Can High‐Performance Human Resource Management Practices Enhance Employee Performance in Developing Countries? A Sequential Mediation Model

ABSTRACT This article advances the understanding of employee performance by unraveling the relationship between high‐performance human resource management practices (HPHRMP) and employee performance. Data were obtained from 510 bank employees across 25 banks operating in Jordan. A sequential mediation model was tested using structural equation modeling (SEM) with bootstrapping. An initial model assessed the pattern of associations for the complete bundle of HPHRMP. Subsequently, separate SEM models evaluate the associations for each individual practice. The findings partially support the hypothesized positive relationship between HPHRMP and employee performance, with continuance commitment and readiness for change operating as intermediate variables. However, the results reveal that in certain developing contexts, some HPHRMP may be perceived as unsupportive, triggering adverse social exchange mechanisms that undermine employee performance. The performed analysis yields valuable insights into how individual practices and practice bundles are associated with employee performance, continuance commitment, and readiness for change in the studied context, while also considering the intermediate role of readiness for change and continuance commitment. From a theoretical perspective, this study extends the ability–motivation–opportunity (AMO) framework and social exchange theory in explaining the outcomes of HPHRMP, suggesting that some inconsistencies in prior AMO‐based findings may be attributable to contextual contingencies. The use of side‐bet theory to understand the relationship between HPHRMP, continuance commitment, and employee performance is another theoretical contribution of this work, while from a practical standpoint, the results indicate that when managers face uncertainty regarding integrative versus additive approaches, implementing a comprehensive bundle of practices is not necessarily optimal.