Leadership in the public sector: A meta‐analysis of styles, outcomes, contexts, and methods

• Published date: 01 November 2022
• Author: Leonie Backhaus,Rick Vogel
• Date: 01 November 2022
• DOI: http://doi.org/10.1111/puar.13516

RESEARCH ARTICLE

Leadership in the public sector: A meta-analysis of styles,

outcomes, contexts, and methods

Leonie Backhaus | Rick Vogel

Faculty of Business, Economics and Social

Sciences, Department of Socioeconomics, Chair

of Public Management, Universität Hamburg,

Hamburg, Germany

Correspondence

Leonie Backhaus, Faculty of Business, Economics

and Social Sciences, Department of

Socioeconomics, Chair of Public Management,

Universität Hamburg, Von-Melle-Park 9, 20146

Hamburg, Germany.

Email: leonie.backhaus@uni-hamburg.de

Abstract

Leadership has attracted growing attention among scholars and practitioners in

public administration. With the rising availability of study results, however, it

becomes increasingly difficult to keep track under which conditions leadership

does or does not make a difference in the public sector. This study provides a

meta-analysis of administrative leadership and various correlates that research has

theorized as outcomes of leadership. The results of a multi-level random-effects

meta-analysis based on 486 effect sizes from 151 studies show that correlations

are stronger for the achievement of beneficial than for the prevention of detrimen-

tal outcomes, as well as for group- and organization-related than for employee-

related outcomes. Moderation analyses reveal that leadership style, administrative

tradition, administrative subfield, and methodological factors explain heterogene-

ity in effect sizes.

Evidence for Practice

•Leadership in the public sector is positively and largely consistently related to

beneficial outcomes, such as performance, and negatively associated with detri-

mental outcomes, such as turnover.

•Within the range of the analyzed leadership styles, it is not particularly important

how public leaders lead —as long as they do lead. Laissez-faire leadership is

confirmed to be a largely ineffective style of leadership. Leaders should be simi-

larly cautious with the exercise of controlling forms of leadership.

•The broad range of leadership styles helps HR practitioners to adjust leadership

trainings to the available talents and given settings. Public leaders may feel

encouraged that there are many ways to become effective leaders.

•Administrative tradition matters, as the leadership-outcome relationship is less strong

in some continental European as well as East Asian traditions than in the Anglo-

American tradition. For example, this applies to transformational leadership.

INTRODUCTION

Leadership is among the core concepts in public adminis-

tration (PA) scholarship (Chapman et al., 2016; Crosby &

Bryson, 2018; Ospina, 2017; Van Wart, 2003,2013a). Two

decades ago, research onleadership in the public sectorwas

at a nascent but still limited stage, severely lagging behind

general leadership studies. Van Wart (2003), in his review of

the then available literature, arrived at the conclusion that

“the needs are great and the research opportunities are

manifold”(p. 225). Since then, PA scholars have frequently

echoed this call and seized the opportunities for leadership

research (Chapman et al., 2016; Van Wart, 2013a;Vogel&

Masal, 2015). The number of available studies has grown con-

siderably, forming a flourishing stream in PA scholarship.

The growing availability and diversity of studies, however,

does not make it easier to answer one of the core questions

of leadership: “To what degree does leadership make a differ-

ence?”(Van Wart, 2003,p.221).Althoughscholarsandpracti-

tioners would largely agree that leadership is an important

Received: 18 June 2021 Revised: 11 April 2022 Accepted: 26 April 2022

DOI: 10.1111/puar.13516

This is an open access article under the terms of the Creative Commons Attribution-NonCommercial License, which permits use, distribution and reproduction in any

medium, provided the original work is properly cited and is not used for commercial purposes.

© 2022 The Authors. Public Administration Review published by Wiley Periodicals LLC on behalf of American Society for Public Administration.

986 Public Admin Rev. 2022;82:986–1003.

wileyonlinelibrary.com/journal/puar

phenomenon that can make a difference in public organiza-

tions, there is little clarity beyond this general consent. The

polyphony begins with different conceptualizations of leader-

ship characteristics and behaviors. A majority of scholars has

focused on transformational leadership (Chapman et al., 2016),

but many more leadership styles have recently been exam-

ined. Some scholars have voiced concerns about the adop-

tion of generic leadership styles and have begun to develop

leadership constructs that account for the specificities of pub-

lic leadership (Fernandez et al., 2010;Tummers&Knies,2016;

Vogel et al., 2020). There is thus a plethora of leadership

styles, and it is unclear if there are differences in the differ-

ence they potentially make.

Conceptual diversity does not only characterize leader-

ship as an independent variable, but also the outcomes of

leadership, that is, the dependent variables in the leader-

ship equation. Consistent with an understanding of

leadership as a process of influencing others (Yukl &

Gardner, 2020), many scholars have examined how leader-

ship influences the attitudes, motivations, and behaviors of

followers, such as job satisfaction (e.g., Park & Rainey, 2008),

work engagement (e.g., Ancarani et al., 2020), or individual

performance (e.g., Hassan, Park, & Raadschelders, 2019).

However, the outcomes differ strongly in kind and may

relate, for instance, to employees (such as work engagement;

Tummers & Knies, 2013), groups (such as team climate;

Kroll & Vogel, 2014), or the organization (such as organiza-

tional performance; Jacobsen & Andersen, 2015). Further-

more, leadership may not only be effective in bringing about

beneficial outcomes but also in preventing detrimental out-

comes, such as turnover intentions (e.g., Park & Rainey, 2008).

It is unclear, however, if leadership makes the same differ-

ence across outcomes of different kinds and valence.

With the growth of leadership studies in PA, more

insights from various subfields of the public sector and

from different national sites have become available. While

this diversity is both needed and welcomed, it also raises

the question whether leadership is effective in different

contexts to similar extents (e.g., Slyke et al., 2006). For

example, leadership has been studied in subfields of the

public sector as diverse as the federal government

(e.g., Caillier, 2014), education (e.g., Bauwens et al., 2019),

and the military (e.g., Hattke et al., 2018). The emergence

and effectiveness of leadership may be contingent on the

distinct professional cultures in each of these subfields.

Leadership may also co-vary with national and adminis-

trative traditions and cultures (House et al., 2004; Painter

& Peters, 2010). Administrative systems are likely to be

characterized by different configurations of “substitutes

for leadership”(Kerr & Jermier, 1978), which may enhance

or neutralize the effects of leadership. In short, the extent

to which insights from leadership studies can be general-

ized and transferred across different administrative sub-

fields and systems remains an open issue. It is therefore

fair to say that scholarship has not yet fully accounted for

the context in which leadership occurs (e.g., Ospina, 2017;

Van Wart, 2013a; Vogel & Masal, 2015).

Considerable variation occurs not only in the what and

where of studying leadership in the public sector, but also in

the how. Leadership scholars can choose from a large pool of

research instruments and designs when they decide how to

gather data and how to arrange the analysis. Among the

decisions with which researchers frequently struggle is whom

to ask for leadership ratings, which data sources to combine,

and when to measure the variables. Such decisions may have

important consequences for the results that will be obtained

and for the claims that can be made. In the broader PA com-

munity, there is increasing awareness of these issues, most

evidently reflected in discussions about common method

bias (George & Pandey, 2017; Jakobsen & Jensen, 2015;

Meier & O’Toole, 2013). How methodological factors interfere

with the association between leadership and outcomes, how-

ever, has not yet been assessed on a broad scale.

To conclude, readers of PA journals face a growing

body of leadership research but are increasingly left puz-

zled with what they can learn from single studies and pass

on to practitioners, given the variety of styles, outcomes,

contexts, and methods in this scholarship. The present

study aims to synthesize previous findings in an up-t o-date

and integrative review of quantitative leadership research

in the public sector. The focus is on administrative leader-

ship, that is, the leadership of non-elected leaders in public

sector settings, rather than on political or community lead-

ership (Van Wart, 2013a). To what extent is administrative

leadership related to outcomes of different kinds and valence?

How do leadership styles as well a s administrative traditions

and subfields moderate these relationships? Which role do

methodological factors, such as rating methods, data sources,

and research designs, play in explaining variance in the

leadership-outcome relationship?

To address these questions, this article presents a meta-

analysis of the relationship between administrative leader-

ship and various correlates that previous scholarship has

conceptualized as outcomes of leadership. The study builds

on 151 studies (N=2,819,591) published in PA journals

since 2000 and conducts a multi-leve l random-effects

meta-analysis with moderator analyses. The detailed find-

ings improve the development of theory and research and

inform administrative practice. Scholars learn about under-

researched phenomena and important contingencies of

leadership in the public sector, thus being provided with

guidance in setting the agenda for future research and

moving the field forward. Practitioners can conclude which

leadership styles are worth developing and where to pay

attention when transferring these styles from one context

to another.

LEADERSHIP IN THE PUBLIC SECTOR: STYLES,

OUTCOMES, CONTEXTS, AND METHODS

This section develops the conceptual framework for the

meta-analysis along the four lines introduced above, each

addressing an important question that needs to be

PUBLIC ADMINISTRATION REVIEW 987