A theory-based analysis of null causality between HRM practices and outcomes: Evidence from four-wave longitudinal data

Chidi Ogbonnaya, Kevin Daniels, Jake Messersmith, Yasin Rofcanin

Research output: Contribution to journalArticlepeer-review

Abstract

The last three decades have seen a growing interest in understanding the influence of human resource management (HRM) practices on employee job satisfaction and organizational performance. While the results have been generally positive, most studies have utilized cross-sectional research designs, which limit causal inferences. Recently, several studies have used longitudinal data but have not consistently found significant causal links between HRM practices and outcomes after controlling for past outcomes. This points to a tension in the literature that merits further investigation. Drawing on general systems theory (GST), we explore this issue by proposing and testing a set of null causal relationships involving HRM practices, organizational performance (i.e., patient satisfaction), and job satisfaction. We show that average scores on HRM practices and outcomes remain relatively stable at the organizational level over time, such that any observed within-organization change is likely negligible or non-significant. Using four-wave longitudinal data (with two, four, and six-year time lags) from the public healthcare sector, we argue that the causal links between HRM practices and outcomes are indeed sensitive to the forces of dynamic equilibrium operating within a highly institutionalized context. We use GST to highlight the self-sustaining nature of HRM systems and discuss the ramifications of this stability for strategic HRM research and practice.

Original languageEnglish
JournalJournal of Management Studies
Early online date26 Oct 2022
DOIs
Publication statusE-pub ahead of print - 26 Oct 2022

Fingerprint

Dive into the research topics of 'A theory-based analysis of null causality between HRM practices and outcomes: Evidence from four-wave longitudinal data'. Together they form a unique fingerprint.

Cite this