Micro CSR intervention towards employee behavioral and attitudinal outcomes: a parallel mediation model

Corporate social responsibility (CSR) has evolved from a focus on legal and economic obligations to encompass moral, social, and environmental responsibilities. Recent work emphasizes micro-CSR—how individual employees perceive, evaluate, and respond to their organization's CSR practices—and its effects on employee attitudes and behaviors. The study addresses gaps regarding the mediating mechanisms linking perceived CSR to employee outcomes and the need to examine both attitudinal (e.g., satisfaction, commitment) and behavioral outcomes (e.g., turnover intention, counterproductive behaviors). Drawing on social exchange theory, the research proposes a parallel mediation model where organizational justice, trust, and identification transmit the effects of perceived CSR to employee outcomes. Objectives: (a) test direct effects of perceived CSR on commitment, job satisfaction, CWBs, and turnover intention; (b) test parallel mediating roles of justice, trust, and identification in these relationships; and (c) synthesize prior evidence via a PRISMA-based meta-analysis.

Prior research links perceived CSR to organizational justice, trust, and identification, which in turn influence employee attitudes and behaviors. From a social exchange and justice perspective, CSR signals fairness and integrity, fostering trust and identification, thereby enhancing commitment and satisfaction and reducing CWBs and turnover intentions. Identification has been shown to mediate CSR–commitment and CSR–satisfaction links; justice and trust serve as antecedents to identification and mediate CSR effects on outcomes. The study formulates hypotheses that perceived CSR positively predicts commitment (H1) and job satisfaction (H2), and negatively predicts CWBs (H3) and turnover intention (H4). It also posits that organizational justice, trust, and identification mediate CSR effects on commitment (H5–H7), job satisfaction (H8–H10), CWBs (H11–H13), and turnover intention (H14–H16). A parallel mediation framework is adopted to capture distinct and simultaneous pathways from CSR perceptions to outcomes.

Design and sample: A positivist, deductive, cross-sectional survey tested a pre-specified model among senior and junior executives of Pakistani cellular networks (Mobilink, Telenor, Ufone). Stratified random sampling was used. Of 798 distributed questionnaires, 767 usable responses were obtained (92% response rate). Measures: A self-administered questionnaire used five-point scales to assess perceived CSR (6 items; Hameed et al., 2016; Hu et al., 2020), organizational justice (8 items; Niehoff & Moorman, 1993), organizational identification (3 items; Sass & Canary, 1991), organizational trust (3 items; Seppänen et al., 2007), organizational commitment (6 items; Meyer & Allen, 1997; Sass & Canary, 1991), counterproductive work behaviors (Bennett & Robinson, 2000), job satisfaction (3 items; Witt & Nye, 1992), and turnover intention (3 items; Cammann et al., 1979). Analysis: Exploratory factor analysis (KMO/BTS adequate) with principal component analysis confirmed substantial loadings after seven iterations. A measurement model demonstrated acceptable fit (e.g., RMSEA 0.08, GFI 0.91, NFI 0.95, CFI 0.97, TLI 0.92, χ²/df ≈ 3.0). Reliability and validity: Composite reliability and Cronbach’s alpha ranged ≈ 0.72–0.81 (>0.70). AVE ranged ≈ 0.53–0.73 (>0.50), indicating convergent validity. Discriminant validity was supported via cross-loadings, Fornell-Larcker (square roots of AVE greater than inter-construct correlations), and HTMT ratio ≈ 0.71 (<0.90). Structural modeling: SEM (AMOS) estimated direct and indirect (parallel mediation) paths; bootstrapping with 5000 samples assessed significance. Predictive strength indicated ≈ 72% variance explained in endogenous outcomes. Mediation was additionally examined using Hayes’ PROCESS macro. Meta-analysis: Following PRISMA (Page et al., 2022), databases (e.g., Sci-Hub, Emerald Insight, ResearchGate, Web of Science) were searched for 2009–2022 using terms such as “CSR perception,” “stakeholder behavioral outcome,” and “stakeholder attitudinal outcome.” Inclusion: quantitative studies examining links between employee CSR perceptions and job-related outcomes. Forty-three studies (total N=29,493) were coded (study details; CSR perception; behavioral and attitudinal outcomes; correlation coefficients). Effect sizes used r with Fisher’s z transformations; variance via Hedges & Olkin; random-effects (ML) models estimated in JASP. Forest plots summarized behavioral and attitudinal associations.

Direct effects (SEM): Perceived CSR positively predicted organizational commitment (β=0.210, t=6.7, p<0.01) and job satisfaction (β=0.153, t=5.7, p<0.01), and negatively predicted CWBs (β=−0.112, t=−5.5, p<0.01) and turnover intention (β=−0.117, t=−5.0, p<0.01). Model predictive strength indicated ≈72% variance explained across endogenous outcomes. Mediation: Organizational justice, trust, and identification significantly mediated CSR effects on commitment—OJ (β=0.081, t=5.77), OT (β=0.402, t=14.2), OI (β=0.022, t=3.19)—with 95% CIs excluding zero. For job satisfaction, OJ (β=0.033, t=3.78), OT (β=0.054, t=3.60), OI (β=0.262, t=6.66) were significant mediators. For CWBs, OJ (β=0.055, t=3.15), OT (β=0.145, t=4.73), OI (β=0.026, t=2.94) were significant mediators. For turnover intention, OJ (β=0.040, t=2.17), OT (β=0.096, t=2.98), and OI (β=0.014, t=2.02) showed small but statistically significant indirect effects with CIs excluding zero. Meta-analysis: Of 43 included studies (N=29,493), 23 reported positive associations between CSR perceptions and behavioral outcomes, and 20 reported positive associations between CSR perceptions and attitudinal outcomes; random-effects models indicated overall positive effects. Measurement quality: Reliability (CR/α ≈ .72–.81), convergent validity (AVE ≈ .53–.73), and discriminant validity (Fornell-Larcker; HTMT ≈ .71) were supported; model fit indices were acceptable (e.g., RMSEA .08; CFI .97).

Findings support social exchange theory: employees perceiving their organization as socially responsible reciprocate with improved attitudes (commitment, satisfaction) and reduced negative behaviors (CWBs, turnover intention). CSR appears to signal fairness and integrity, strengthening justice perceptions, trust, and organizational identification—key socio-cognitive mechanisms transmitting CSR’s effects to outcomes. The parallel mediation framework demonstrates that multiple, distinct affective-cognitive pathways jointly explain CSR’s impact. Contextually, in Pakistan’s telecom sector, micro-CSR perceptions align with improved employee outcomes, echoing global evidence. Practically, organizations should implement and communicate CSR initiatives that emphasize fairness, stakeholder care, and authentic values to enhance trust and identification, thereby improving employee outcomes. The study also integrates primary SEM results with a meta-analytic synthesis, reinforcing generalizability of positive CSR–employee links.

Perceived micro-CSR directly enhances organizational commitment and job satisfaction and reduces CWBs and turnover intention. Organizational justice, trust, and identification function as parallel mediators, significantly transmitting CSR effects to commitment, satisfaction, and CWBs, with small but significant indirect effects also observed for turnover intention in the bootstrapped analyses. A PRISMA-guided meta-analysis of 43 studies corroborates positive CSR–employee outcome relationships across behavioral and attitudinal domains. The work advances micro-CSR by articulating a parallel mediation model grounded in social exchange theory and underscores the managerial value of authentic, well-communicated CSR programs to foster positive employee attitudes and behaviors. Future research should broaden mediators/moderators and refine causal inference and measurement of actual versus perceived CSR.

Several limitations apply: (1) The model omitted potentially important mediating and moderating variables due to limited empirical support in micro-CSR; future studies should test additional mechanisms (e.g., employee CSR practices, innovation, performance). (2) Cross-sectional design limits causal inference; longitudinal or field experiments are recommended. (3) Method bias and contextual factors warrant methodological triangulation. (4) The study focused on perceived CSR; the differential effects of actual vs. perceived CSR and related determinants should be examined. (5) Sectoral and national context (Pakistan telecom) may limit generalizability; replication in diverse contexts is advised.