustained resource erosion induces a protective mode, wherein individuals streamline objectives and constrict strategies to forestall further losses, perpetuating a depletion cycle [60]. Consequently, amid HPWSs-induced pressures, employees facing resource scarcity adopt pragmatic, utilitarian responses, fostering a bottom-line mentality—an outcome-centric mindset that privileges results over ethical deliberations [51].

This constrained orientation shapes behavioral patterns: to evade sanctions for underperformance, employees render ethical standards flexible, rationalizing compromises for goal attainment; when conventional tactics falter against rigorous benchmarks, they exploit metric loopholes via atypical, potentially unethical means. Although such actions, termed UPB, deliver short-term organizational advantages [14], they originate from self-preservation rather than altruism, masking misconduct as loyalty to avert failure’s penalties.

Thus, the bottom-line mentality mediates the HPWS-UPB linkage: rigorous demands inherent in HPWSs incite resource strain, nurturing this mentality that elevates performance imperatives and prompts risk-averse UPB engagement. This rationale supports our proposition:

H2c: 

A bottom-line mentality mediates the relationship between HPWSs and employee UPB.

2.3. The Moderating Role of Moral Identity

Moral identity encompasses an individual’s self-concept centered on enduring moral attributes, denoting the extent to which ethical qualities define their core identity [61], thereby reshaping how employees reconcile organizational imperatives with moral standards and yielding distinct patterns in ethical reasoning and conduct [62]. Individuals possessing robust moral identity prioritize virtuous elements—such as integrity, empathy, and equity—within their self-perception, fostering alignment with altruistic principles, whereas those with diminished moral identity marginalize such traits, amplifying self-interested inclinations and eroding adherence to ethical norms [63]. These disparities engender varied ethical appraisals: stronger moral identities heighten sensitivity to and rejection of improprieties, while weaker ones facilitate rationalization of violations for pragmatic ends [64], positioning moral identity as a pivotal moderator that bolsters resilience against ethically corrosive influences, particularly in low-safety contexts induced by performance pressures. While prior research has firmly established the protective role of moral identity [24], this study addresses a critical gap by examining whether the buffering effect of moral identity is uniform across the distinct reciprocity-driven and pressure-driven pathways.

In modulating the linkage between psychological contract fulfillment and UPB, robust moral identity internalizes ethical tenets as integral to self-concept, prompting employees to scrutinize resource exchanges through a principled framework; fulfillment thus amplifies moral self-regulation, channeling reciprocity toward conscientious efforts rather than expedient outcomes, thereby attenuating self-serving motives and curbing UPB propensity [65,66]. Conversely, feeble moral identity renders employees prone to instrumental interpretations of fulfillment, prioritizing rewards over scruples and fostering moral disengagement that rationalizes transgressions like data falsification for contractual goal achievement, consequently heightening UPB engagement [67].

Similarly, moral identity tempers the association between bottom-line mentality and UPB: employees with elevated moral identity exhibit fortified defenses against mentality formation amid high-performance work systems (HPWSs) demands, leveraging internalized ethics to sustain integrity and avert unethical shortcuts despite resource strains and diminished safety [68]. In contrast, those with attenuated moral identity succumb more readily to outcome-focused cognition under depletion, justifying metric manipulations to evade repercussions and escalating UPB likelihood [69].

Based on this theoretical framework, we propose:

H3: 

Moral identity negatively moderates the relationship between psychological contract fulfillment and unethical pro-organizational behavior.

H4: 

Moral identity negatively moderates the relationship between bottom-line mentality and unethical pro-organizational behavior.

Building on the established mediation and moderation mechanisms, this study proposes that moral identity serves as a boundary condition for the mediating effects of both psychological contract fulfillment and bottom-line mentality. Due to the alignment between moral principles and self-concept, individuals with strong moral identity demonstrate remarkable consistency in ethical judgment across situations, relying primarily on internalized moral standards rather than external cues [70]. In contrast, those with weak moral identity exhibit greater moral plasticity, showing weaker self-regulation and greater susceptibility to situational pressures in ethical decision-making [71].

This differential manifests in two key ways: First, high moral identity employees treat ethical principles as non-negotiable behavioral constraints, making them less likely to: (1) pursue organizational rewards through unethical means, or (2) compromise standards to avoid personal risks. Second, their steadfast adherence to moral standards weakens the transmission of HPWS effects through both mediators (psychological contract fulfillment and bottom-line mentality). Conversely, low moral identity employees demonstrate stronger mediation effects, as their greater reliance on external cognition makes them more responsive to both the reward signals of psychological contract fulfillment and the pressure of bottom-line mentality.

Integrating our mediation hypotheses (H1c, H2c) with moderation hypotheses (H3, H4), we propose the following moderated mediation hypotheses:

H5: 

Moral identity negatively moderates the mediating effect of psychological contract fulfillment on the relationship between HPWSs and UPB.

H6: 

Moral identity negatively moderates the mediating effect of bottom-line mentality on the relationship between HPWSs and UPB.

In summary, integrating a systems perspective with the dual-pathway logic of the JD-R model, this study proposes a moderated mediation framework (Figure 1) in which HPWSs serve as the independent variable, UPB as the dependent variable, psychological contract fulfillment and bottom-line mentality as parallel mediators, and moral identity as the moderating variable.

3. Materials and Methods

3.1. Sample and Procedures

To enhance the generalizability of our findings, this study adopted a multi-site purposive sampling strategy, drawing its sample from organizations across three of China’s distinct and economically significant regions. The selection of provinces was deliberate to ensure heterogeneity: Zhejiang was chosen to represent the highly developed, export-oriented economy of the Yangtze River Delta; Henan was selected for its role as a major agricultural and logistical hub, typifying the vast inland economy of Central China; and Hebei was included to reflect the heavy-industry characteristics of North China. This stratified purposive approach ensures the sample is not confined to a single regional context, thereby strengthening the external validity of the results. Collaborations were subsequently established with the human resources or administrative departments of approximately 25 organizations across these provinces to facilitate on-site data collection.

The initial distribution of 600 questionnaires yielded 545 returns, representing a preliminary response rate of 90.83%. To ensure data integrity, these questionnaires underwent a meticulous screening process, during which responses that were substantially incomplete or contained significant logical inconsistencies were excluded. This validation procedure resulted in a final, usable sample of 473 questionnaires, corresponding to an effective response rate of 78.83%, which formed the basis for all subsequent statistical analyses.

The demographic characteristics of the respondents are presented below. The sample was relatively balanced in terms of gender, with 51.4% male and 48.6% female participants. The age distribution was as follows: 28.5% were 25 years old or younger, 29.8% were between 26 and 35, 15.6% were between 36 and 45, 19.9% were between 46 and 55, and 6.1% were over 55. Regarding educational attainment, 16.5% of participants held a high school diploma (or equivalent) or less, 17.5% had a junior college degree, 42.1% had a bachelor’s degree, 22.6% had a master’s degree, and 1.3% held a doctoral degree. In terms of marital status, 46.7% of the respondents were unmarried, while 53.3% were married. The distribution of work tenure was as follows: 22.6% of participants had worked for up to one year, 25.8% for one to three years, 15.6% for three to five years, 18.4% for five to ten years, 11.6% for ten to twenty years, and 5.9% for over twenty years. In terms of job function, administrative roles were the most common (28.5%), followed by marketing/sales (18.2%), technical/R&D (13.5%), production (8.7%), logistical services (7.6%), and finance (5.7%); other functions accounted for the remaining 17.8%. Finally, with respect to job level, general employees constituted the majority of the sample (55.2%), followed by first-line managers (e.g., team leaders, group managers) at 29.6%, middle managers at 12.7%, and senior managers at 2.5%. Overall, the sample is considered representative of the target population and suitable for the subsequent data analysis.

3.2. Variable Measurement

All constructs were measured using validated scales from established literature, rated on a 5-point Likert scale (1 = “Strongly Disagree” to 5 = “Strongly Agree”). Complete questionnaire items are provided in Appendix A.

• HPWSs: Measured using an 18-item Chinese-adapted scale across five dimensions by Su et al. (e.g., “Employee compensation is linked to individual performance”) [72].
• UPB: Assessed via a 6-item unidimensional scale by Umphress et al. (e.g., “If it would help my organization, I would exaggerate the truth about my company’s products or services to customers and clients “) [73].
• Psychological Contract Fulfillment: Evaluated using a 5-item scale (e.g., “My employer has fulfilled most recruitment promises”) [74].
• Bottom-Line Mentality: Measured with Greenbaum’s 4-item scale (e.g., “I focus exclusively on meeting performance targets”) [51].
• Moral Identity: Captured through the 5-item internalized scale by Karl et al. (e.g., “These moral characteristics define my self-concept”) [61].
• Control Variables: Following evidence that demographic factors influence organizational perceptions [75,76,77], we controlled for gender, age, education, marital status, work tenure, job category, and job grade.

4. Results

We conducted reliability and validity analyses for the five core study variables: HPWSs, UPB, psychological contract fulfillment, bottom-line mentality, and moral identity. Internal consistency reliability was assessed using Cronbach’s alpha (SPSS 26.0), while confirmatory factor analysis (Amos 24.0) was used to assess convergent validity. As shown in

Table 1. Results of reliability and convergent validity tests.

Scale	Cronbach’s α	CR	AVE
High-Performance Work System	0.896	0.966	0.610
Unethical pro-organizational behavior	0.853	0.855	0.498
Psychological contract fulfillment	0.867	0.868	0.569
Bottom-line mentality	0.856	0.857	0.599
Moral identity	0.897	0.898	0.637

, all Cronbach’s alpha coefficients exceeded 0.70, demonstrating strong internal consistency.

Convergent validity was assessed against established criteria for standardized factor loadings, composite reliability (CR > 0.70), and average variance extracted (AVE > 0.50). It is noted that convergent validity may still be considered acceptable if AVE is below 0.50, provided that the corresponding CR is above 0.60 [78]. In this study, the results confirmed strong convergent validity: all CR values surpassed 0.80 and all AVE values exceeded 0.50. These results affirm the robustness of the scales used.

To establish the discriminant validity of the measures, we conducted a confirmatory factor analysis (CFA) using AMOS 24.0. We compared the fit of our hypothesized five-factor model against four more parsimonious, alternative models. In the hypothesized model, items for HPWSs, UPB, psychological contract fulfillment, bottom-line mentality, and moral identity each loaded onto their own distinct factor. The alternative models combined conceptually related constructs: a four-factor model (combining psychological contract fulfillment and bottom-line mentality), a three-factor model (combining psychological contract fulfillment, bottom-line mentality, and UPB), a two-factor model (combining all variables except HPWSs), and a one-factor model (combining all five variables).

As presented in

Table 2. Fit Indices for Competing Measurement Models.

Model	Factor	χ2/df	RMSEA	IFI	TLI	CFI
Five-factor model	HPWSs, PCF, BLM, UPB, MI	1.389	0.029	0.971	0.969	0.971
Four-factor model	HPWSs, PCF + BLM, UPB, MI	3.324	0.072	0.826	0.814	0.825
Three-factor model	HPWSs, PCF + BLM + UPB, MI	4.216	0.085	0.755	0.743	0.754
Two-factor model	HPWSs, PCF + BLM + UPB + MI	6.127	0.107	0.609	0.590	0.607
Single-factor model	HPWSs + PCF + BLM + UPB + MI	6.482	0.111	0.580	0.561	0.579

Note: HPWSs = High-Performance Work Systems; UPB = Unethical Pro-Organizational Behavior; PCF = Psychological Contract Fulfillment; BLM = Bottom-Line Mentality; MI = Moral Identity.

, the results showed that the hypothesized five-factor model achieved an excellent fit to the data (χ²/df = 1.389, RMSEA = 0.029, CFI = 0.971, TLI = 0.969, IFI = 0.971) and was superior to all other models. This finding provides strong evidence for the discriminant validity of the five study variables, confirming they represent distinct constructs.

We further assessed discriminant validity using the Fornell-Larcker criterion. According to this criterion, the square root of a construct’s AVE must be greater than its correlation with any other construct in the model. The results are presented in

Table 3. Discriminant validity assessment.

	HPWSs	PCF	BLM	UPB	MI
HPWSs	0.781
PCF	0.109 ***	0.755
BLM	0.142 ***	0.216 ***	0.774
UPB	0.135 ***	0.135 ***	0.238 ***	0.705
MI	−0.129 ***	−0.080 *	−0.219 ***	−0.061 *	0.798

Note: Bolded numbers on the diagonal are AVE square root values for each variable. *** p < 0.001, * p < 0.05.

, which displays the inter-construct correlation matrix with the square root of the AVE values on the diagonal. The data confirm that this criterion was met for all variables. These findings provide strong, additional support for the discriminant validity of our measures. Taken together, these comprehensive psychometric assessments confirm that the measurement scales employed in this study are reliable, valid, and capture empirically distinct constructs. This robust measurement foundation provides the necessary confidence to proceed with testing the structural relationships outlined in our hypotheses.

4.1. Common Method Bias Analysis

Given that our data were self-reported and collected at a single point in time, common method bias (CMB) is a potential concern. To address this, we took both procedural and statistical precautions. Procedurally, during the survey design, we guaranteed participant anonymity and counterbalanced the order of the measurement scales to reduce response patterning.

Statistically, we conducted Harman’s single-factor test ex post. An unrotated exploratory factor analysis (EFA) was performed on all measurement items. The analysis extracted nine factors with eigenvalues greater than 1.0, which collectively accounted for 69.19% of the total variance. More importantly, the first and largest factor accounted for only 23.84% of the variance, a value well below the established 40% threshold. These results suggest that common method bias is not a significant concern in this study and does not threaten the validity of our findings.

4.2. Correlation Analysis

We conducted a Pearson correlation analysis to examine the relationships among the study variables. The means, standard deviations, and inter-correlations are presented in

Table 4. The mean, standard deviation and correlation coefficient of each variable.

Variable	1	2	3	4	5	6	7	8	9	10	11	12
1. HPWSs	1
2. UPB	0.368 **	1
3. PCF	0.234 **	0.258 **	1
4. BLM	0.298 **	0.423 **	0.288 **	1
5. MI	−0.291 **	−0.195 **	−0.124 **	−0.332 **	1
6. Gender	−0.061	−0.139 **	−0.050	−0.055	0.118 **	1
7. Age	−0.126 **	0.043	−0.039	−0.017	−0.019	−0.259 **	1
8. Education	0.100 *	0.028	0.042	0.017	−0.020	0.274 **	−0.556 **	1
9. Marital status	−0.127 **	0.017	−0.033	−0.011	0.014	−0.259 **	0.707 **	−0.580 **	1
10. Work tenure	0.024	0.158 **	−0.002	0.078	−0.113 *	−0.303 **	0.737 **	−0.392 **	0.595 **	1
11. Job category	0.063	0.046	−0.042	0.070	0.059	−0.014	−0.007	−0.065	−0.019	0.005	1
12. Job grade	−0.041	0.065	−0.101 *	−0.035	−0.050	−0.100 *	0.569 **	−0.206 **	0.399 **	0.561 **	−0.028	1
Mean	3.407	2.246	2.797	2.529	3.332	1.490	2.450	2.750	1.530	2.880	3.530	1.630
S.D.	0.601	0.696	0.916	0.920	0.861	0.500	1.260	1.025	0.499	1.528	2.202	0.801

Note: N = 473; ** p < 0.01, * p < 0.05.

. The correlational matrix offers a preliminary but telling glimpse into the proposed paradoxical model. As hypothesized, HPWSs were significantly and positively correlated with both psychological contract fulfillment ($r=0.23,p<0.01$) and bottom-line mentality ($r=0.30,p<0.01$). This initial evidence provides the first empirical signal of the paradox at play, demonstrating that HPWSs implementation is associated with the antecedents of both the reciprocity-driven path (psychological contract fulfillment) and the pressure-driven path (bottom-line mentality). In turn, both psychological contract fulfillment ($r=0.26,p<0.01$) and bottom-line mentality ($r=0.42,p<0.01$) were positively associated with unethical pro-organizational behavior (UPB). Conversely, moral identity demonstrated significant negative correlations with psychological contract fulfillment ($r=-0.12,p<0.01$), bottom-line mentality ($r=-0.33,p<0.01$), and UPB ($r=-0.20,p<0.01$), hinting at its potential protective function against performance-driven pressures. These results are consistent with our hypotheses and provide preliminary support for our proposed theoretical framework (Table 4). The mean, standard deviation and correlation coefficient of each variable are shown.

4.3. Hypothesis Testing

4.3.1. Analysis of Mediating Effects

To test our mediation hypotheses, we employed a manifest-variable path model in AMOS to simultaneously examine the proposed mediating roles of psychological contract fulfillment and bottom-line mentality. The model demonstrated a strong fit to the data (χ²/df = 1.594, RMSEA = 0.035, IFI = 0.961, TLI = 0.957, CFI = 0.961), indicating that the proposed theoretical structure accurately represents the relationships in the data.

The subsequent analysis unpacks the two distinct mechanisms linking HPWSs to UPB, beginning with the reciprocity-driven path. As detailed in Figure 2 and

Table 5. Standardized Path Coefficients for the Direct Effects.

Path	Estimate	S.E.	C.R.	p
HPWSs→PCF	0.325	0.103	5.395	***
HPWSs→BLM	0.405	0.082	6.491	***
PCF→UPB	0.155	0.045	3.039	**
BLM→UPB	0.454	0.063	8.235	***

Note: *** p < 0.001, ** p < 0.01.

, the path analysis supported the hypotheses that HPWSs had a significant positive effect on psychological contract fulfillment ($\beta =0.33,p<0.001$), which in turn was a positive predictor of UPB ($\beta =0.16,p<0.01$), supporting H1a and H1b. This initial result suggests that while HPWSs succeed in fostering a sense of fulfilled obligations, this very success is associated with a higher propensity for misconduct.

Next, we examined the pressure-driven path. Similarly, HPWSs significantly influenced bottom-line mentality ($\beta =0.41,p<0.01$), which subsequently had a strong positive effect on UPB ($\beta =0.45,p<0.001$), providing support for H2a and H2b. This highlights a more direct and pernicious consequence, wherein the performance pressures of HPWSs cultivate a results-at-all-costs mindset that strongly predicts unethical behavior.

To formally test the significance of these two indirect pathways, we conducted a bootstrapping analysis in AMOS using 5000 resamples. The results, presented in

Table 6. Results of Bootstrapping Analysis for Indirect Effects.

Path	Estimate	S.E.	Bias-Corrected 95%CI			Percentile 95%CI
			Lower	Upper	Lower	Lower	Upper	p
HPWSs→PCF→UPB	0.050	0.050	0.014	0.101	0.004	0.013	0.100	0.005
HPWSs→BLM→UPB	0.063	0.024	0.023	0.117	0.003	0.019	0.112	0.005

, provide robust support for our mediation hypotheses. First, the indirect effect of HPWSs on UPB through psychological contract fulfillment was estimated at 0.050. The 95% bias-corrected confidence interval [0.014, 0.101] and the percentile confidence interval [0.013, 0.100] both excluded zero. This indicates a statistically significant mediating effect, thus supporting H1c. Second, the indirect effect of HPWSs on UPB through bottom-line mentality was estimated at 0.063. The 95% bias-corrected confidence interval [0.023, 0.117] and percentile confidence interval [0.019, 0.112] also excluded zero. This result confirms a significant mediating role for bottom-line mentality, supporting H2c.

4.3.2. Moderating Effect Test

This section examines the proposed role of moral identity as a crucial buffer against the adverse effects of the two mediating variables. First, to test the moderating effect of moral identity on the reciprocity-driven path, a hierarchical regression analysis was employed. An interaction term was created using the centered variables for psychological contract fulfillment and moral identity were used to create an interaction term. As shown in

Table 7. Test results of the moderating effect of moral identity (psychological contract fulfillment as a mediating variable).

Variables	UPB
	Model 1	Model 2	Model 3
Control Variables
Gender	−0.122 *	−0.094 *	−0.084
Age	−0.093	−0.082	−0.081
Education	0.082	0.066	0.059
Marital status	−0.050	−0.034	−0.028
Work tenure (years)	0.249 **	0.201 **	0.185 **
Job category	0.047	0.066	0.052
Job grade	0.005	0.036	0.044
Mediating variables
Psychological contract fulfillment		0.236 ***	0.224 ***
Moderator variable
Moral identity		−0.133 **	−0.135 **
Interaction term
Psychological contract fulfillment × Moral identity			−0.147 **
R2	0.056	0.135	0.156
DR2		0.079	0.137
F	3.943 ***	8.002 ***	8.510 ***
DF		21.021 ***	11.452 **

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

, the analysis revealed that the interaction term was a significant negative predictor of unethical pro-organizational behavior ($\beta =-0.15,p<0.01$), providing support for Hypothesis 3.

To further probe this interaction, a simple slopes analysis was performed, and the results were plotted (Figure 3). The analysis compellingly illustrated the buffering effect that for employees with high moral identity (+1 SD), the positive relationship between psychological contract fulfillment and UPB was weak and not statistically significant (simple slope $=0.06$, $p>0.05$). In contrast, for employees with low moral identity (−1 SD), this relationship was strong and highly significant (simple slope $=0.028$, $p<0.001$). These findings confirm that a strong moral identity weakens—and in this case, effectively neutralizes—the positive association between psychological contract fulfillment and UPB.

Subsequently, we turned our attention to the pressure-driven path, examining the moderating role of moral identity on the relationship between bottom-line mentality and UPB was examined. A hierarchical regression analysis was conducted using an interaction term created from the centered variables. As reported in

Table 8. Test results of the moderating effect of moral identity (bottom-line mentality as a mediating variable).

Variables	UPB
	Model 1	Model 2	Model 3
Control Variables
Gender	−0.122 *	−0.103 *	−0.095
Age	−0.093	−0.048	−0.022
Education	0.082	0.071	0.084
Marital status	−0.050	−0.038	−0.034
Work tenure (years)	0.249 **	0.157 *	0.142 *
Job category	0.047	0.023	0.017
Job grade	0.005	0.036	0.029
Mediating variable
Bottom line mentality		0.391 ***	0.359 ***
Moderating variable
Moral identity		−0.033	−0.052
Interaction term
Bottom-line mentality * Moral identity			−0.153 ***
R2	0.056	0.214	0.236
DR2		0.158	0.022
F	3.943 ***	14.005 ***	14.267 ***
DF		46.521 ***	13.275 ***

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

, the interaction term was found to be a significant negative predictor of UPB ($\beta =-0.15,p<0.001$), providing support for Hypothesis 4.

To further interpret this finding, a simple slopes analysis was performed, with the results plotted in Figure 4. The analysis revealed that the positive relationship between bottom-line mentality and UPB was significant under both conditions, but its magnitude differed substantially. For employees with high moral identity (+1 SD), the relationship was significant but relatively contained (simple slope $=0.16$, $p<0.01$). In contrast, for employees with low moral identity (−1 SD), the relationship was substantially stronger (simple slope $=0.39$, $p<0.001$). These results confirm that moral identity significantly attenuates the positive effect of a bottom-line mentality on an employee’s engagement in UPB, thus supporting H4. This finding is particularly consequential, as it highlights moral identity’s capacity to mitigate the most direct and potent pathway leading from HPWSs to unethical conduct.

4.3.3. Tests with Moderating Mediating Effects

Building upon the previous moderation analyses, this section investigates whether the buffering effect of moral identity extends to the entire indirect pathways from HPWSs to UPB. We sought to determine if the strength of these indirect effects was conditional on employee moral identity. This was accomplished using a bootstrapping procedure in SPSS 26.0 with 5000 resamples to estimate the conditional indirect effects at different levels of the moderator. The results of this analysis are presented in

Table 9. The moderating mediating effect test.

Mediating Variable	Moral Identity	Effect Value	Boot SE	Boot LLCI	Boot ULCI
Psychological contract fulfillment	Low (M − 1SD)	0.044	0.020	0.009	0.088
	Medium (M)	0.025	0.013	0.001	0.052
	High (M + 1SD)	0.009	0.016	−0.023	0.041
Bottom line mentality	Low (M − 1SD)	0.124	0.034	0.064	0.196
	Medium (M)	0.091	0.024	0.048	0.140
	High (M + 1SD)	0.063	0.025	0.018	0.114

.

The investigation first explored the reciprocity-driven path through psychological contract fulfillment (H5). For employees with low moral identity (−1 SD), a significant indirect effect from HPWSs to UPB was found (Effect = 0.044, 95% CI [0.009, 0.088]). In contrast, for those with high moral identity (+1 SD), this indirect effect was not statistically significant (Effect = 0.009, 95% CI [−0.023, 0.041]). While this pattern suggests potential moderation, the definitive test lies in the Index of Moderated Mediation. In this case, the index was non-significant (Index = −0.019, 95% CI [−0.050, 0.006]). Because the confidence interval for the index contains zero, it is concluded that the difference between the two conditional indirect effects is not statistically significant. Therefore, the hypothesis that moral identity moderates the indirect effect via psychological contract fulfillment was not supported; thus, H5 is rejected.

In contrast, a different pattern of results emerged for the pressure-driven pathway through bottom-line mentality (H6). At low levels of moral identity (−1 SD), a significant and strong indirect effect was observed (Effect = 0.124, 95% CI [0.064, 0.196]). Although the indirect effect was also significant at high levels of moral identity (+1 SD), it was substantially diminished (Effect = 0.063, 95% CI [0.018, 0.114]). Critically, the significance of the attenuation was confirmed by a statistically significant Index of Moderated Mediation. The index was found to be statistically significant (Index = −0.033, 95% CI [−0.074, −0.002]), as its confidence interval does not contain zero. This finding directly supports the assertion that moral identity significantly weakens the indirect effect channeled through bottom-line mentality, confirming the particularly consequential nature of its buffering role on this detrimental pathway. Therefore, Hypothesis 6 was supported.

5. Discussion

The findings of this study reveal a significant paradox: HPWSs, traditionally viewed as catalysts for organizational excellence, also operate as systemic mechanisms that produce inherent, contradictory outcomes. Our research demonstrates that these systems do not merely influence individual choices but actively shape an organizational context where UPB becomes a rational, adaptive response. This conclusion challenges the decontextualized, predominantly positive perception of HPWSs. Our dual-pathway model details how this systemic conditioning unfolds. The first pathway operates through a logic of distorted social exchange; by fulfilling the psychological contract within a high-pressure environment, the system inculcates a transactional form of obligation, leading employees to construe UPB as a legitimate means of reciprocity. The second pathway operates through system-induced cognitive narrowing; by fostering a bottom-line mentality, the system encourages an instrumental rationality that frames ethical norms as negotiable costs in the relentless pursuit of performance. By empirically validating these two parallel pathways, our findings extend the foundational research that identified a single, reciprocity-based mechanism, presenting a more complex and systemic picture of how the HPWS paradox functions in practice.

Furthermore, this research illuminates the asymmetrical power of individual character in resisting these systemic forces. While a strong moral identity proved effective at attenuating the indirect effect operating through a bottom-line mentality, it was notably ineffective at buffering the path through psychological contract fulfillment. This asymmetry suggests that while an individual’s moral compass can effectively counter an internal cognitive state of narrow, results-only focus, it is substantially less effective against the powerful situational pressures for social exchange that the organizational system itself generates. Thus, our findings point to an uncomfortable conclusion: UPB should be understood not as a series of isolated individual moral failures, but as a predictable, systemic outcome of well-intentioned management architectures that inadvertently prioritize performance to the detriment of ethical conduct and psychological safety. This finding provides a finer-grained insight than previously available, moving beyond simply confirming that moral identity acts as a buffer to explain how its efficacy differs across competing systemic pressures.

5.1. Theoretical Contributions

5.1.1. Illuminating the Dark Side of HPWSs

This study contributes to the literature by identifying a significant positive relationship between HPWSs and employees’ engagement in UPB. While most prior research has focused on the positive outcomes of HPWSs, such as improved performance and well-being, their potential negative consequences have been largely overlooked. By demonstrating that an intense focus on performance goals can motivate employees to prioritize ends over ethical means, this research reveals how the instrumental logic embedded in HPWSs can erode ethical norms. This finding challenges the one-sided, positive perception of HPWSs, provides a new perspective on their complex impact, and underscores the need for scholars and practitioners to critically examine the inherent ethical risks of these systems.

Furthermore, our findings contribute a significant sense of urgency to this critical examination by connecting the systemic antecedents of UPB to its documented downstream consequences. Our research reveals how a prevalent HRM system can systematically foster UPB. This is particularly alarming when viewed in light of recent findings on UPB’s long-term effects. For example, research has shown that engaging in UPB can lead to a “moral slippery slope,” where individuals become morally disengaged and more likely to commit further, even self-serving, unethical acts [79]. Additionally, if such behaviors are enacted or tolerated by leaders, they can be learned and replicated by employees through social learning, creating a pervasive unethical climate [80]. Therefore, our study suggests that intervening at the systemic level of HPWSs design is not just about preventing isolated incidents of UPB; it is about preventing the very catalyst for a potential downward spiral of organizational ethics. The identification of specific, competing mechanisms allows our research to move beyond simply labeling HPWSs’ negative consequences as a ‘dark side’, offering instead a more granular explanation of their systemic and paradoxical nature. The broader theoretical implication, therefore, is the suggestion that HPWSs function as inherently paradoxical systems. This perspective shifts the central research question from a binary assessment of whether these systems are “beneficial’ or “harmful” to a more critical inquiry into how organizations can manage the intrinsic performance-ethics tension they inevitably create.

5.1.2. A Dual-Pathway Mediation Model

Drawing on the JD-R model, this study develops and empirically validates a novel dual-pathway model explaining how HPWSs influence UPB. This model significantly extends prior research that conceptualized the HPWS-UPB link through a single, reciprocity-based mechanism. We identify a “resource-gain” path, where HPWSs enhance psychological contract fulfillment, and a “demand-imposition” path, where they foster a bottom-line mentality. Our findings confirm that these two mechanisms operate in parallel: HPWSs provide resources that motivate social exchange while simultaneously imposing demands that induce cognitive narrowing, with both pathways ultimately promoting UPB. This integrated framework not only extends the application of the JD-R model but also provides a more holistic and complex account of the psychological mechanisms linking HPWSs to employee behavior. The key theoretical contribution of this model, therefore, is that it provides a blueprint for how a single, well-intentioned management system can produce unintended and contradictory outcomes. It offers an integrative mechanism to explain the tensions inherent in many organizational initiatives, thereby moving beyond a binary perspective of beneficial versus detrimental effects.

5.1.3. Beyond Altruism: The Instrumental Motives of UPB

This research reconceptualizes the motivational underpinnings of UPB by challenging the predominantly altruistic view. Whereas existing theories often attribute UPB to organizational loyalty or reciprocity, our findings reveal its instrumental and self-interested nature. Specifically, the validation of the pressure-driven pathway (via a bottom-line mentality) provides robust empirical evidence for these instrumental motives. We demonstrate that UPB is not purely pro-social but is also a strategic behavior driven by instrumental motives, including “active gain” (to secure rewards by demonstrating commitment) and “passive risk avoidance” (to prevent negative outcomes associated with performance failure). This insight deepens our understanding of the complex motivations for UPB and provides a strong theoretical foundation for future research to analyze this behavior through the lens of a “motivational duality” between altruism and self-interest. More fundamentally, this finding shifts the locus of explanation for certain unethical behaviors from an individual’s moral disposition to their rational adaptation to systemic pressures. It suggests that organizational systems can create conditions where unethical conduct becomes a seemingly logical, strategic choice, thereby challenging theoretical models that situate the primary responsibility for ethical decision-making exclusively within the individual.

5.2. Practical Implications

Our findings offer several practical implications for organizations. The central challenge for managers is to harness the performance-enhancing benefits of HPWSs while mitigating the ethical risks they can inadvertently create. To prevent the instrumental rationality embedded in these systems from eroding ethical standards, organizations must proactively balance performance imperatives with robust ethical safeguards. To this end, we offer the following recommendations.

5.2.1. Proactively Re-Architecting HPWSs to Enhance Psychological Safety

To mitigate the inherent “performance-versus-ethics” conflict that HPWSs can create, the very architecture of these systems must be proactively optimized to prevent the negative consequences of an overemphasis on results, which begins with restructuring performance management to embed ethical constraints and foster psychological safety. Instead of relying solely on traditional KPIs, organizations should adopt a two-dimensional appraisal system that explicitly evaluates both performance outcomes and ethical conduct, thereby countering the tendency to prioritize results at any cost. To buffer employees from the intense pressure that leads to resource depletion and erodes psychological safety, firms should implement not only flexible work arrangements but also mandatory leave policies that signal a genuine commitment to well-being. This should be paired with a goal-setting process that emphasizes phased, achievable targets over singular, high-stakes outcomes, which discourages an “ends-justify-the-means” mindset. Finally, by aligning long-term incentives such as promotions and equity with sustained ethical performance—rather than solely with short-term bonuses—organizations can effectively curb instrumental rationality and cultivate a culture where ethical conduct is integral to success.

5.2.2. Unmasking “Pseudo-Altruism” to Foster a Genuinely Ethical Climate

To foster a genuinely healthy ethical climate, organizations must learn to unmask the “pseudo-altruism” of UPB, which often conceals instrumental motives behind a veneer of pro-organizational commitment. This requires building a robust ethical risk prevention system. One key component is an early warning system that uses behavioral data to create an ethical risk index, enabling the identification of abnormal patterns—such as policy shortcuts justified as “for the good of the company”—and allowing for timely intervention. At the same time, leaders must explicitly define and communicate the boundaries between legitimate organizational contributions and unethical transgressions. This can be reinforced by regularly assessing the alignment between employee and organizational expectations to prevent commitment from being weaponized as a justification for misconduct. Finally, leaders must actively dismantle a “performance-at-all-costs” culture. This involves openly criticizing results-only thinking, reflecting ethical priorities in key decisions, and restructuring performance appraisals to reduce zero-sum competition and increase the emphasis on collaborative goals. Together, these actions reduce the intense individual pressures that undermine psychological safety and fuel UPB.

5.2.3. Fostering Ethical Identity as an Internal Defense Mechanism

Organizations can fortify employees’ moral identity, transforming it into a robust internal mechanism for self-regulation against unethical conduct. This requires ethics training that is grounded in specific cultural and organizational contexts. Rather than relying on abstract rules, such training should leverage principles of character-based ethics, whether drawn from Eastern philosophies like the Confucian principle of yi-li (balancing righteousness with profitability) and the ideal of shendu (maintaining integrity while unobserved), or from Western traditions of virtue ethics that emphasize the cultivation of moral character. These principles can be brought to life through facilitated discussions of real-world ethical dilemmas from within the company, prompting employees to critically analyze the underlying motives and consequences to deepen their own ethical self-awareness. This approach is also a cornerstone of modern practices like values-based leadership, where managers are expected to consistently model and champion core ethical values. Crucially, these learned values must be systematically reinforced by structurally aligning career progression, promotions, and significant rewards with demonstrated ethical conduct. Doing so creates an incentive structure where ethical behavior and professional advancement are mutually supportive, not in conflict. Finally, this ethical framework must be woven into the very fabric of the organizational culture. Leaders should champion the philosophy that “Ethics is Competitiveness,” embedding this into the company’s mission and daily operations. By creating psychologically safe forums for employees to share their own ethical challenges, organizations can normalize ethical discourse and foster a collective commitment to integrity that transcends mere compliance. Ultimately, the goal is to shift the organizational consensus away from a tacit acceptance of UPB, thereby empowering observers to voice concerns rather than maintain a silence born from a misguided sense of loyalty [81].

5.2.4. Beyond the Firm: A Call for Systemic Governance and External Scrutiny

Finally, our research suggests that mitigating the ethical risks inherent in performance-driven systems cannot be left to organizations alone, as they operate within a larger economic ecosystem where intense market pressures often incentivize a race to the bottom. In such an environment, ethical considerations are frequently sidelined for competitive advantage, which necessitates a broader, systemic governance perspective. For policymakers and regulatory bodies, it is therefore crucial to design regulations that hold organizations accountable not merely for their financial outcomes, but also for their ethical climate and broader societal impact. Such measures could include mandatory, audited reporting on employee well-being and psychological safety metrics, alongside stronger whistleblower protections that shield employees who expose systemic pressures to act unethically.

Concurrently, for civil society, institutional investors, and the media, our study underscores the profound importance of external scrutiny. By demanding greater transparency and systematically incorporating ethical performance into investment criteria, such as through robust ESG ratings, these external stakeholders can generate powerful incentives for organizations to transcend a short-term, bottom-line focus. Ultimately, curbing system-induced UPB must be understood as a collective responsibility, requiring coordinated efforts from both within and outside the corporate structure to fundamentally reshape the very definition of “high performance”.

5.3. Research Limitations and Prospects

While this study offers several contributions, its limitations should be acknowledged, which in turn provide promising avenues for future research.

First, this study relies on cross-sectional data collected through a single instrument where all variables were self-reported by the same respondent, a single-source, single-point-in-time design that introduces several potential limitations. Although a Harman’s single-factor test was conducted as a preliminary diagnostic, we acknowledge that this test is considered somewhat outdated and its possibility of common method bias cannot be entirely ruled out. Consequently, the cross-sectional nature of the data limits our ability to establish stronger inferences regarding causal relationships. Furthermore, relying exclusively on employee perspectives for sensitive variables such as unethical behavior may reduce the robustness of the findings due to potential social desirability bias.

Future research would benefit from employing multi-source data collection (e.g., supervisor or peer ratings of UPB) and a longitudinal design to more accurately capture the dynamic interplay between HPWSs and employee behavior over time. Additionally, employing experimental designs, such as situational simulations, could further strengthen causal inferences regarding the relationships among the variables, while future studies using multi-level modeling could also account for the potential variance from the nested structure of our multi-organization sample.

A second limitation stems from the study’s exclusive focus on UPB as the behavioral outcome. From a paradox theory standpoint, this focus captures only one facet of the system’s full range of consequences. Consequently, while our dual-pathway model delineates the antecedents of a detrimental behavior, its explanatory scope does not extend to the concurrent processes that may yield constructive ones. Future inquiries could therefore advance a more symmetrical conceptualization by examining a positive dependent variable alongside UPB. For example, a compelling avenue would be to investigate whether the resource-gain pathway, mediated by psychological contract fulfillment, not only mitigates unethical conduct but also actively cultivates behaviors such as Organizational Citizenship Behavior (OCB). Such a model would reveal HPWSs as a system that concurrently fosters OCB through a resource-providing mechanism while driving UPB via a demand-imposing one. Investigating these divergent outcomes in tandem would yield a more holistic appreciation of the HPWS paradox, thereby offering more nuanced guidance for managerial intervention.

Third, while our findings highlight the instrumental nature of UPB, we did not explore the complex interplay between employees’ self-interested and pro-social motives. It is likely that these motivations co-exist and that their relative salience is dynamic and context-dependent. For instance, an employee’s altruistic motives might dominate during an organizational crisis, whereas self-serving motives may become more prominent in a high-stakes promotion context. Future research could utilize qualitative methods, such as in-depth interviews, to deconstruct the complexity and situational dependence of UPB motivations, which would provide a richer understanding of the phenomenon and help organizations develop more nuanced and differentiated intervention strategies.

Finally, concerning our theoretical framework, we acknowledge that the JD-R model, while useful for deconstructing the dual pathways of HPWSs, is not a traditional business ethics theory. The model’s primary focus on well-being and motivation means its direct explanatory power for complex ethical choices like UPB has its limits. Our study represents an attempt to extend this model into a new domain; however, we concede that a more explicit business ethics framework could provide a richer interpretation. Future research could greatly benefit from integrating our findings with theories that distinguish between “business frames” and “ethical frames” [82] to explore how the demand and resource paths of HPWSs respectively activate these competing cognitive frameworks.

6. Conclusions

This research reframes unethical pro-organizational behavior, moving beyond a narrative of isolated, individual moral failures to recognize it as a systemic outcome of performance-oriented management architectures. Our findings compellingly demonstrate that UPB can be a predictable result of systems that inadvertently prioritize performance over ethical conduct.

Consequently, the primary implication is not a call for managers to simply be more ethical, but for them to become more critical architects of the systems they oversee. The task shifts from policing individual character to scrutinizing the unintended ethical consequences embedded within performance appraisals, incentive structures, and cultural norms. Proactively fostering an environment that supports buffers like employee moral identity is therefore not a soft initiative, but a core design principle for risk mitigation. Ultimately, this study posits that creating a resilient organization—one that achieves both high performance and a sustainable ethical climate—depends less on sanctioning individual misconduct and more on the deliberate, thoughtful design of the systems that shape behavior in the first place.