Social Trust and the Management of Risk

George Cvetkovich, Ragnar E. Lofstedt

• Publisher: Routledge
• Year: 2013
• eBook ISBN: 9781134190218
• Edition: 1
• Topic: Business
• Subtopic: Insurance
• Index: Business

Social trust is a crucial issue to many aspects of modern society. Policy makers continually aspire to winning it and corporations frequently run the risk of losing it. The 'trust deficit' raises vital questions and problems to which until recently there have been few answers or solutions.

Experts from both sides of the Atlantic explore the importance for trust of various influences, from individual perceptions to organizational systems, and consider the conditions involved in building or undermining trust. Several authors examine practical hazard management issues, including medical vaccination programmes and popular participation in pollution control and waste management as strategies for enhancing social trust.

This book provides insightful analysis for researchers and students of environmental and social sciences and is essential reading for those engaged in risk management in both the public and private sectors.

1
Social Trust and Culture in Risk Management

Timothy C. Earle and George Cvetkovich¹

We have argued in a variety of reports over the past few years (Cvetkovich and Earle, 1992; Earle and Cvetkovich, 1994a, b, c; Earle and Cvetkovich, 1994b; Earle and Cvetkovich, 1994c), and in a recent monograph (Earle and Cvetkovich, 1995), that the development of useful ways of understanding social trust has been retarded by an all-but-universal acceptance of a dominant traditional interpretation. According to that tradition, social trust is a rational process based on competence and responsibility. Further, social trust traditionally is empirically-based, requiring evidence of competence and responsibility. In our critique of traditional social trust (Earle and Cvetkovich, 1995) we maintain that a basic function of social trust is the reduction of cognitive complexity. Because it is empirically-based, the traditional account of social trust increases complexity instead of supplying simplicity. We also argue that social trust must accommodate cultural variation. But traditional social trust is itself culture-specific, recognizing only competence and responsibility as legitimate bases. We conclude that, due to these and other flaws, the traditional interpretation acts to undermine social trust and to generate its functional equivalent, social distrust.

In this chapter we describe two studies that test the validity of the traditional account of social trust relative to our proposed alternative. In our formulation (Earle and Cvetkovich, 1995), social trust is based on value similarity, with the value basis varying across people, contexts and time. This cultural-values interpretation of social trust is a ‘groundless’ social trust, requiring no justification for its use. Instead of being deduced from evidence, social trust is inferred from value-bearing narratives.² People tend to trust other people and institutions that ‘tell stories’ expressing currently salient values, stories that interpret the world in the same way they do. Within this general construction, then, traditional social trust is simply a special case.

Social Trust and Cultural Values

Thinking about social trust as a ‘construction’ – as a human product that has evolved as a tool for achieving human goals – enables us to move beyond the rationalist tradition.³ Social trust is a social construction that is based on varying sets of cultural values – the values of specific persons and institutions living in certain times and places – as expressed in cultural narratives. This is the fundamental contrast. Between cultural singularity (rationalist, the presumed normative) and cultural pluralism. Between what it is claimed by authorities that people should do and what it is they actually wind up doing.

The simple contrast between cultural singularity and cultural pluralism provided the basis for the two studies we describe below. In both studies, respondents were asked to read texts that consisted of two parts, one that varied across respondents and one that stayed the same. In the texts, descriptions of individuals and institutions were the same for all participants. The variable part of the texts consisted of brief policy descriptions written to express different sets of cultural values. Some respondents, for example, may have read policies expressing value set A while others read policies expressing value set B (or G or D). After reading a text, each respondent was asked to produce a social trust judgement for the institution described in it. If respondents’ social trust judgements were based on the competence and responsibility of individuals and institutions (qualities we attempted to hold constant across texts), we would expect little variation in social trust judgements. This would be interpreted as support for the traditional interpretation of social trust. But if participants’ social trust judgements were based on expressed cultural values, we would expect wide variation in judgements – specifically as a function of the similarity between the values expressed in the text and the values favoured by respondents. This would be interpreted as support for the cultural values interpretation of social trust.

To demonstrate that individuals base their social trust judgements on varying sets of cultural values, we must devise ways to reliably describe cultural pluralism – among individuals, their judgements and expressions. We present two ways of doing this. In Study One, we draw upon the cultural theory developed, in the main, by Douglas and Wildavsky (Douglas and Wildavsky, 1982), Dake (Dake, 1992), Thompson (Thompson, Ellis and Wildavsky, 1990), and Rayner (Rayner, 1992). In Study Two, our approach is based on the four elementary forms of sociality described by Fiske (Fiske, 1991a; Fiske, 1991b; Fiske, 1992).

Study One

Method

The Survey Experiment. The fundamental claim of our cultural-values interpretation of social trust is that social trust is based on value similarity: people tend to trust other people and institutions that ‘tell stories’ expressing currently salient values, stories that interpret the world in the same way they do. One way to test this cultural-values hypothesis is through the use of what we call the ‘survey experiment’, a procedure in which survey-like methods and materials are used – but with a twist: The contents of the materials (questionnaires) are manipulated so that they create the structure of an experiment. Thus, different forms of a basic questionnaire are used to create cells in an experimental design.

One version of the survey experiment, employed successfully by several research groups, incorporates systematic variations in both questionnaires and respondents. Clary and colleagues (Clary, Snyder, Ridge, Miene and Haugen, 1994), for example, varied the motives expressed in messages and the motives that were personally relevant to respondents (eg knowledge, social adjustment, value expression, ego defence and utilitarian). These researchers demonstrated that matches between message and respondent motives were more persuasive than mismatches. Matches on motives also produced higher levels of trust. Similarly, Arad and Carnevale (1994) showed that partisan individuals judged the trustworthiness of a neutral third party on the basis of whether his proposal matched their position. By using a method that can accommodate variation in both experimental materials and respondents, studies like these can explore interactions between the two sets of variables. The basic assumption of these studies, and of ours, is that the effects of a message, for example, depend critically on how it is interpreted by the person reading it. In our studies, we examined the effects of the culture of individuals on their interpretations of risk messages that expressed varying sets of cultural values. Based on our cultural-values theory of social trust, we expected matches between the culture of individuals and the culture of messages to produce higher trust judgements than mismatches.

The Culture of Individuals. Our first attempt to describe and measure the culture of individual respondents was based primarily on the work of Karl Dake, who has expressed the categories (or ‘cultural biases’) of cultural theory in the form of questionnaire items (Dake and Wildavsky, 1990; Dake, 1992). There are three primary categories in cultural theory: Egalitarian, Individualistic and Hierarchical. A fourth category, Fatalism, was not included in this study. The items we used, given in Table 1.1, consisted of a mixture of those described by Dake and Wildavsky (1990) and our own. Standard statistical analyses (Jessor and Hammond, 1957; Judd, Jessor and Donovan, 1986) were used to select the items, based on data from 449 respondents.

Individual respondents were assigned scores on each of the three scales based on the mean of their responses to the items contained in a scale. Thus, each respondent had an Egalitarian score, an Individualistic score and a Hierarchical score. Respondents were classified in this way: if a respondent’s

Table 1.1 Items Used to Measure Culture of Individuals: Culture Theory^a

Egalitarian score was greater than or equal to his/her Individualistic score and greater than or equal to his/her Hierarchical score and, in addition, was greater than or equal to 4 (the midpoint on the response scale), then that individual was classified as Egalitarian. Analogous procedures were used to classify respondents as Individualistic and Hierarchical. A total of 402 respondents were classified into the three cultural categories: 282 Egalitarian; 76 Individualistic; 44 Hierarchical. These individuals were unpaid volunteers recruited from both university student and general populations in western Washington State. Their average age was 28.3 years, ranging from 15 to 72. Gender distribution was even, with 50.4 per cent male, and there was no relation between gender and culture.

The Culture of Messages. The messages used in this study were written in the form of brief, simulated newspaper stories.⁴ Each story was composed of two parts. The first part was a core story that was read by all respondents. Since the data were collected late in 1992, a (fictional) story concerning the prospective high-level nuclear waste management policy of the incoming Clinton administration was selected as timely and moderately involving to a broad range of potential respondents. A key part of the core story stated that ‘a new decision-making procedure’ will be used. The second part of each message provided a description of that procedure in one of three forms designed to succinctly express the relevant Egalitarian, Individualistic or Hierarchical values identified in cultural theory.

Social Trust and Trust Values. Our cultural-values interpretation of social trust claims that social trust is based on value similarity: individuals tend to trust institutions that express currently salient values. A simple test of this notion would include: a) a measure of judged value similarity; b) a trust judgement; and c) a measure of the relation between a) and b). Our measure of judged value similarity consisted, first, of the following question referring to the nuclear waste management message the respondents had read:

The proposed new federal organization described in the story is the Nuclear Waste Management Agency (NWMA). Based on what you have read here, how do you feel about the NWMA?

This question was followed by a six-item trust-values scale. Each of the six items consisted of a seven-point response scale anchored by bipolar descriptors of the relations between the story and the respondent:

Standard procedures (Jessor and Hammond, 1957; Judd, Jessor and Donovan, 1986) were used to select these items from an initial large pool. A trust-values score, consisting of the mean of the six response scales, was computed for each respondent. Our measure of social trust was a single judgement made in response to this question:

Based on what you have read here, would you say you would trust the NWMA?’

The correlation between respondents’ social trust judgements and their trust-values scores gives us a measure of the relation between the two, a simple test of our hypothesis that social trust is based on shared cultural values. In addition, of course, the social trust judgements and the trust-values scores are the dependent variables in our experimental test of this hypothesis.

Procedure. The newspaper-style stories on nuclear waste management and the questionnaire items measuring social trust, trust values and the culture of individuals were presented to respondents in the form of a booklet. The booklet contained an unrelated task in addition to the one described in this report. The nuclear waste management task appeared first in half the booklets. Order had no effects on respondents’ judgments, and it is ignored here. In the booklet, a general set of instructions dealing with the mechanics of responding to items was followed by the first task and the questionnaire items related to it; then came the second task and its items. The social trust items described in this report are a subset of the nuclear-waste-management-task items. The items used to measure respondent culture followed the two tasks, and the booklet ended with a set of questions about the respondents’ backgrounds. The booklets were randomly assigned to respondents, and the respondents worked on the booklets individually, self-paced.

Results

The mean social trust and trust values judgements for the nine experimental groups (three respondent cultures crossed with three story cultures) are given in Table 1.2. The striking thing about this table of means is the dominance of the main diagonal: cultural matches between respondents and stories produced the highest judgements of social trust and trust values for all groups of respondents. This very strong result is reflected in significant analyses of variance for the respondent-culture-by-story-culture interactions (F(4,393) = 8.39, p < .001 for social trust; F(4, 393) = 5.61, p < .001 for trust values).

Table 1.2 Social Trust and Trust Values by Culture of Respondents and Culture of Stories: Culture Theory

These experimental results provide strong support for the cultural-values social trust hypothesis: the respondents in this experiment did tend to give higher levels of social trust to institutions that shared their cultural values. In addition, there was a high level of correlation between social trust and trust values across all respondents (r = .66, p < .001). This result indicates that individuals have some means of comparing the values expressed in a story with the values expressed in their favoured cultural narratives. And it suggests that they may use that information in making judgements of social trust.

Study Two

Method

The main purpose of our second study was the same as that of the first –...

Table of contents

• Cover
• Title
• Copyright
• Contents
• List of Tables, Figures and Boxes
• Acronyms and Abbreviations
• Foreword
• Preface
• Introduction
• 1 Social Trust and Culture in Risk Management
• 2 Risk, Trust and Democratic Theory
• 3 Perceived Risk, Trust, and Democracy
• 4 The Attribution of Social Trust
• 5 Trust Judgements in Complex Hazard Management Systems: The Potential Role of Concepts of the System
• 6 Environmental Regulation in the UK: Politics, Institutional Change and Public Trust
• 7 Perceived Competence and Motivation in Industry and Government as Factors in Risk Perception
• 8 Institutional Trust and Confidence: A Journey into a Conceptual Quagmire
• 9 Trust and Public Participation in Risk Policy Issues
• 10 Social Trust, Risk Management, and Culture: Insights from Native America
• 11 Who Calls the Shots? Credible Vaccine Risk Communication
• 12 Conclusion: Social Trust: Consolidation and Future Advances
• References
• Index