HOW TO MAKE PUBLIC NETWORKS REALLY WORK: A QUALITATIVE COMPARATIVE ANALYSIS
| Published date | 01 March 2016 |
| DOI | http://doi.org/10.1111/padm.12192 |
| Date | 01 March 2016 |
| Author | JOSIP MARKOVIC,DANIELA CRISTOFOLI |
doi: 10.1111/padm.12192
HOW TO MAKE PUBLIC NETWORKS REALLY WORK:
A QUALITATIVE COMPARATIVE ANALYSIS
DANIELA CRISTOFOLI AND JOSIP MARKOVIC
Many studies have striven to understand which factors affect the performance of public networks.
However, there are very few studies in the eld of public management that investigate the joint,
interactive effects of different determinants on network performance. This article uses the relatively
new method of qualitative comparative analysis (QCA) to investigate the complex causality of deter-
minants and network performance. It examines the combination of resource municence, central-
ization of the network structure, formalization of coordination mechanisms, network management
and their joint effects on network performance. An analysis of 12 Swiss networks providing home
and social care services shows that there are a range of possible paths. Various combinations of the
above-mentioned factors can in fact lead to high network performance. The paths provide insight
into how to make public networks really work.
INTRODUCTION
Ever since Provan and Milward (1995) published their seminal article on four public men-
tal health care networks in the US, scholars in the eld of public management have tried
to assess whether public networks really work and which determinants affect their per-
formance. There have been increasing efforts by scholars to shed light on a plethora of
factors that affect network performance (for a review see Turrini et al. 2010, among oth-
ers), but they have rarely investigated the possibility of these factors having a combined
effect on network success (Turrini et al. 2010; Raab et al. 2015; Verweij et al. 2013).
Exploring this aspect is the aim of this article. In particular, we are interested in the
way in which four of the predictors of public network performance identied by previous
studies (Provan and Milward 1995; Kickert et al. 1997; Agranoff and McGuire 2003; Provan
and Kenis 2009) combine to give high network performance. These predictors are resource
municence, centralization of the network structure, formalization of coordination mech-
anisms and network management.
The congurational approach of qualitative comparative analysis (QCA, Ragin 1987)
was chosen in order to conduct a systematic comparison of 12 home care networks in
Switzerland (along with the above-mentioned factors and their performance) and identify
the different paths that lead to network success.
The article is divided into four parts. The rst section reviews the relevant literature
for our research and outlines the theoretical background on which the research design
was based. Different determinants of public network performance have been extracted
and operationalized along with the performance measurement criteria. In the next section,
there is a presentation of the method applied (qualitative comparative analysis) and the
empirical setting, along with details of the case selection and data collection processes. The
third section contains a summary of the ndings. The QCA follows an iterative approach
and constantly goes back and forth between empirical cases, theories and the actual results,
so the ndings will be interpreted in the context of relevant theories and cases at the end
of the article. The results will contribute to the existing literature by shedding new light
Daniela Cristofoli and Josip Markovic are in the Department of Political Economy at the University of Lugano, Switzer-
land.
Public Administration Vol.94, No. 1, 2016 (89–110)
© 2015 John Wiley & Sons Ltd.
90 DANIELA CRISTOFOLI AND JOSIP MARKOVIC
on long-standing theories about the joint effects of certain determinants of public network
performance.
This is the latest in a series of three articles exploring the joint effects of differentdetermi-
nants of network performance (Cristofoli et al. 2014, 2015). In particular, moving on from
Turrini et al. (2010) and the authors’ invitation to explore whether an interaction effect
among the determinants of network performance can be supposed, the rst article made
us more condent about the existence of a relationship between network structure, mecha-
nisms and management that jointly affects network performance (Cristofoli et al. 2015). On
that basis, the second article focused on a specic form of network governance and showed
that success can be achieved in shared-governance networks through the simultaneous
presence of formalized coordination mechanisms and a group of ‘network administrators’
(Cristofoli et al. 2014).
Following these ideas and expanding their scope, the current article sheds light on
various combinations of factors that also lead to network success. It thus conrms and
enriches the results of previous studies. First of all, it conrms the existence of a relation-
ship between some characteristics of network structure, mechanisms and management.
Second, it conrms the existence of joint effects among the above-mentioned factors: they
simultaneously affect network performance. Third, it makes it possible to identify at least
two different paths leading to high network performance in resource-municent con-
texts: the rst involves centrally governed network structures and network management
(facilitating, mediating and/or leading), while the second involves shared-governance
networks and formalized coordination mechanisms.
LITERATURE REVIEW
The success of public networks is a long-standing theme in public network literature (for
example, see Mandell 1984; Agranoff 1986; Provan and Milward 1995; Ferlie and Pettigrew
1996; Kickert et al. 1997; Provan and Sebastian 1998; Mandell 2001; Provan and Milward
2001; Herranz 2009; Kenis and Provan 2009). The existing studies usually focus on deter-
minants of network performance without considering the effects of varying combinations
of them. The idea of analysing joint, interactive effects of these combinations of factors is
relatively new to the eld (Turrini et al. 2010). In the following literature review,1we will
only focus on relevant studies that suggest conjunctional relationships between at least a
couple of the determinants, thus leading us to suppose that there are multiple causal paths
for network success and allowing us to derive a theoretical framework for the qualitative
comparative analysis.
In their seminal article, Provan and Milward (1995) identied network context and net-
work structure as key variables for network success. They argued that resource muni-
cence is paramount for maintaining effective networks, but it is not sufcient to guaran-
tee effective provision of services. Therefore, external factors have to be accompanied by
appropriate network structures to foster high network effectiveness.
Mentioning network structure as a key determinant of network performance, Provan
and Milward (1995) were also able to demonstrate a positive relationship between the
degree of network integration, the degree of network centralization and network perfor-
mance. They showed that centralized integration is a key determinant of network perfor-
mance, as it will be signicantly enhanced if a network is integrated through a central core
agency (Provan and Milward 1995). Expanding on the rst article, Provan and Sebastian
(1998) identied another path toward high network performance. They showed that if a
Public Administration Vol.94, No. 1, 2016 (89–110)
© 2015 John Wiley& Sons Ltd.
To continue reading
Request your trial