Whether we are concerned with individuals or social systems, trust is fundamentally about expectations about the future when we cannot know the outcome or consequences of a decision to act in a particular way (Messick & Kramer 2001 p 91, Misztal 1996). In other words, making a choice involves risk. Risk is a concept that describes the uncertainty of the outcomes of human action in a complex world (Luhmann 1979). Risks, from a social perspective, only exist in the context of ‘decision and action. They do not exist by themselves. If you refrain from action you run no risk’ (Luhmann 1988 p 100). If a course of action is chosen ‘in spite of the possibility of being disappointed by the actions of others, you define the situation as one of trust’ (Luhmann 1988 p 97). Trust also requires knowledge upon which to base a decision to trust. In regard to individuals, the knowledge may be about a person’s circumstances, behaviour or motivations, based on prior interactions or recommendations of others. In regard to a system, it is knowledge of ‘the ability of the system to maintain conditions and to perform its functions’ (Misztal 1996 p 21), built up through ‘continual, affirmative experiences with the system’, or the experience of ‘taken-for-granted routines’ (Mollering 2006 p 72). Mollering (2006) argues that we cannot possibly have all the knowledge required to make a fully informed decision to trust individuals or systems, so the decision to trust necessarily involves a ‘leap of faith’ or the ‘suspension of doubt’. The word risk is not only used to describe a quality of the relationship between people and social systems, the perspective that is being discussed here. It is also used to describe ‘objective risk’ or the risk posed by factors external to people and social relationships; for example, risk posed by infective agents or by other environmental hazards. In this chapter trust is not being discussed in relation to these ‘objective’ risks.
The emphasis on knowledge and rational decisions reflects one tradition within the study of trust. The alternative tradition views trust as a quality of relationships that have emotional and ethical roots. From the latter perspective the decision to trust is based on expectations that the other will abide by ethical rules relevant to the situation (Messick & Kramer 2001 p 91). One key ethical rule is the principle of not knowingly doing harm to the person bestowing trust. Ethical rules are an element of culture and people gain knowledge of them through social learning.
Papadakis (1999 p 75) argues that many changes in Australia in the last two decades of the 20th century have lead to significant change in perception of the ‘effectiveness and value of many institutions’. At least some of these changes were made to the institutions that help sustain trust in democratic government. Using data from the Australian Values Survey (1983) and the World Values Survey (Australia) (1995) Papadakis was able to demonstrate ‘a sharp decline in confidence in governmental and non-governmental organisations between 1983 and 1995’ (Papadakis 1999 p 75). The significance of the substantial decline in confidence (confidence is conflated with trust in this work) is arguable, but concern is widespread (Hudson 2006, Job 2005, Levi 1998, Papadakis 1999).
IN WHAT WAY IS TRUST A PROBLEM IN PUBLIC POLICY?
The issue of trust in government and public institutions is a significant issue in many countries; in Britain (Alaszewski 2003, Taylor-Gooby 2006a), Europe (Organisation of Economic Cooperation and Development [OECD] 2001), the United States (ANES 2006), and in Australia (Job 2005). The factors influencing citizen trust in government and public institutions are complex and, to some extent, dependent on historical and social context. For example, in the emerging democracies of Eastern Europe trust is influenced by social and political history in ways not experienced in the established democracies of the West (e.g. Mishler & Rose 2005). In the West, factors influencing trust in government can be classified as economic (citizen evaluations of the national economy), socio-cultural (crime rates and child poverty), and political (citizen evaluation of incumbents, institutions, scandals and corruption) (e.g. Chanley et al. 2000). The recurring themes underlying concern about trust in government are legitimacy and effectiveness of elected government, and public institutions consisting of the public service and publicly funded organisations such as hospitals and health centres that provide services to the community. The following discussion will focus on trust in public institutions with particular emphasis on the health sector.
If one is interested in the condition of trust in public policy the issues that might lead to it being defined as a problem depend on whether it is viewed through the lens of the rational or relational policy traditions (Job 2005, Taylor-Gooby 2006a). From a rational perspective, embodied in the new public management (NPM), rational calculation of performance provides the foundation for trust in systems or public institutions. If the public is informed about an institution (such as a health service), and it is meeting performance targets, then citizens are able to calculate the benefits to them and their trust in the institution should increase. From the relational perspective a good experience of those systems or institutions, by the people who use them, should create trust. There is substantial evidence that, in reality, trust in public institutions is founded on both rational and relational factors (Job 2005, Taylor-Gooby 2006a).
Taylor-Gooby (2006a p 4) explores the question of why trust in the British National Health Service (NHS) is low at the same time that there is more funding, greater transparency and higher output, suggesting that the NHS is both efficient and competent. Further, trust in health professionals in the NHS remains high. Under these conditions trust in the NHS could be expected to rise, but it has not. From Taylor-Gooby’s perspective the problem is that the dual nature of trust, the cognitive and relational foundations, has not been addressed in a balanced way. The NHS:
reforms draw on a particular tradition in economics and economic psychology, transmitted through organizational theory, that stresses the role of cognitive factors. The central focus is on service providers and service consumers as market actors, behaving in response to their interests as they understand them. This is a world in which action is essentially driven by rational deliberative processes, both for the provider in seeking greater efficiency and for the consumer in deciding where to place trust, and in which the non-rational aspects of trust are discounted. Crucially, the stress on a competitive and consumerist logic may undermine a core component of trust, since the motivation of providers is declared to be self interest, in response to market signals, rather than public interest.
To explore in more depth the issue raised by Taylor-Gooby we will turn to some of the evidence available about public trust in MMR (mumps, measles and rubella) vaccine in the United Kingdom (UK). Vaccination is a private good for individuals who gain immunity, and a public good when a sufficiently large proportion of the population (for MMR 95% of the population) is vaccinated to make it impossible for the infective agent to survive in the community. In Britain, vaccination is voluntary and high vaccination rates declined after publication of a report suggesting a link between MMR vaccination, autism and inflammatory bowel disease (Brownlie & Howson 2006, Petts & Niemeyer 2004). The link between MMR vaccination and specific ‘objective’ health risks, and the necessity for parents to decide whether or not to have a child vaccinated, raises the question of trust in primary healthcare practitioners and the system in which they are embedded. In the UK the primary healthcare system has been reformed in line with the rational principles described above by Taylor-Gooby. No longer does the primary healthcare system rely on professionals to act voluntarily in the public good, instead a system of ‘auditable rules and procedures’ has been instituted to ensure they do so (Brownlie & Howson 2006). Specifically, general practices have targets to achieve in terms of the proportion of eligible children vaccinated and receive financial rewards and penalties based on target attainment (Brownlie & Howson 2006). For parents the financial incentives raised questions about the trustworthiness of general practitioners, and for general practitioners and health visitors who assisted parents to make informed choices, they raised questions about the trustworthiness of the information available through the NHS to parents (Brownlie & Howson 2006).
Petts and Niemeyer (2004) undertook a qualitative study of the information strategies used by parents of young children to make sense of the risks of immunisation, especially MMR. Necessarily, the question of who and what information to trust arose. The sources of information used included: official publications (40% of informants), discussion with family and friends (60%), and discussion with a general practitioner (25%) (Petts & Niemeyer 2004). In general, parents actively evaluated information received and the trustworthiness of the source was important. Broadly speaking there were two major foundations of trust against which sources were evaluated. One was empathy and concern for the wellbeing of the parent and child, and the second was scientific expertise. The most trusted sources, in descending order, were: the national Department of Health, the local health authority and local general practitioners (Petts & Niemeyer 2004 p 15). All of these were considered to combine concern for the wellbeing of parents and children and scientific expertise, although the question of immunisation targets made some informants sceptical of general practitioners’ concern for the parents and children. Friends and family were a frequent source of information and were considered trustworthy in terms of their concern but not in terms of their scientific expertise. Politicians and the mass media were towards the untrustworthy end of the scale on the grounds that they were viewed as ‘in it for themselves’ and not concerned about the wellbeing of the parents and children (Petts & Niemeyer 2004 p 15). Reflected in this study, as in others, are the two foundations of trust in public institutions: a focus on the wellbeing of the public and competent performance. The incentives created by immunisation targets was considered, by a subset of the public, as shifting the focus of general practitioners away from public wellbeing toward private benefits for general practitioners, therefore diminishing trust in general practitioners.
The discussion of MMR illustrates some of the major issues influencing public trust in a national health system, in particular the way that rational and relational issues appear to work in combination. For a broader analysis we can turn to the work of Straten et al. (2002) who developed a valid and reliable measurement scale, using standard scale development methods, to measure trust in the Dutch national health system. Out of this work emerged six dimensions of trust; that is, areas that were strongly related to public trust in the Dutch health system. The dimensions are:
3. Healthcare providers’ expertise, meaning the technical competence of providers.
6. Quality of cooperation, meaning that providers will work together for the patient’s benefit (Straten et al. 2002 p 231).
Straten et al. (2002 p 233) argue that only two of these dimensions, expertise and quality of care, relate to provider practice. Arguably, patient focus could also be included in this group of factors. The remaining dimensions relate to the health system. The overview of factors influencing trust is summarised by Straten et al. thus:
At the micro level, people are worried about information supply, and, in more general terms, about the behavior of providers. They wonder whether providers will pay enough attention to them, take enough time, and listen and take their problem seriously. At the meso-level, the worries are about cooperation in the health system among providers. Can the patient be sure that that providers will work together and cooperate? At the macro-level, people fear developments with possible consequences for accessibility and quality of healthcare. These last two dimensions emphasise the fact that public trust is not solely related to the relation between patients and healthcare providers.
From this study it appears that policy changes that impact on the six dimensions identified by Straten et al. have the potential to change the institution of healthcare in ways that influence public trust.
Does it matter if people lose trust in the health system? It most certainly does! In the example of MMR the challenge to trust lead to declining immunisation rates in the UK. In studies of risk communication, low levels of trust in an information source lead to less attention to public health risk communication messages (Walls et al. 2004) and lower rates of compliance with recommended public health actions under risk conditions (Lasker 2004). Straten et al. (2002 p 228) cite studies that suggest that a public that has low levels of trust in a health system will also: more frequently seek second opinions, use ‘alternative’ therapies and services, will try to use services with the ‘best’ reputations and avoid others, and demonstrate lower levels of compliance with recommended health treatments and practices. In Australia, in the early and mid 1990s immunisation rates were considered too low to prevent outbreaks of infectious diseases in children (Blair et al. 1997). Immunisation became a policy problem and a national immunisation strategy was successfully implemented. Current childhood vaccination rates are substantially higher (Australian Institute of Health and Welfare [AIHW] 1996 p xiii). In the early and mid 1990s there was a vigorous community-level debate about the pros and cons of childhood immunisation. Key anti-immunisation arguments were: vaccines are dangerous, vaccines cause other diseases, vaccines are ineffective, and that this information was being hidden from the public (Leask & Chapman 1998 pp 19–20). Each of these arguments challenged beliefs about the centrality of parent and child welfare to providers of immunisation services. Amongst parents who rejected immunisation for their children ‘general skepticism about health professionals’ (that is a low level of trust, or even distrust, in them) was an important factor in their decision (Leask & Chapman 1998 p 24). Trust was an important issue in the immunisation debate of the 1990s. Although trust was not discussed overtly, many of the discursive responses to the anti-immunisation interests were consistent with a trust-building approach.
Walls et al. (2004 p 148) suggest that the public can be thought of as having three major orientations to trust in institutions. The three categories constituting the concept of trust in an institution are: uncritical trust or emotional acceptance of the institution; distrust or rejection of the institution; and, critical trust or a condition in which people are aware that they rely on the institution to perform its functions but they are sceptical about its ‘effectiveness, motivations or independence’. Critical trust sits on a continuum between uncritical acceptance and outright rejection of an institution, and accommodates the reasoning of a large proportion of the population. Within critical trust are notions of tolerance, ambiguity, and the conditions under which an institution can be trusted.
POLICY RESPONSES TO THE PROBLEM OF TRUST
In this section I will discuss two broad categories of response to the problem of trust in public institutions. The first is the concept of clinical governance, the second the notion of universality.
Clinical governance
Health reform programs in UK (and Australia) are based on market principles and involve:
decentralising budgetary responsibility, linking resource allocation to treatment of service users and expanding patient or GP choice … The reforms are intended to encourage competition between providers to offer services that are attractive to users (or their proxies).
Altering incentives to service providers, introducing quality controls and mechanisms to make information available to the public, is viewed as a way of improving both efficiency and quality of services (Taylor-Gooby 2006b). The intended result is a high-quality, highly trusted health service (Alaszewski 2003). The strategy is embodied in the concept of clinical governance. Clinical governance, in the NHS, has four components:
1. Clear lines of responsibility and accountability for overall quality of care;
2. A comprehensive program of quality improvement activities;
3. Clear policies aimed at managing risk; and,
4. Effective procedures to identify and remedy poor performance.
It is argued that the NHS has shifted from a system in which patients had blind trust in health professionals and services to one in which they have the information to make rational decisions about trustworthiness.
As argued earlier, trust has both rational and affective elements. Rational elements are, broadly speaking, indicators of competence; the ability to provide effective and efficient services. The process of establishing trust in competence is fundamentally a rational one, based on good information. Affective elements are, in general, matters of caring in which people feel valued and respected, and have positive feelings towards the service (Taylor-Gooby 2006a). Trust in caring is based on experience rather than information. The UK reform program, encapsulated in clinical governance, established dynamics that risk creating a system ‘that buttresses one of the twin pillars of trust, but ignores the other … The outcome may well be a reformed health service that actually delivers the goods and which makes sense in terms of rational judgement, but which no one likes or trusts’ (Taylor-Gooby 2006b p 101). Davies and Mannion (1999) argue that clinical governance requires a balance between checking (measuring and monitoring performance) and trusting (establishing a quality focused and learning-oriented culture). In this context ‘trust is not the abandonment of controls but is instead the re-emphasising of the power of intrinsic motivations over external drivers; controls based on shared values and goals, rather than incentives and fault finding’ (Davies & Mannion 1999 p 14).
Communities’ experience of health institutions
The experience communities have of relating to health institutions can provide evidence of competence upon which a rational decision to trust can be made, but it also provides an emotional experience of being respected and cared for which, it is argued, is a parallel but equal foundation for trust (Taylor-Gooby 2006a). For example, in their qualitative study of the Health and Safety Executive (HSE) in the UK, Walls et al. (2004 p 139) found that the high levels of trust expressed in the HSE were primarily based on the perception that the agency was altruistic – it worked for the public good. Informants were critical of some aspects of HSE work but still regarded it as trustworthy. The acts criticised could be explained as the fault of government and therefore did not threaten the perception that the HSE itself was motivated to achieve public good (Walls et al. 2004).
Using 2001 Eurobarometer data, Hudson (2006) sought to test whether satisfaction with life, across European Union (EU) countries, is related to the quality of institutions and whether this is related to different levels of trust. He found that trust in some key institutions, for example, government and the law, impacted positively on people’s wellbeing. However, negative experience of any institution included in the study lead to a generalised loss of trust in all institutions (Hudson 2006 p 58). Presumably, the wellbeing effects of trust reflect a positive experience of institutions. Hudson (2006) also argues that experience of institutions appears to be socially patterned. For example:
… direct experience of institutions is likely to vary between people in a systematic manner because people of different backgrounds may have different experiences with institutions such as police, the law and voluntary organisations. For example, a young man, particularly from a low income family, living in a large city can often be expected to have different interactions with police than a relatively wealthy, older man or woman, living in a rural village. In part an institution, such as the police, may itself differ between urban and rural locations and in part the agents of that institution may treat different people in different ways, giving more respect to rich, successful and educated citizens.
Lasker (2004) also argues that responses to public health authorities in regard to major risks are likely to be similarly patterned. It appears that trust in institutions is in part a consequence of factors internal to the institution and in part a consequence of characteristics of the population.
A key issue is whether we have evidence of policy approaches that create experiences of health institutions that most citizens value. Using Swedish survey data Kumlin and Rothstein (2005) show that welfare institutions that provide universal services increase interpersonal trust (used as a proxy for social capital) whereas those that are needs-based undermine it. The difference can be explained in terms of citizen experience of procedural justice in their relationships with the institution. Universal services are targeted at an entire population or an easily identified segment of it. Needs-based programs require institutional employees to interpret a body of regulation to ascertain whether an applicant qualifies, and this can be done unfairly. Applicants might also attempt to manipulate information to influence the eligibility decision.
Because of these complex and controversial decision-making processes, needs testing and bureaucratic discretionary power are often more difficult to reconcile with principles of procedural justice, compared with universal public services. Because selective welfare institutions must test each case individually, they are to a greater extent subject to suspicions of cheating, arbitrariness, and discrimination compared with universal public agencies (Kumlin & Rothstein 2005 p 349).
The Swedish study obtained data on frequency of contact with universal and needs-based services and measured interpersonal trust as a proxy for social capital. It demonstrated that contact with universal services increased measured trust and needs-based services decreased it (Kumlin & Rothstein 2005). The causal link appears to be whether or not people felt they had been treated fairly, and received entitlements due to them, through their contact with the institutions.
The argument that trust in institutions is an outcome of cognitive assessment of competence plus positive affective experience is common in the political science, risk management and social science literature. Furthermore, there is some evidence of the connections between the two, the consequences of trust, and the causal pathways leading to it.
FUTURE DIRECTIONS FOR HEALTH POLICY DEVELOPMENT
The policy trend in Western democracies has been to view trust in public institutions and trust in government as based on rational consideration of performance evidence. Hence, in the health sector we have a performance indicator explosion, patients’ charters, league tables of institutions, clinical governance programs, and so on (Davies & Lampel 1998, Davies & Mannion 1999, Taylor-Gooby 2006a). However, there is now a body of work indicating that competent performance is insufficient to create public trust in institutions. Competent performance is very important but it needs to be accompanied by a positive experience for the public interacting with the institution. Public trust in institutions is unlikely to increase so long as policy is based exclusively on the concept of service users as market actors (Taylor-Gooby 2006a).
If performance indicators, that gather evidence about practice to reward and sanction institutions based on the competence of their performance, are an insufficient mechanism to achieve trust in health institutions, additional measures need to be considered. One area for policy development is the relationship between health service managers and the people that provide front-line services. Davies and Lampel (1998) argue that information systems used to monitor performance need to be accompanied by management systems that create a positive experience of excellence for service providers, which in turn translates into a positive experience for citizens. They argue that the principles of learning organisations need to be promoted such that ‘… knowledge is mobilised at the operational level to improve performance rather than wielded at a managerial level to demand (but not facilitate) ill-specified improvements’ (Davies & Lampel 1998 p 161).
Learning organisations function in ways that are very different from traditional bureaucracies, and are necessarily managed differently. Development of learning organisations in a health system would be a significant policy intervention in its own right.
In organisational studies trust and the exercise of power to constrain untrustworthy behaviour are increasingly described as necessary companions (Bachmann 2001, Das & Teng 2001, for example). Davies and Mannion (1999 p 14) describe this as ‘striking a balance between checking and trusting’, Braithwaite (1998) as ‘institutionalising distrust, enculturating trust’. Both of these papers argue for organisational systems that mobilise altruism to achieve the benefits of trust for institutions whilst continuing to monitor and identify breaches of trust. The dual approach of enhancing trust in order to facilitate cooperative action and developing control mechanisms to constrain breaches of trust, is in fact how health services structure cooperative relations (Walker et al. 2007). Braithwaite (1998 p 353) argues for strong regulatory institutions that utilise the regulatory pyramid; with virtuous actors they use dialogue and persuasion, with calculating but rational actors they use deterrence, and with incompetent or irrational actors they use incapacitation. He argues:
This is the basic idea of the regulatory pyramid: Trust first, and thereby get the efficiency benefits of trust in most cases, but motivate trust as obligation by signaling very clearly a preparedness to escalate intervention to progressively less trusting interventions when trust is abused.
Of course it is not possible to get the benefits of a ‘checking and trusting’ strategy unless both sides are attended to simultaneously. Policy needs to simultaneously facilitate learning organisations and establish appropriate regulatory regimes, as two sides of the same coin.
The apparent difference in citizen experience of universal and needs-based services appears to be an effect of ‘street-level bureaucrats’ exercising the discretion required of them in ways that result in citizens feeling unfairly treated. When a population or population segment eligible for a service is clearly defined and service providers’ role is to facilitate access, citizens are more likely to experience the service as respectful and fair. If, however, the role of service providers is to separate eligible from ineligible citizens, using guidelines that need interpretation in light of individual cases then it is less likely that citizens will experience the service as fair. Although dealing with citizens on a case-by-case basis may sound fair, the evidence suggests that it may not be experienced that way by large sections of the population. Evidence from Sweden suggests that universality of services, where this is possible, may have a generalised effect on trust in institutions and the state, that may not be achievable through needs-based services.
In Australia, trust in health institutions is, in Kingdon’s (1984) terms, a condition that has not yet overtly become a policy problem. Having said that, trust appears to have been a sub-text in the Australian National Immunisation Strategy, evident but not ‘spoken about’. It can be argued that a number of policy interventions, such as those that facilitate patient-centred care and team-based services (these are qualities of health systems Straten et al. [2002] identified in which the public trust), implicitly support the development of trust in health institutions. However, because they are implicit they can also be accompanied by initiatives likely to diminish trust.
