Preference for Managerial Boundary Setting in Relation to Empowerment: Adding Clarity to the Role of Boundaries

Abstract

This study explores the role of manager and employee preference for managerial boundary setting in empowerment. Research has shown a clear relationship between managers’ empowerment practices and employee psychological empowerment, but confusion persists in the empowerment literature about the role played by boundaries in creating empowerment. We add clarity to the role of boundary setting by considering how the individual difference variable of manager and employee preference for managerial boundary setting impacts empowerment. Results indicate that higher preference for managerial boundary setting was associated with greater utilization of empowerment practices by managers and with greater psychological empowerment of employees. For managers there was a positively-accelerating quadratic relationship between preference for managerial boundary setting and empowerment practices. We also confirm the positive relationship between managers’ empowerment practices and employee psychological empowerment, and we found that employee preference for boundary setting did not moderate this relationship, except in the model for competence.

Keywords:

• empowerment
• preference
• structure
• boundaries

Notes

1. Following Kenny, Kashey, and Cook’s (2006) recommendation to test dyadic data with multilevel modeling, we repeated the analysis using Hierarchical Linear Modeling (HLM). HLM is preferred to ordinary least-squares regression because HLM simultaneously accounts for variances and covariances both within and between groups (Raudenbush & Byrk, 2002). Dyadic data are a special case of hierarchically nested data, and as such can be modeled with HLM (Kenny et al., 2006). Normally in HLM, within-group differences are shown by differences in both intercepts and slopes. However, for dyadic data, models must be constrained to include only the fixed effects at the lower (subordinate) level because dyads do not contain enough members to allow random effects across dyads. Constraining the model in this way allows member scores within dyads to be modeled via the intercepts only. This method does not bias the estimates, and thus allows for the use of HLM with dyadic data (Kenny et al., 2006).

2. This hypothesis involves a partial replication of a hypothesis from a previous study using the same data (Randolph & Kemery, 2011). However, there are two important differences between the analyses. First, the previous study’s analysis was merely correlational, while this study employs a multivariate technique. Second, the previous test for this hypothesis examined only the overall measure of psychological empowerment, while this study examines both the overall measure and the four subdimensions (meaning, competence, self-determination, and impact) of psychological empowerment.

3. We gratefully acknowledge an anonymous reviewer for pointing this out.

Additional information

Notes on contributors

Edward R. Kemery

Edward R. Kemery, PhD, is an associate professor of management at the University of Baltimore. He can be reached at [email protected].

W. Alan Randolph

W. Alan Randolph, PhD, is a professor of leadership and international business at the University of Baltimore. He can be reached at [email protected].

Lisa T. Stickney

Lisa T. Stickney, PhD, is an associate professor of management at the University of Baltimore. She can be reached at [email protected].