Abstract
This study examined whether compliance training quality (CTQ) enhanced operational efficiency (OE) through employee compliance capability (ECC) among cross-border e-commerce exporters, and identified which capability dimensions most strongly conveyed CTQ effects. A quantitative descriptive design was applied to 433 employees from five Guangdong exporters serving the Philippines and the United States. A five-point Likert instrument aligned with ISO 10015/37301 and U.S. DOJ guidance had been pilot-tested for reliability. Descriptive statistics, Pearson correlations, ordinary least squares, and parallel multiple mediation (PROCESS Model 4; bootstrap 5,000) were employed. CTQ was rated high, ECC remained low, and OE was very low. CTQ showed a positive association with ECC, and ECC showed a stronger positive association with OE. Mediation analyses indicated that ECC significantly conveyed CTQ influence on OE overall. At the sub-dimension level, Compliance Awareness and Compliance Culture and Continuous Improvement produced near-full mediation because direct paths became non-significant, whereas Incident and Dispute Management Capability and Process Execution Accuracy produced partial mediation because direct paths remained positive. Generalizability was constrained by one province, five firms, and two destination markets. The cross-sectional, self-report design limited causal and longitudinal inferences, and partner-process performance and detailed control variables were not incorporated. Training quality alone did not directly lift efficiency; gains materialized when knowledge had been absorbed as stable capabilities, particularly incident handling, execution accuracy, and a culture of continuous improvement, consistent with human-capital, organizational information processing, total quality management, and learning-organization perspectives. Recommendations included prioritizing rule-library and HS/HTS governance, protected training time with role-based micro-learning, incident playbooks with evidence management, simulator-supported quality gates, and mechanisms for peer coaching and lessons learned; policy support for data standards and targeted incentives was also emphasized.
