Although context is widely seen as essential for leadership research, the field is thought to suffer from a context deficit. Building on Johns (2006) and Rousseau and Fried (2001), this systematic review investigates whether empirical leadership studies go beyond merely acknowledging context and engage in contextualization by truly embedding context in the research design, analysis, and interpretation. We argue that contextualization is best assessed as a research practice in its own right. Yet, the value of contextual findings also depends on how a study handles two further challenges in leadership research: conceptual proliferation and causal ambiguity. We systematically review 209 empirical leadership studies (2003−2023), coding them not only on a large set of context variables, but crucially also on the degree of contextualization, and as complementary diagnostics, on conceptual clarity and causal inference. Our review shows that many leadership studies do typically incorporate context, but the degree to which they engage in high levels of contextualization remains limited. In addition, and irrespective of the degree of contextualization, the review brings to the fore that the analysis of the leadership concepts used or of causal inference in our sample of 209 papers is underdeveloped. The dual challenge that emerges from our review for leadership scholars is therefore not only to deepen contextualization as such, but to do so while also strengthening leadership conceptualization and causal inference.
Leadership approaches defined by inaction, such as laissez-faire leadership, remain understudied, leaving theory ill-equipped to address organizational challenges where knowing when to step back may be as important as knowing when to act. Where such work exists, it largely frames leader inaction as dysfunctional, reducing laissez-faire leadership to a one-dimensional negative construct. This framing constrains theoretical development by obscuring when, why, and under what conditions leader non-intervention might be appropriate or effective. To address this gap, we conduct a problematizing review of the laissez-faire leadership literature to examine how implicit assumptions have constrained theoretical development. We identify six such assumptions: (1) leadership necessarily involves action; (2) existing definitions adequately capture the construct; (3) negative findings are independent of measurement choices; (4) laissez-faire leadership is inherently negative; (5) current theoretical approaches sufficiently capture leader inaction; and (6) laissez-faire leadership reflects a stable leader tendency. We argue that these assumptions have restricted theoretical development and reinforced a pattern of largely unchallenged negative findings. Drawing on signaling theory and leader role theory, we reconceptualize laissez-faire leadership as leader behavioral non-intervention. Building on this reconceptualization, we outline a step-by-step agenda for future research that advances new conceptual, theoretical, and methodological approaches to studying leader non-intervention and clarifies when and why laissez-faire leadership may produce a range of outcomes. Together, this framework opens new avenues for theory development, empirical research, and leadership practice.
Leader-driven psychological safety has attracted growing organizational investment and a rapidly expanding empirical literature, yet the causal quality of the evidence underpinning the performance case, that is, the claim that leader behavior improves work outcomes through psychological safety, has never been systematically evaluated. This review takes a critical and methodological stance, conducting the first systematic causal audit of 47 articles published in highly ranked peer-reviewed journals between 1990 and January 2026. Applying a diagnostic protocol that evaluates concepts, measures, and study design, we find that 82% of coded entries operationalize the leader antecedent through follower ratings, rendering the evidence causally indeterminate at the point of measurement. No observational study employs a design-based identification strategy, and no study provides causal evidence for the full theoretical chain from leader behavior through experienced psychological safety to work outcomes. Ninety percent of coded entries are diagnosed as implausible causal inference. Only 10% of coded entries reach partial causal evidence, establishing a credible estimate for the antecedent-to-mediator link and leaving the mediator-to-outcome relationship descriptive only. The associations reported across the literature therefore neither confirm nor refute the theoretical argument. We offer theoretical and methodological recommendations for future research, provide a prescriptive roadmap with five pathways toward a more causally rigorous research program, and introduce a diagnostic flowchart to help scholars and practitioners assess the trustworthiness of available evidence.
With increased abilities and accessibility of Large Language Models (LLMs), subordinates increasingly rely on output generated by these powerful AI tools. Yet, existing leadership theory cannot meaningfully explain how subordinates' LLM use influences their supervisors' trust in them. In six studies (spanning five incentivized economic games and one field setting), we investigated how subordinate LLM use affects supervisory trust. First, we found that subordinates' use of LLM-generated output leads to less trust compared to when the subordinate generated the output themselves, but more trust compared to when the subordinate plagiarized the output (Study 1). As potential boundary conditions, we explored whether organizational LLM policy (Study 2), task criticality (Study 3), and supervisors' prior LLM experience (Studies 1-4 and 6) moderate the subordinate LLM use-supervisory trust linkage. Of these, only policy meaningfully moderated the effect on trust, such that when subordinates were encouraged (vs. forbidden or no policy stated), LLM use did not significantly reduce trust. We also found significant downstream consequences of LLM use (via supervisory trust) on perceived subordinate performance, organizational citizenship behavior, and promotability (Studies 1 and 5). Lastly, in an ecologically valid setting with supervisors working in academia, we found that breadth of subordinate LLM use reduces supervisory trust in a subordinate (Study 6). Collectively, this research contributes to scholarly understanding of the novel and understudied phenomenon of supervisor (trust) reactions to subordinate LLM use.
