Decoding effectiveness and efficiency in AI-enabled public services: a configurational pathway to citizen and employee satisfaction

The integration of AI in public services often poses a paradox: while it can streamline operations, it may simultaneously undermine service effectiveness, ultimately shaping how both citizens and employees perceive service quality. Seeking to address gaps in our understanding of service-related factors in AI-driven settings, this study employs fuzzy-set Qualitative Comparative Analysis (fsQCA) on survey data collected in Indonesia over an 8-month period from 457 citizens and 429 employees. The results reveal six configurations underpinning employee satisfaction and dissatisfaction, alongside four configurations driving citizen satisfaction and dissatisfaction. On the employee side, satisfaction thrives when service accessibility, operational effectiveness, and resource utilization are well-coordinated, whereas dissatisfaction emerges in the presence of fragmented workflows and the lack of key enablers. From the citizen perspective, satisfaction is fueled by trust-building elements such as service accuracy, transparency, and perceived service value, whereas their absence or misalignment leads to dissatisfaction. By unraveling these pathways, this study extends public administration and AI adoption literature, offering insights into how AI-enabled services can balance organizational objectives with user-centric needs. The findings highlight the importance of harnessing AI's efficiencies without sacrificing core service qualities, ultimately guiding strategies to optimize public service outcomes.