The research, conducted by Matthew L. Cole and Patricia Castelli of Lawrence Technological University and Yong Li of DTE Energy, surveyed 223 employed adults across 13 countries, roughly 45 percent of whom worked in hospitality. Participants rated their immediate supervisors using the 48-item Managerial Practices Survey, a validated instrument built on the Yukl taxonomy that separates leadership behavior into three meta-categories: task-oriented behaviors such as planning and clarifying, relations-oriented behaviors such as supporting and recognizing, and change-oriented behaviors such as envisioning and encouraging innovation. Employee engagement was measured with the 17-item Utrecht Work Engagement Scale, capturing vigor, dedication, and absorption, while retention was assessed through a three-item intention-to-remain scale. Confirmatory factor analysis confirmed that the three constructs were statistically distinct, with the proposed three-factor model outperforming rival specifications on every fit index: RMSEA of 0.042, CFI of 0.982, TLI of 0.979, and SRMR of 0.031.

The headline finding is a reversal of the authors’ own previous work. In an earlier study of a general multi-industry sample, change-oriented leadership, the envisioning, innovation-encouraging style beloved of pre-pandemic leadership literature, was the strongest predictor of engagement. This time, when the three behavior types were tested simultaneously, task-oriented leadership was the only significant predictor of engagement (β = 0.335), while relations-oriented and change-oriented behaviors showed no significant effects. The authors caution that the three behavior types are highly intercorrelated, with variance inflation factors indicating substantial multicollinearity, so the result identifies task-oriented behavior as the strongest correlate rather than an isolated cause. Still, the implication is provocative: the kind of supervision that best fuels engagement may depend on the operating environment, not on any universal hierarchy of leadership virtues.

The study then traced the causal-looking chain from supervision to staying power. Leadership behavior, modeled as a composite of the three types, significantly predicted engagement (β = 0.562) and retention (β = 0.205), and engagement mediated the leadership-retention relationship with a significant indirect effect of 0.298. When the composite was decomposed, only the indirect path running through task-oriented behavior remained significant. A further analysis of the three engagement dimensions revealed something subtler still: task-oriented supervision boosted vigor, dedication, and absorption alike, but only dedication, the experience of meaning, enthusiasm, and pride in one’s work, carried the effect forward to retention. The engagement that keeps people in their jobs, in other words, is not raw energy or immersion in tasks but a sense that the work matters.

Industry context turned out to be a powerful amplifier. The indirect effect of task-oriented leadership on retention through engagement was significant in both hospitality workers (indirect effect = 0.421) and non-hospitality workers (0.211), but the index of moderated mediation was significant, confirming the effect was roughly twice as strong in hospitality. The authors attribute this to the structure of frontline service work: hospitality employees are physically present, operationally dependent on their supervisors for task guidance and resources, and lack the remote-work autonomy, flexible scheduling, and professional identities that buffer engagement independently of supervision in other sectors. When the supervisor controls much of the daily work experience, the supervisory pathway to retention carries more of the load. Notably, the direct association between leadership and retention did not differ by industry; what varied was the mediated route.

Age added a second layer of nuance. The direct association between task-oriented leadership and retention strengthened significantly with age, from a non-significant effect among 18-to-29-year-olds to a substantial effect among workers aged 50 and above. The authors interpret this through career stage and reciprocity: older workers, with greater accumulated investment in their organizations, may read supervisory quality as a signal of organizational commitment. Meanwhile, the indirect effect through engagement was significant in every age group and largest among the youngest workers, though the formal moderation test for that pattern did not reach conventional significance. For organizations with age-diverse workforces, the practical suggestion is a differentiated retention strategy: older employees may stay or go based on supervisory quality itself, while younger employees may be retained primarily when the work feels meaningful.

The theoretical stakes extend beyond hospitality. The authors argue that claims about which leadership behaviors matter most may need to be conditioned on the environment in which they were observed. Pre-pandemic leadership models were developed and validated under operational stability, and their dominance in the literature may partly reflect that context. Crisis research during COVID-19 found that effective leaders shifted toward operational presence, direct communication, and task structuring, precisely the behaviors the new study links to engagement and retention. Consistent with contingency accounts of leadership, which hold that no single behavioral profile is universally effective, the findings suggest that operational disruption is a condition under which task-oriented behavior assumes primacy, displacing the visionary styles that dominate textbooks and training curricula.

The authors are careful about causal language. The design was cross-sectional and correlational, data came from a single period of sustained disruption with no within-sample variation in operating conditions, and all measures were self-reported from the same source, leaving open the possibility of common-method bias. The sample, while geographically diverse, cannot establish whether the hospitality pattern holds across different cultural orientations toward leadership, and industry served only as a proxy for unmeasured structural features such as supervisory contact frequency and schedule autonomy. External-oriented leadership behaviors, such as networking and environmental monitoring, were not examined and may have played an unmeasured role during the pandemic’s supply-chain and regulatory upheavals. Longitudinal and multi-source studies, the authors suggest, are needed before organizations can treat these associations as levers they can pull.

Still, the practical implications are concrete. A substantial share of managers report never having received formal management training, and the findings offer a basis for prioritizing where scarce development resources go: frontline, operationally intensive settings where supervisory contact is highest. Assessment frameworks that emphasize vision and inspiration may underweight the clarity, planning, and monitoring behaviors most closely associated with engagement in demanding environments. And because dedication, rather than engagement broadly, was the dimension tied to retention, organizations may get more from cultivating the meaningfulness of frontline work than from generic energy-boosting initiatives, whether through better task design, clearer role expectations, or recognition that connects daily service to a larger purpose.

The post-pandemic era has not resolved the hospitality workforce crisis; quit rates remain among the highest of any sector, and the structural vulnerabilities that COVID-19 amplified have not disappeared. What this study contributes is a sharper map of where organizations should look: not to charismatic visionaries, but to supervisors who bring order, clarity, and stability to daily work, especially in industries where employees have nowhere else to turn for structure. In a crisis, the study suggests, employees do not stay for the dream. They stay because someone made the present manageable, and made their work feel like it mattered.

Subject of Research: How task-oriented leadership behavior influences employee engagement and retention in the hospitality industry during and after the COVID-19 pandemic

Article Title: Leadership behavior, employee engagement, and retention: lessons from the COVID-19 pandemic in the hospitality industry

Article References: Cole, M. L., Castelli, P., & Li, Y. (2026). Leadership behavior, employee engagement, and retention: lessons from the COVID-19 pandemic in the hospitality industry. Frontiers in Organizational Psychology, 4, Article 1851922. https://doi.org/10.3389/forgp.2026.1851922

Image Credits: AI Generated

DOI: 10.3389/forgp.2026.1851922

Keywords: employee engagement, employee retention, leadership behavior, hospitality industry, COVID-19, task-oriented leadership, moderated mediation, organizational psychology, Great Resignation, dedication, frontline workers, Utrecht Work Engagement Scale