Comparing Measures of Citizen Trust and User Satisfaction as Indicators of ‘Good Governance’: Difficulties in Linking Trust and Satisfaction Indicators
| Author | Steven van de Walle,Geert Bouckaert |
| Published date | 01 September 2003 |
| Date | 01 September 2003 |
| DOI | http://doi.org/10.1177/0020852303693003 |
| Subject Matter | Journal Article |
Comparing measures of citizen trust and user satisfaction
as indicators of ‘good governance’: difficulties in linking
trust and satisfaction indicators
Geert Bouckaert and Steven Van de Walle
Abstract
Until recently, public administration mainly used so-called ‘hard indicators’, such as
resources and outputs, to monitor performance. Increased attention on accountability
and issues around impacts and outcomes have stimulated the introduction of ‘soft’
indicators — e.g. citizen and user satisfaction targets. Moreover, there is increased
demand for information on performance in relation to ‘governance’ as a whole,
including ‘quality of life’ indicators. Politicians, journalists and citizens increasingly
express their worries about a decreasing level of trust in government and the
detrimental effects this has on government and on the cohesion of society — they
appear to assume that more trust and more satisfaction equal better governance.
Increasing the quality of governance will thus also lead to citizens who are more
satisfied and more trusting. This article shows that current attempts to measure trust
and satisfaction in government are misleading if they claim to be measuring good
governance for two reasons. First, satisfaction is difficult to measure and very service-
specific. Second, trust in government is easier to measure but its linkages with good
governance are far from clear. Even when trust in government can be measured, it is
not at all clear whether changes in the level of trust are actually influenced by
government-related factors. We suggest, finally, the hypothesis that trust could be
insufficient but necessarily part of a set of indicators which are unnecessary but
sufficient for good governance.
Introduction
Until recently, public administration mainly used so-called ‘hard indicators’ (such
as, for example, resources and outputs) to monitor performance. Increased atten-
tion on accountability and the problems that have emerged in relating inputs,
activities and outputs to their effects (impacts, outcomes, etc.) have stimulated the
introduction of ‘soft’ indicators, e.g. satisfaction targets in the budget (Govern-
ment of Western Australia, 2000). Information on the functioning of separate
agencies and programmes is no longer regarded as enough by citizens, politicians
Geert Bouckaert is the Director of the Public Management Institute, Katholieke
Universiteit Leuven, Belgium, and Steven Van de Walle is a Researcher at the Public
Management Institute, Katholieke Universiteit Leuven, Belgium. CDU: 35.072.7.
The authors want to thank Elke Löffler, Tony Bovaird, Jarl Kampen and two anonymous
reviewers for their helpful comments.
International Review of Administrative Sciences [0020–8523(200309)69:3]
Copyright © 2003 IIAS. SAGE Publications (London, Thousand Oaks, CA and New
Delhi), Vol. 69 (2003), 329–343; 036082
02_IRAS69/3 articles 15/8/03 10:37 am Page 330
330
International Review of Administrative Sciences 69(3)
and, indeed, researchers. Instead, there is a demand for information on ‘govern-
ance’ as a whole. The recent attention on ‘quality of life’ indicators indicates this
(Bennett et al., 2001). The turn towards so-called ‘soft’ indicators is a sign of the
emancipation of public management and public administration studies from eco-
nomics, business and management faculties.
The World Bank defines governance as ‘the manner in which power is exer-
cised in the management of a country’s economic and social resources for
development’ (World Bank, 1992). Governance refers to steering and has to do
with the ability of human institutions to control their societies (Peters, 1995). In
its more practical application, however, ‘good governance’ often turns out to be a
very broad and value-laden concept (Kofele-Kale, 1999). It is used in the public
arena to discourage certain government actions and to promote others, even
though the definitions of governance usually do not, in themselves, contain the
ethical/moral criteria which would justify such conclusions.
Politicians, journalists and citizens increasingly express their worries about a
decreasing level of trust in government and the detrimental effects this has on
government and on the cohesion of society (Bok, 2001). The underlying hypothe-
sis of these politicians, public administrators and journalists, but also of most
researchers, is that more trust and more satisfaction will equal better governance.
Increasing the quality of governance will thus also lead to citizens who are more
satisfied and more trusting. When the concepts of trust and satisfaction are used in
the general discourse, they are assumed to measure citizens’ evaluation of the
quality of the ‘steering’ of society and of the direction in which society is
‘steered’.
This discourse contains a very mechanistic reasoning: increasing the quality of
governance will increase satisfaction and trust and, therefore, trust and satisfac-
tion indicators can be used as proxies for good governance. This article aims to
show that current attempts to measure trust and satisfaction in government are
misleading if they claim to be measuring good governance for the following two
reasons:
• Satisfaction has shown itself to be elusive to measurement and very service-
specific — making it a rather less meaningful indicator when aggregated. It can
only be interpreted in relation to the significance of a given service and responses
are easily swayed by the broader public mood.
• Trust in government is perhaps easier to measure but its linkages with
good governance are far from clear. Even when trust in government is indeed
measured, it is not at all clear whether the level of trust is actually influenced by
government-related factors.
A careful research design may help to deal with these problems but even then a
number of problems remain:
• in using satisfaction as an indicator,
• in using trust as an indicator and
02_IRAS69/3 articles 15/8/03 10:37 am Page 331
Bouckaert and Van de Walle: Trust and user satisfaction indicators
331
• in relating satisfaction with the functioning of public services, trust in
government and ‘good governance’.
The following sections will deal with each of these issues.
Satisfaction
The attention given to satisfaction indicators, and indeed subjective indicators in
general, is not new. At the end of the 1970s and the beginning of the 1980s, there
was an increase in social surveying, in general, and in satisfaction surveying more
specifically, especially with regard to local services (Stipak, 1979). Since the
beginning of the 1990s, developments in the instruments available for social
measurement and the levels of research in university economic and marketing
departments, together with the increased attention upon assessing customer orien-
tation in the public sector, stimulated further interest in satisfaction surveying.
The data generated by these surveys are often used to create benchmarks, to com-
pare services, to evaluate the quality of the management or to reconsider funding.
The renewed attention in the 1990s upon customer satisfaction started in North
America, with the most important landmark being the development of Servqual,
an instrument for measuring customer satisfaction in the service industries
(Zeithaml et al., 1990). In the context of government, we should also mention
the American Customer Satisfaction Index (ACSI), the Canadian Common
Measurements Tool (Schmidt and Strickland, 1998) and many local initiatives.
These initiatives have been followed swiftly in Europe, as is shown by a broad
range of applications in the UK (Moore et al., 1998), the Swedish Customer
Satisfaction Index and by an increased attention in Central Europe (SIGMA,
1998). In The Netherlands, the Court of Audit did a meta-analysis of client survey
initiatives (Algemene Rekenkamer, 1997) and a fiscal monitor (van Kruchten and
Peereboom, 1997) together with instruments to guide local authority initiatives
(Programmabureau Overheidsloket, 2000) have been developed. In Belgium, the
use of a standardized instrument at the federal level, the Quality Barometer, has
somewhat receded but, at the same time, a steering committee has been inaugu-
rated within the Ministry of the Flemish Community to coordinate all internal
initiatives (Bouckaert et al., 2001).
Variations in the measurement instruments, however, mean that the aggrega-
tion of satisfaction data remains difficult. To know what these data mean and how
to interpret them is even more important. The most important question, however,
is construct validity — what does ‘satisfaction’, in fact, measure?
What does satisfaction mean?
In satisfaction surveys, a direct causal relation is presupposed between the quality
of a certain service delivery and user satisfaction. If service quality increases,
satisfaction is assumed to increase as well. In reality, however, this is not always
the case (Bouckaert, 1995) as there are
• differences in producer and consumer views on quality,
02_IRAS69/3 articles 15/8/03 10:37 am Page 332
332
International Review of Administrative Sciences 69(3)
• changes not only in quality and perceptions of it but also in expectations and
• different service characteristics.
All these factors thus have to be taken into account. A single satisfaction indicator
does not tell us enough about the reasons for a certain level of satisfaction.
In methodologically sound satisfaction surveys (which are still the exception
rather than the rule), clients are not only...
To continue reading
Request your trial