When and why voice to higher-up? Declaring the psychological mechanisms of subordinate's voice behavior in the public sector

Summary

This study examines the psychological mechanisms underlying voice behavior among frontline public sector employees in China, with particular focus on how supervisor support, public service motivation, and supervisor-subordinate guanxi (SSG) interact to predict whether subordinates speak up to higher-ups. Drawing on Leader-Member Exchange (LMX) theory, the authors propose that perceived supervisor support positively influences subordinate voice behavior, that public service motivation mediates this relationship, and that the closeness of the supervisor-subordinate guanxi relationship moderates the overall pathway. A longitudinal survey of 136 grassroots administrative law enforcement officers across 12 teams in a Chinese county was conducted at two time points six weeks apart. Structural equation modelling via Mplus 7.4 and Bayesian estimation were employed. Results confirmed that supervisor support significantly predicts voice behavior and that public service motivation serves as a partial mediator. The SSG moderation effect was also supported, with high-guanxi conditions amplifying the link between supervisor support and voice behavior. The study concludes that both formal workplace support and informal relational ties function as complementary mechanisms activating proactive voice in hierarchical public organizations, and calls for greater attention to non-role-based motivational dynamics in public administration research.

Strengths and limitations

Strengths: The longitudinal two-wave design partially addresses common method bias by separating the measurement of predictors and outcomes across time points. The use of Bayesian estimation with 30,000 iterations adds robustness to the coefficient estimates. The theoretical integration of LMX, PSM, and guanxi constructs is coherent and fills a recognised gap in public sector voice behavior research. Limitations: The final longitudinal sample of 136 respondents drawn from a single Chinese county is small and geographically restricted, severely limiting generalisability. The study does not adequately account for multilevel nesting effects — 136 individuals across 12 teams — which risks inflated standard errors or misattributed effects. Three of the four key variables were collected at the same time point (Time 1), reintroducing common method variance concerns despite the longitudinal framing. The partial mediation finding (hypothesis 2 only 'partly supported') is underexplained and warrants further theoretical elaboration. The SSG scale used (Chen and Peng, 2008) was originally developed for a different construct domain (university student internet use and academic performance), raising questions about construct validity for workplace guanxi measurement. The study is conducted exclusively in China, limiting cross-cultural applicability despite claims of broader relevance. No independent measure of team-level effects or cross-level analyses was conducted, acknowledged by the authors as a limitation.

What this implies

The findings point to a dual-pathway model of voice activation in public organizations — one formal (supervisor support as perceived organizational endorsement) and one informal (guanxi as relational trust buffer) — suggesting that PMS designs focused solely on structural incentives may underestimate the role of relational capital in eliciting discretionary employee contributions. The partial mediation result indicates that public service motivation alone does not fully account for the supervisor support–voice behavior link, implying that motivational profiling of employees may need to be coupled with relational management practices to produce voice behavior outcomes. The study also highlights a gap in cross-sector comparability of voice behavior mechanisms, as the dynamics documented here may differ substantially from those in private sector PMS research.

Key points

• Perceived supervisor support positively predicts subordinate voice behavior among frontline public officers, consistent with LMX theory predictions about insider status and psychological safety.
• Public service motivation functions as a partial — not full — mediator between supervisor support and voice behavior, indicating that direct relational effects of supervisor support on voice persist beyond motivational pathways alone.
• Supervisor-subordinate guanxi (informal, non-work social ties) moderates the supervisor support–voice behavior relationship, with high guanxi conditions producing stronger voice behavior outcomes, highlighting the role of informal relational trust in hierarchical Chinese public sector contexts.

What to take away

• Organizations in hierarchical public sector environments may observe higher rates of subordinate voice behavior in contexts where supervisors actively demonstrate support, as this shapes both motivational states and perceptions of safety for speaking up.
• Selection and cultivation of employees with strong public service motivation may be associated with higher levels of proactive voice behavior, particularly in environments where supervisor support is already present.