Promotive and prohibitive ethical voice in groups: the effect of faultlines and role ambiguity

The study addresses why and when groups exhibit ethical voice—defined as speaking up about violations of ethical norms and making suggestions to sustain ethical norms—by distinguishing promotive (what should be done) and prohibitive (what should not be done) ethical voice. Drawing on regulatory focus theory (promotion vs. prevention focus) integrated with social identity theory (categorization and identity processes), the authors argue that intra-group structures and identity dynamics shape the emergence of ethical voice at the group level. They propose that activated demographic faultlines (perceived subgroup divides based on age and gender) undermine collective ethical voice by fostering psychological boundaries and out-subgroup hostility, thereby narrowing the scope of voice to subgroup interests. They further posit that group role ambiguity (uncertainty about role expectations) weakens the positive influence of both promotive and prohibitive ethical voice on group-level outcomes—organizational citizenship behaviors and task performance—because uncertainty diverts attention from identity enhancement and ethical voice elaboration toward uncertainty reduction. The purpose is to identify faultlines as antecedents and role ambiguity as a boundary condition for group ethical voice, clarifying mechanisms and conditions under which ethical voice contributes to group functioning.

• Group ethical voice: Based on regulatory focus theory, group ethical voice is conceptualized in two forms—promotive (suggestions to improve ethical functioning) and prohibitive (concerns to prevent unethical or harmful actions). In group communication, promotive voice uses a positive tone highlighting ethical opportunities (e.g., emphasizing equality), whereas prohibitive voice uses a preventive tone highlighting potential violations (e.g., avoiding discrimination).
• Group faultlines: Faultlines are subgroup divides arising from the alignment of multiple demographic attributes (here, age and gender), creating within-subgroup similarity and between-subgroup differences. When activated, subgroup boundaries become salient, increasing polarization, discrimination, and behavioral disintegration, which can impair information exchange and collective processes.
• Linking faultlines to ethical voice: From a social identity perspective, salient subgroups shift members’ alignment from group norms to subgroup norms. Out-subgroup voices are devalued or resisted, constraining the pool of ethical ideas available to the group. Thus, stronger faultlines should reduce both promotive and prohibitive ethical voice.
• Moderating role of group role ambiguity: Uncertainty (role ambiguity) undermines group identification and the motivation to engage with ethical voice. It is theorized to dampen the benefits of both types of ethical voice for GOCB and task performance. During high ambiguity, uncertainty reduction motives supersede identity-enhancement motives, reducing attention to promotive voice benefits; and the negativity alignment between ambiguity and prohibitive voice can intensify negative affect and disengagement. Hypotheses: H1: Group faultlines are negatively related to group promotive ethical voice. H2: Group faultlines are negatively related to group prohibitive ethical voice. H3a/b: Group role ambiguity weakens the positive relationships between group promotive ethical voice and (a) GOCB, (b) group task performance. H4a/b: Group role ambiguity weakens the positive relationships between group prohibitive ethical voice and (a) GOCB, (b) group task performance.

Design and setting: Multi-source, multi-wave field study in three Chinese organizations. Surveys administered online with confidentiality assured. Sampling and procedure: At Time 1 (T1), 779 employees were invited to report perceived age and gender subgroups and provide demographics. At Time 2 (T2; 2 weeks later), participants reported group promotive and prohibitive ethical voice. At Time 3 (T3; 2 weeks after T2), participants reported group role ambiguity. Separately, 77 direct leaders were invited to evaluate followers’ GOCB and task performance. Employee and leader responses were matched via last four digits of phone numbers. Final sample: 632 employees (81% response rate) nested in 62 groups (62 leaders; 81% response rate). Mean group size = 10.19 (range 7–19). Mean age = 36.43 years; age distribution: <30 (23.1%), 30–39 (50.8%), 40–49 (21.7%), >49 (4.4%). Females = 44.9%. Measures (6-point scales; back-translation used):

• Group promotive ethical voice (3 items; Chen & Treviño, 2022; α=0.81). Example: emphasis on opportunities to care for others. Aggregation supported: Rwg=0.73; ICC(1)=0.19; ICC(2)=0.71.
• Group prohibitive ethical voice (3 items; Chen & Treviño, 2022; α=0.85). Example: emphasis on stopping harmful behavior. Aggregation supported: Rwg=0.71; ICC(1)=0.22; ICC(2)=0.75.
• Activated group faultlines (age and gender): Faultline strength computed using Shaw’s (2004) FLS algorithm for age and gender; perceived subgroups measured with a 4-item scale (Jehn & Bezrukova, 2010); activation operationalized as the product of FLS strength and perceived subgroup scores, averaged to form activated identity faultlines.
• Group role ambiguity (4 items; Cicero et al., 2010; α=0.82; Rwg=0.84; ICC(1)=0.11; ICC(2)=0.54). Reverse-coded as needed so higher scores reflect greater ambiguity when used in analyses.
• Group-level organizational citizenship behaviors (GOCB; 5 items; Euwema et al., 2007; α=0.85; Rwg=0.83; ICC(1)=0.25; ICC(2)=0.78). Leader-rated for each member and aggregated to the group level (mean).
• Group task performance (3 items; Lam et al., 2002; α=0.82; Rwg=0.90; ICC(1)=0.12; ICC(2)=0.57). Leader-rated for each member and aggregated to the group level (mean). Control variables: Group size, age diversity, gender diversity. Analytic strategy: Confirmatory factor analysis supported distinct constructs (five-factor model: χ²/df=1.94; CFI=0.98; TLI=0.97; RMSEA=0.04; SRMR=0.04) over alternatives. ANOVAs indicated no significant between-organization variance, so organization was not controlled. Hypotheses were tested via hierarchical regression at the group level, examining main effects of faultlines on ethical voice and interaction effects of role ambiguity with each ethical voice form on GOCB and task performance. Simple slope analyses probed significant interactions.

• Faultlines → ethical voice: Activated group faultlines negatively predicted both forms of group ethical voice.
  • Promotive ethical voice: β = -0.59, p < 0.01; ΔR² = 0.13; ΔF = 8.64, p < 0.01.
  • Prohibitive ethical voice: β = -0.55, p < 0.05; ΔR² = 0.09; ΔF = 6.18, p < 0.05.
• Moderation by group role ambiguity on promotive ethical voice effects:
  • Interaction (role ambiguity × promotive voice):
    • GOCB: β = -0.47, p < 0.05.
    • Group task performance: β = -0.37, p < 0.01.
  • Simple slopes: At low role ambiguity, promotive voice positively related to outcomes (GOCB: β = 0.34, p < 0.01; task performance: β = 0.23, p < 0.01). At high role ambiguity, effects were non-significant (GOCB: β = 0.02, p > 0.05; task performance: β = -0.02, p > 0.05).
• Moderation by group role ambiguity on prohibitive ethical voice effects:
  • Interaction (role ambiguity × prohibitive voice):
    • GOCB: β = -0.73, p < 0.05.
    • Group task performance: β = -0.47, p < 0.05.
  • Simple slopes: For task performance, positive relation at low role ambiguity (β = 0.24, p < 0.01), non-significant at high ambiguity (β = -0.08, p > 0.05). For GOCB, prohibitive voice showed non-significant relations at both ambiguity levels.
• Measurement model fit was strong (five-factor CFA: χ²/df=1.94; CFI=0.98; TLI=0.97; RMSEA=0.04; SRMR=0.04). Between-organization effects were non-significant for focal variables.

The findings support the integrated regulatory focus–social identity framework. Activated demographic faultlines foster salient subgroup identities and psychological boundaries, reducing the likelihood that members express both promotive and prohibitive ethical voice to the group as a whole. This addresses the research question of why group ethical voice may fail to emerge: subgroup categorization leads members to privilege in-subgroup norms and devalue out-subgroup input, constraining collective ethical discourse. Furthermore, group role ambiguity—capturing uncertainty about roles—diminishes the positive impact of ethical voice on group outcomes. Under high ambiguity, members prioritize uncertainty reduction over identity enhancement and ethical information elaboration, attenuating the benefits of promotive voice for GOCB and task performance, and of prohibitive voice for task performance. These results highlight the significance of intra-group structure (faultlines) and uncertainty (role ambiguity) as key mechanisms and boundary conditions shaping the effectiveness of ethical voice, extending prior work focused on individual upward voice or macro-level uncertainty. The results also reconcile mixed prior findings by showing that ethical voice is not uniformly beneficial; its effects depend on the certainty of the group context.

This study identifies activated demographic faultlines as a barrier to the emergence of both promotive and prohibitive group ethical voice and shows that group role ambiguity weakens the positive relationships between ethical voice and group outcomes (GOCB and task performance). The research advances theory by integrating regulatory focus with social identity perspectives, demonstrating how subgroup categorization and uncertainty processes shape collective ethical voice and its effectiveness. Practically, leaders should manage group composition to reduce faultline activation (e.g., balance age/gender alignments, foster cross-subgroup relations, emphasize common in-group identity) and minimize role ambiguity through clear job descriptions, timely role clarification, and proactive communication—thereby enabling ethical voice to translate into better group functioning. Future research could generalize across cultures, incorporate additional social identity variables (e.g., leader group prototypicality), and use multi-method designs to deepen understanding of antecedents and moderators of group ethical voice.

• Potential common method variance for variables reported by the same source (faultlines perceptions and ethical voice), partially mitigated by temporal separation; future work should employ multiple sources/methods.
• Cultural context: Data from Chinese organizations with high uncertainty avoidance may limit generalizability; replication in lower uncertainty-avoidance contexts is needed.
• Model scope: Focused on one antecedent (faultlines) and one moderator (role ambiguity); other social identity-relevant factors (e.g., leader group prototypicality) may further explain ethical voice dynamics.