Organizational size and public service performance: a meta-analysis and an agenda for future research

ABSTRACT

The question whether small or large organisations are associated with the best public service performance has long been a subject of debate in public administration research, and has had profound ramifications for practice. This article seeks to bring clarity to this debate by conducting a meta-analysis of studies scrutinising the relationship between organisational size and public service performance (45 articles, 122 effect sizes). Meta-analytical and meta-regression results show mostly null findings. We discuss the circumstances in which organisational size matters for public service performance, and propose rekindling venerable research agendas about nonlinearity and contingency in the size-performance relationship.

KEYWORDS:

• Organizational size
• public service performance
• meta-analysis
• null findings

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1. Studies that included class size as a measure of organisational size were excluded from our analysis because a class is a group of individuals within an organisation, rather than a sub-unit or department (see Meier & O’toole, 2013a, 2013b.

2. For example, Andersen et al. (2016) searched Scopus using the terms “performance” and “public” and returned 794 articles, which after close reading revealed that only 61 were suitable for the purposes of analysis.

3. We search for the use of organisational size as an independent variable, a control and for non-linear and moderated relationships. Non-linear, mediated and moderated relationships are not examined in enough empirical studies of organisational size and public service performance to undertake a meta-analysis.

4. Identified studies were coded by two of the authors. They coded each study independently and then shared their coding results. Disagreements were dealt with through discussion until agreement was reached. This approach was taken because of the need to identify studies that included measures of organisational size as a control.

5. For example, the organisational size-public service performance relationship was studied in the education arena. These studies used clients to operationalise organisational size, and label this school size.

6. Integration studies, including those on questions of organisational size, frequently tackle concepts that are operationalised in a variety of ways, see for example Boyne (2003), Damanpour (1993), Gerrish (2016), Walker and Andrews (2015) and Yeung (2009). Indeed, if the studies reviewed in a meta-analysis were homogeneous on questions of population, measures and settings etc. little evidence would be generated to support casual generalisations, and might introduce method bias (Matt, 2003).

Additional information

Notes on contributors

Richard M. Walker

Richard M. Walker is the Chan Hon Pun Professor of Behavioural and Policy Sciences, Dean of the College of Liberal Arts and Social Sciences and Chair Professor of Public Management in the Department of Public and International Affairs at City University of Hong Kong. Richard is Director of the Laboratory for Public Management and Policy and Co-director of the Centre for Public Affairs and Law.

Rhys Andrews

Rhys Andrews is Professor of Public Management at Cardiff Business School, Cardiff University.

Bert George

Bert George is an Associate Professor in the Department of Public and International Affairs at City University of Hong Kong.

Xuan Tu

Xuan Tu is an Associate Professor in the Department of Public Administration, Jiangsu Administration Institute.