Examining the dynamics of pro-social rule-breaking among grassroots public servants

The study investigates grassroots civil servants’ attitudes and behaviors regarding pro-social rule-breaking (PSRB)—intentional violations of formal organizational policies to benefit the organization or its stakeholders. While traditional views treat rule-breaking as uniformly negative, recent work recognizes both destructive and constructive motives. Drawing on agency and stewardship perspectives, social exchange and social learning theories, deterrence theory, and the theory of planned behavior, the authors propose a framework linking social support, coworkers’ rule-breaking, bureaucratic structure (centralization and formalization), and disciplinary controls (certainty of detection and punishment) to PSRB attitudes and behaviors. The context is Pakistan’s public welfare programs, where frontline employees balance formal compliance with discretionary client service. The study addresses a gap by focusing on external situational antecedents of PSRB rather than individual traits.

The literature distinguishes rule-breaking from broader deviant constructs and notes its dual evaluation in public management. PSRB is defined as constructive, pro-social violations of written rules (Morrison, 2006), related to constructive deviance, pro-social behavior, and organizational citizenship behavior, and associated with benefits such as creativity, efficiency, client satisfaction, and improved image. Motivations include performing duties efficiently, helping colleagues, and better customer service. Social support, framed by social exchange theory, fosters extra-role pro-social behaviors and may encourage PSRB by enhancing wellbeing, shared norms, and information access. Coworkers’ behaviors shape attitudes and actions through social information processing and social learning; observing coworker rule-breaking may normalize and increase PSRB. Bureaucratic centralization concentrates authority, reduces autonomy, and typically deters deviation; formalization clarifies expectations and standardizes processes, limiting discretion and discouraging rule violations. Disciplinary controls—perceived certainty of detection and severity of punishment—are theorized by deterrence models to reduce deviance, including PSRB. Hypotheses: H1a/b social support positively relates to PSRB attitudes/behaviors; H2a/b coworker rule-breaking positively relates to PSRB attitudes/behaviors; H3a/b centralization negatively relates to PSRB attitudes/behaviors; H4a/b formalization negatively relates to PSRB attitudes/behaviors; H5a/b perceived punishment negatively relates to PSRB attitudes/behaviors; H6a/b perceived detection negatively relates to PSRB attitudes/behaviors.

Design: Two-stage descriptive online survey to reduce common method bias. Population/Context: Grassroots civil servants delivering services under Pakistan’s Ehsaas Welfare Programme (e.g., emergency cash, Kafalat, shelters, community kitchens, public health). Approvals obtained from relevant ministries and authorities. Sampling and procedure: Initial target N=600. Stage 1 collected PSRB attitudes (PSRBA), PSRB behaviors (PSRBB), controls, and identifiers; valid n=562. Stage 2 (≈2 months later) collected independent variables from the same respondents; matched valid n=522 (overall response rate 87%). Confidentiality/anonymity emphasized to mitigate social desirability bias. Measures: 7-point Likert scales (1=strongly disagree to 7=strongly agree). PSRBA and PSRBB (6 items each; adapted from Morrison, 2006; Dahling et al., 2012), α=0.91 and 0.93. Social Support (SS; 6 items; Spreitzer & Quinn, 1996), α=0.87. Perceived Coworkers’ Rule-Breaking (CRB; 6 items; adapted from Morrison, 2006; Dahling et al., 2012), α=0.94. Bureaucratic Centralization (BC; 4 items) and Formalization (BF; 5 items) from Dewar et al. (1980) adapted from Aiken & Hage (1968); α=0.86 and 0.89. Bureaucratic Rule-Breaking Punishment (BRBP; 4 items; Martin et al., 2013; Pifferi, 2022), α=0.88. Bureaucratic Rule-Breaking Detection (BRBD; 5 items; Hollinger & Clark, 1983; Abernethy & Brownell, 1997), α=0.80. Controls: gender, age, education, experience. Validity and reliability: EFA (principal components, Varimax) indicated AVE>0.50 and items loading on intended factors; discriminant validity supported (square roots of AVE exceeded inter-construct correlations). Composite reliabilities >0.90; Cronbach’s alphas 0.86–0.94. CFA (Amos) showed good fit for PSRBA model (χ2=484.63, df=216, χ2/df=2.24, CFI=0.96, TLI=0.97, SRMR=0.03, RMSEA=0.04) and PSRBB model (χ2=493.24, df=217, χ2/df=2.27, CFI=0.93, TLI=0.92, SRMR=0.03, RMSEA=0.06). Harman’s single-factor test indicated no serious common method bias (largest factor ≈35% variance). Analysis: Primary hypothesis tests via OLS regression with PSRBA and PSRBB as separate outcomes. Ordered probit models estimated as robustness checks with substantively similar results. Demographics described: 83.5% male; age primarily 36–40 (46.7%); education: 26.1% intermediate, 28.2% undergraduate, 45.8% graduate+; experience distribution centered at 6–10 years (27%).

OLS results (N=522):

• Social Support (SS) positively associated with PSRBA (β=0.153, p<0.05) and PSRBB (β=0.138, p<0.05), supporting H1a/H1b.
• Perceived Coworkers’ Rule-Breaking (CRB) positively associated with PSRBA (β=0.283, p<0.001) and PSRBB (β=0.254, p<0.001), supporting H2a/H2b.
• Bureaucratic Centralization (BC) negatively associated with PSRBA (β=−0.058, p<0.05) but not significantly with PSRBB (β=−0.046, p>0.05), supporting H3a and not H3b.
• Bureaucratic Formalization (BF) negatively associated with PSRBA (β=−0.146, p<0.01) and PSRBB (β=−0.128, p<0.001), supporting H4a/H4b.
• Perceived Bureaucratic Rule-Breaking Punishment (BRBP) negatively associated with PSRBA (β=−0.264, p<0.001) and PSRBB (β=−0.220, p<0.01), supporting H5a/H5b.
• Perceived Bureaucratic Rule-Breaking Detection (BRBD) negatively associated with PSRBA (β=−0.137, p<0.05) and PSRBB (β=−0.104, p<0.05), supporting H6a/H6b.
• Controls: Only gender showed a significant positive relationship with both outcomes; age, education, and experience were non-significant. Model fit: R²≈0.773 (PSRBA) and 0.771 (PSRBB).

Findings indicate that relational and social-context factors foster PSRB: when employees feel supported and observe peers breaking rules, they are more likely to develop favorable attitudes toward and engage in PSRB, aligning with social exchange, social information processing, and social learning theories. Conversely, bureaucratic structures (centralization, formalization) and disciplinary controls (perceived detection and punishment) deter PSRB, consistent with deterrence and planned behavior frameworks. The partial effect for centralization (attitudes only) suggests central authority may dampen willingness but not necessarily translate into behavior in the same way. Practically, public organizations should recognize PSRB’s constructive potential yet manage it carefully: improve rule design to reduce red tape, foster supportive climates and judicious discretion, and calibrate disciplinary systems to maintain compliance without stifling pro-social initiative.

Using two-wave survey data from grassroots civil servants in Pakistan’s welfare programs, the study shows that social support and coworker rule-breaking increase PSRB attitudes and behaviors, while centralization (attitudes), formalization, and perceptions of detection and punishment reduce PSRB. The work advances PSRB research by centering social, relational, and bureaucratic antecedents and suggests balancing supportive, decentralized environments with clear, effective rules and proportionate controls to harness beneficial PSRB. Future work should validate PSRB with multi-source and longitudinal designs, explore personality and leadership effects, differentiate formal vs. informal sanctions, and examine PSRB’s impacts on performance and stakeholder outcomes.

Self-reported, survey-based design risks social desirability and common method bias (mitigated by two-wave design, confidentiality, Harman’s test). Cross-sectional inferences limit causal claims. Measures reflect perceptions; external ratings were not used. Generalizability is limited to similar public welfare contexts. Future research should employ multi-source, longitudinal or mixed-method designs; examine personality and leadership contingencies; refine PSRB measurement (motives, beneficiaries, frequency); and distinguish formal and informal punishment mechanisms.