Journal of Environmental Management 137 (2014) 61e68
Contents lists available at ScienceDirect
Journal of Environmental Management journal homepage: www.elsevier.com/locate/jenvman
Social trust, risk perceptions and public acceptance of recycled water: Testing a social-psychological model Victoria L. Ross a, *, Kelly S. Fielding b,1, Winnifred R. Louis c, 2 a
Smart Water Research Centre, Griffith University, Gold Coast Campus, Edmund Rice Drive, Southport, QLD 4222, Australia Institute for Social Science Research, The University of Queensland, Level 4, GPN3 (Building 39A), Campbell Road, St Lucia, QLD 4072, Australia c School of Psychology, The University of Queensland, McElwain Building, St Lucia, QLD 4072, Australia b
a r t i c l e i n f o
a b s t r a c t
Article history: Received 6 August 2013 Received in revised form 11 January 2014 Accepted 17 January 2014 Available online 4 March 2014
Faced with a severe drought, the residents of the regional city of Toowoomba, in South East Queensland, Australia were asked to consider a potable wastewater reuse scheme to supplement drinking water supplies. As public risk perceptions and trust have been shown to be key factors in acceptance of potable reuse projects, this research developed and tested a social-psychological model of trust, risk perceptions and acceptance. Participants (N ¼ 380) were surveyed a few weeks before a referendum was held in which residents voted against the controversial scheme. Analysis using structural equation modelling showed that the more community members perceived that the water authority used fair procedures (e.g., consulting with the community and providing accurate information), the greater their sense of shared identity with the water authority. Shared social identity in turn influenced trust via increased source credibility, that is, perceptions that the water authority is competent and has the community’s interest at heart. The findings also support past research showing that higher levels of trust in the water authority were associated with lower perceptions of risk, which in turn were associated with higher levels of acceptance, and vice versa. The findings have a practical application for improving public acceptance of potable recycled water schemes. Ó 2014 Elsevier Ltd. All rights reserved.
Keywords: Recycled water Risk Fairness Identification Trust
1. Introduction Positive public perceptions and acceptance of water reuse are now recognized as key factors for the successful introduction of wastewater reuse projects (Dolnicar and Hurlimann, 2011; Dolnicar et al., 2011, 2010; Friedler and Lahav, 2006; Nancarrow et al., 2008). Regardless of the strength of the scientific evidence, public opposition can cause wastewater reuse projects to fail at any stage of their implementation (Friedler and Lahav, 2006; Uhlmann and Head, 2011). There is ample evidence of this from potable water (i.e., for drinking water purposes) reuse projects that have failed to be implemented in the U.S. and Australia. For example, in the 1990s the San Diego County Water Authority’s plans to mix recycled water with the drinking water supply was subject to negative
* Corresponding author. Tel.: þ61 7 5552 7808; þ61 407 583 054 (mobile). E-mail addresses: victoria.ross@griffith.edu.au (V.L. Ross), k.fi[email protected] (K.S. Fielding), [email protected] (W.R. Louis). 1 Tel.: þ61 7 3346 8725. 2 Tel.: þ61 7 3346 9515. http://dx.doi.org/10.1016/j.jenvman.2014.01.039 0301-4797/Ó 2014 Elsevier Ltd. All rights reserved.
public campaigns highlighting the health risks associated with the project and dubbing it “Toilet to Tap”(Po et al., 2003). Despite significant investment the project was abandoned (Hurlimann and McKay, 2004). In South East Queensland, Australia, in particular, where the combination of rapid population growth and prolonged drought placed pressure on the government to seek out rainfallindependent water supplies, there have been a number of noteworthy examples of proposed potable wastewater reuse schemes that have failed to be implemented. During the mid-1990s, potable recycled water projects proposed by the Maroochy and Caloundra councils in Queensland were both strongly opposed by their communities and ultimately rejected (Uhlmann and Head, 2011). Public opposition to the projects was fuelled by campaigns from opposition groups such as CADS (“citizens against drinking sewage”) who warned of alleged health risks associated with drinking recycled water. In 2006, in the droughtstricken regional city of Toowoomba, Queensland, a referendum was held for a proposed potable reuse scheme to augment dam supplies. The proposed scheme created intense opposition and publicity from the CADS group, who cited health risks and a negative impact on Toowoomba’s image as their opposition to the
62
V.L. Ross et al. / Journal of Environmental Management 137 (2014) 61e68
project. The project became politicized, and as noted by the mayor of Toowoomba at the time, moved the focus from water to politics and vested interests (Hurlimann and Dolinicar, 2010). Even with severe water shortages and dam levels dropping to record lows of only 23%, the scheme was abandoned after 62% of residents voted against it (Hurlimann and Dolnicar, 2010). More recently, the Western Corridor Recycled Water Project in South East Queensland (the largest water project to be built in the Southern Hemisphere), was put on hold, despite completion of $2.5 billion of infrastructure. Extensive media speculation regarding possible health risks impacted significantly on community confidence (Roberts, 2008) and in response to the drop in public confidence (as well as unexpected rainfall which restored dam levels) the government changed its policy so that recycled water would only be introduced to drinking water supplies in an emergency, that is, when combined dam levels in the region drop to below 40% (Queensland Water Commission, 2009). Given the critical importance of public perceptions to the implementation of potable reuse schemes, the current study examines the factors that are related to public risk perceptions and acceptance of recycled water. Specifically, the study develops a model that investigates the role of trust in predicting risk perceptions and acceptance and the factors that help to promote trust. In doing this the study makes an important theoretical contribution to the literature on the social dimensions of recycled water management and also contributes more broadly to the study of environmental management and to the social psychological literature on trust.
management, it has been argued that in order for water management schemes to be successful, the community requires trust in the responsible authorities to deliver them a safe water supply (Hurlimann and McKay, 2004; Marks and Jadoroznyj, 2005). The relationship between trust, risk and acceptance has been empirically explored in the risk communication literature in the context of new technologies (Eiser et al., 2002; Pavlou, 2003; Siegrist, 2000; Siegrist et al., 2007), and more specifically in relation to public acceptance of potable reuse (Hurlimann et al., 2008; Nancarrow et al., 2009). Thus, past research and theory provide a strong foundation for the hypothesis that greater trust in the water authority to deliver safe recycled water will be associated with lower risk perceptions and greater acceptance of the recycled water scheme. What past research has not explored though, are the factors that underpin trust in a water authority. This is a critical question given that water resources are likely to become more vulnerable in the future and water authorities will need to explore alternative water sources, including recycled water (Dolnicar and Hurlimann, 2011; Hurlimann et al., 2009). Understanding how trust can be developed will be essential information to aid in this process. In the present model, we explore three inter-related antecedents: perceived procedural fairness, social identity, and source credibility. We focus on these variables as, theoretically, they have been linked to trust, although these linkages have not previously been tested in relation to recycled water.
2. The hypothesized model
According to the procedural justice literature, the belief that one has been treated fairly by authorities enhances acceptance of legal decisions, obedience to laws, and evaluations of public policies; while the belief that one has been treated unfairly prompts protest behaviour (Lind, 2001; Tyler, 2001; Van den Bos and Lind, 2002). Within the context of water management, there is evidence that fair procedures are a major predictor of acceptability of the schemes and compliance with urban and rural water management and policy issues (Hurlimann et al., 2008; Nancarrow et al., 2002; Syme et al., 1999), including community intentions to drink water from potable reuse schemes (Nancarrow et al., 2009). Hence, past research suggests that the more community members consider a recycled water scheme to be fair, the greater their acceptance of the scheme. The relational model of authority (Tyler and Lind, 1992) provides a theoretical explanation for these findings, proposing that people care more about how decisions are made than they do about the actual decisions because procedural treatment provides them with important information about their relationship with authorities (Skitka and Mullen, 2002). Tyler and Degoey (1996) suggest that the way people are treated by authorities provides them with information about whether they are respected members of the group and whether they should feel pride in the group as a whole. In other words, fair procedures indicate to community members whether and how much they share an identity with authorities. Shared social identity in turn increases the likelihood that group members trust an authority and accept their decisions (Tyler and Degoey, 1996; Williams, 2001). Hence, according to the relational model of authority, shared social identity is a key mechanism by which fair procedures influence trust in authorities. It was therefore hypothesized that perceiving that the water authority treats community members fairly predicts a greater sense of shared identity between community members and the water authority. In the current study shared social identity is reflected by the extent to which people see the water authority as a member of their group (i.e., the community)
Below we outline the theoretical and empirical basis for relationships proposed in the hypothesized model. 2.1. Risk perceptions, trust and acceptance A strong body of research has shown the centrality of social factors in water management (Dolnicar et al., 2010; Dolnicar and Schäfer, 2009; Hurlimann et al., 2009; Lienert et al., 2013). In the failed reuse cases discussed above, the public perceived the possible health risks associated with recycled water as unacceptable despite the reassurances provided by authorities and by scientists (Dolnicar et al., 2010; Hurlimann and Dolnicar, 2010; Uhlmann and Head, 2011). The conclusion that perceived risk is a key predictor of acceptance of recycled water management schemes accords with the broader risk literature that shows a clear relationship between risk perceptions and acceptance (Eiser et al., 2002; Mankad and Tapsuwan, 2011; Robinson et al., 2012; Siegrist et al., 2007). In turn, in the risk communication literature, it is generally agreed that trust in authorities to manage risk is a critical factor in the perception and acceptance of risks (Earle et al., 2007; Lofstedt and Cvetkovich, 2008). Trust is a multidimensional, complex construct (Fischhoff, 1999; Grabner-Kräuter and Kaluscha, 2003), but in the current study we use a specific operationalization drawn from the literature (Frewer et al., 1996; Lewicki et al., 2006; Rousseau et al., 1998; Siegrist et al., 2000): a psychological state comprising the intention to accept vulnerability based upon positive expectations of the intentions or behaviour of the authority responsible for the recycled water scheme. Siegrist et al. (2000) have argued that many individuals lack the resources such as knowledge, time and interest to make decisions and take action in relation to science and technology and therefore they rely on trust in the relevant authorities or government agencies to make decisions. In the context of urban water
2.2. Predictors of trust: fairness, identity, credibility
V.L. Ross et al. / Journal of Environmental Management 137 (2014) 61e68
and whether they perceive a sense of shared values with the authority, both of which are important dimensions of social identity (Hogg and Abrams, 1998; Tajfel and Turner, 1986; Tyler and Degoey, 1996). Past research has also conceptualized value similarity as a reflection of social bonds and has shown, consistent with our reasoning that it is significantly related to trust (Siegrist et al., 2000, 2003; Vaske et al., 2007). This relational framing of trust accords with social identity theory which posits that social groups provide members with a social identity: a definition of who one is and evaluation of what that entails (Hogg and Abrams, 1998; Tajfel and Turner, 1986). Importantly, research from the social identity perspective has shown that people who belong to our groups (i.e., ingroup members) are perceived in more desirable ways (e.g., more trustworthy or honest) than outgroup members and that shared social identity reduces uncertainty and produces in-group trust (Brewer, 2007; Hogg, 2003; Hogg and Abrams, 1998). Drawing on social identity theory we further reason that in the context of acceptance of recycled water schemes, the relationship between shared social identity and trust will be mediated by source credibility, that is, the extent to which the water authority is perceived to be competent and has the public’s interests rather than vested interests at heart (Frewer et al., 1996; Siegrist et al., 2003; Tyler and Degoey, 1996). Past research has demonstrated that sources are trusted more when they are perceived to be competent and credible (Frewer et al., 1996; Sutherland et al., 2013; Tyler and Degoey, 1996). Given the technical nature of water recycling processes, and the potential impacts on the health of the population if the process were to go wrong, this variable seems likely to be a critical predictor of trust. Our analysis though, does not see this variable as an independent predictor but rather, we believe that judgments of credibility will be influenced by social identity processes. There is compelling evidence that links shared social identity to more positive evaluations of ingroup members across a variety of group relevant dimensions (Hogg, 2007; Tanis and Postmes, 2005; Williams, 2001). We predict, therefore, that social identity would also influence judgments of the source credibility of water authorities and in turn, trust. We reason that the more that people perceive a shared social identity with the water authority (perceive the water authority as a member of their community who shares their values), the more people will perceive the water authority as a competent entity that has the interests of the group at heart. To summarize the model: We hypothesize that community members who perceive that the water authority uses fair procedures will feel a greater sense of shared social identity with the water authority and this will be related to greater trust in the water authority via increased perceptions of credibility. Greater trust in turn will predict lower risk perceptions and greater acceptance of recycled water. Interestingly, previous research has often drawn distinctions between instrumental and relational approaches to trust, contrasting competency judgments and social relational concerns as the main basis for trustworthiness (Saparito et al., 2004; Tyler and Degoey, 1996; Zaheer and Venkatraman, 1995). Our model, suggests, however, that instrumental and relational perspectives may be strongly inter-related, with social identity influencing credibility judgments. In this way our propositions are consistent with previous researchers who argue for the compatibility of instrumental and relational approaches to understanding trust (Cho, 2006; Edwards and Kidd, 2003; Kerkhof et al., 2003; Tyler and Degoey, 1996).
63
To our knowledge the current study is the first study to advance a theoretical model of the role of trust in authorities and its antecedents and how that relates to risk perceptions and acceptance of proposed recycled water management schemes. Moreover, as the research was conducted in a context in which residents were about to vote for or against a wastewater reuse scheme, the issue was not hypothetical but instead very salient to them. Past research seeking to understand public responses to potable recycled water schemes has tended to focus on the hypothetical introduction of recycled water (e.g., Dolnicar and Schäfer, 2009; Nancarrow et al., 2008). Hence, the research provides a unique opportunity to gain insight into the social psychological variables that are important drivers of the acceptance or rejection of recycled water projects. Through identifying the variables that are related to trust in water authorities, the research can provide critical information to water authorities about how they can develop greater public trust and thereby promote more positive responses to potable recycled water schemes. 3. Method 3.1. Study context The hypothesized model was tested in the context of a proposal to introduce a potable recycled water scheme in the regional city of Toowoomba. Known as “the Garden City” for its parks and gardens, Toowoomba is located approximately 127 km west of Queensland’s capital city of Brisbane. With an estimated population of 157,023 Toowoomba is Australia’s second most populous inland city. As described earlier in the introduction, after years of drought and dam levels dropping to 23%, the water authority proposed a potable wastewater reuse project to augment dwindling supplies. The scheme was strongly opposed by the community action group “CADS” and later rejected by the public in a referendum. The data for this research was collected only weeks before the referendum, so the issue was extremely salient to the community at the time (Hurlimann and Dolnicar, 2010; Price et al., 2012). At the same time, the present study asked participants a broader question than the referendum, which focused on support for adding recycled water to the city’s water supply. The present study examined willingness to use recycled water over a range of contexts (e.g., outdoors, washing, showering, cooking and drinking), allowing a broader analysis of predictors of support for recycled water. 3.2. Participants and procedure Four hundred and one residents from Toowoomba were interviewed by telephone several weeks before the referendum for the proposed recycled water project was held. An attempt was made to obtain equal numbers of males and females, resulting in 220 (55%) females and 181 (45%) males. Of the 923 people asked to participate, there were 522 refusals, indicating a 43% response rate. The age of respondents were spread relatively evenly across predetermined grouped age categories which ranged from “18e19 years”, “20e24 years”, “25e29 years”, “30e39 years”, “40e49 years” etc, up to and including “75 years and older”. Of the 401 respondents, two cases were excluded because of significant missing data, and 19 cases were excluded because they were multivariate outliers (i.e., cases with unusual combination of scores on two or more variables that differ from the range of observations exhibited for other cases). This meant that the final number of respondents included in the analysis was 380. The study was conducted by a Social Research Centre using Computer Assisted Telephone Interviewing (CATI), with each interview taking approximately 15 min to complete. The sample
64
V.L. Ross et al. / Journal of Environmental Management 137 (2014) 61e68
was randomly selected from telephone listings. Telephone interviewers explained that they were conducting an independent study on Toowoomba residents’ thoughts and beliefs about their water supply. Participants were initially asked if they were connected to Toowoomba’s main water supply (and would therefore be affected by the proposed potable reuse scheme). Interviews did not proceed for those not connected. Before being asked about the project, participants were provided with details about the proposed recycled water scheme including information that the wastewater would be treated to a standard higher than safe drinking standards and the process by which it would be released to the dam and then treated again prior to being added to the water supply. This explanation was followed by questions in a format which required participants to rate their level of agreement. Participants were then asked to provide demographic information (i.e., categories of age, income, education, and household type).
Table 1 Scale reliabilities of constructs. Construct
Items
Alpha
Source credibility
The water authority is competent enough to manage our water supply. The water authority provides the Toowoomba community with everything they want to know about their water supply. The water authority does not listen to concerns raised by people like me. The water authority acts in the public interest when it comes to water quality. The water authority makes fair decisions about water provision. The water authority makes an effort to treat everyone fairly. I believe the water authority is a good representative of the Toowoomba community. The water authority is an important member of the Toowoomba community. In relation to the Toowoomba community, I believe that the water authority is “one of us”. I believe that the water authority has the same opinion as me about how to provide good quality water to Toowoomba. In relation to providing a good quality water supply, the water authority has similar values to me. In general, the water authority has similar values to me. I have confidence that the water authority will deliver a good water supply. I think that the water authority has good intentions in managing Toowoomba’s water supply. I can depend on the water authority to provide a good quality water supply. I have complete trust in the water authority to provide me with good quality water supply. How likely do you think it is that there would be any possible problems or risks associated with the recycling scheme? How risky do you think using a water supply that contains the recycled water is to you personally? How risky do you think using a water supply that contains the recycled water would be to other members of the Toowoomba community? I would be willing to use the recycled water for outdoor use. I would be willing to use the recycled water for washing clothes and showering. I would be willing to use the recycled water for drinking and food preparation. How acceptable is the proposed water recycling scheme to you?
.80
Fair procedures
Group membership
3.3. Measures The questionnaire assessed the proposed model variables. In the research context, the authority responsible for administering the recycled water scheme was the local city council. Table 1 lists the items used to measure each of the constructs. The survey assessed the extent to which the water authority’s procedures were considered to be fair, the perceived group membership of the water authority and shared values with the water authority (to measure social identity), source credibility of the water authority, trust in the water authority, risk perceptions in relation to the proposed recycled water scheme and acceptance of the recycled water scheme. The scales for each of the trust model variables were developed as theoretically distinct constructs. The items comprising the procedural fairness, perceived group membership and shared values scales were drawn from the theoretical frameworks presented by Tyler and Degoey (1996), Siegrist et al. (2000), Poortinga and Pidgeon (2003), Frewer et al. (1996) and Metay (1999). The source credibility scale was developed to measure the technical competence of the agency, as well as the agency having the public’s interests, rather than vested interests at heart (Frewer et al., 1996; Lewis and Weigert, 1985). The items for procedural fairness, shared values and source credibility were adapted from Poortinga and Pidgeon (2003). The trust scale was designed to capture Rousseau’s (1998) theoretical construct of trust as the “intention to accept vulnerability” and “positive expectations of the intentions or behaviour of the other party”. Items comprising the risk perception scale was drawn from Hsee and Weber (1997) and Weinstein (1999), and included items about the perceived level of risk to participants personally, as well as to other members of their community. Acceptance of the recycled water scheme was developed in terms of willingness to use the recycled water across a range of uses, from outdoor use, to increasingly personal uses (McKay and Hurlimann, 2004). The perceived group membership of the water authority and value similarity with the water authority were used as indicators of the social identity variable. In addition, background questions and basic demographic information were assessed. All items relating to the trust model variables required participants to rate their degree of agreement with statements on a Likert scale (where 1 ¼ strongly disagree/extremely unacceptable; 2 ¼ disagree/unacceptable; 3 ¼ neutral; 4 ¼ agree/acceptable and 5 ¼ strongly agree/extremely acceptable). The scales for the model variables were created by taking the mean of the items that comprised them. The Cronbach’s alpha was used to assess the reliability of the scales. The individual items that comprise each construct and the scale reliability are shown in Table 1.
Shared values
Trust
Risk perceptions
Acceptance
.64
.75
.88
.89
.90
.82
4. Results 4.1. Descriptive statistics for trust model variables Table 2 shows the means and standard deviations and zeroorder correlations amongst the model variables in relation to the proposed recycled water scheme. Significant correlations were found between a number of the model variables, particularly between value similarity and the other variables. As proposed, the highly correlated variables of value similarity and group membership were reconceptualized into the overarching construct of social identity. Some high intercorrelations remain among the model variables, however, as will be explained in the discussion, there is theoretical and empirical support for conceptualizing these as distinct constructs and the items’ measurement has distinct face validity (see Table 1).
V.L. Ross et al. / Journal of Environmental Management 137 (2014) 61e68
The means for trust and acceptance of the scheme were 3.51 and 3.77 respectively, indicating that respondents fell between neutral and some agreement as to whether they would trust the water authority and would find the scheme acceptable. Means for fair procedures, group membership, value similarity and source credibility were all between “neither agree nor disagree” and “agree”, suggesting that respondents fell between neutral and some agreement/acceptance on these variables. Mean perceptions of risk, at 2.68, fell between the points representing “unlikely” and “neither likely nor unlikely”. A breakdown of the percentage of responses for each response option for each of the questions that comprise the scales can be found in Table 3. To illustrate the differences between participants who were high and low in acceptance of recycled water we show the mean responses to the model variables for participants who were low in acceptance (i.e., scored 3 or below on the scale, this constituted 26.8% of the sample) and those who were high in acceptance (i.e., scored 4.75 or above on the scale, this constituted 26.4% of the sample). As Table 3 shows, high accepters had significantly higher mean responses to all of the model variables than the low accepters, except for risk perceptions where they had lower mean risk perceptions. 4.2. Data analytic strategy To test the hypothesized relationships between the model variables, Structural Equation Modeling (SEM) was conducted using the AMOS 20 program. All variables had less than 5% missing data, and missing data were replaced with the variable mean. The data were analysed using the two-step approach to SEM (Anderson and Gerbing, 1988). The first step involved analysing the measurement model to establish the relationships between the latent variables and the observed indicators. The second step involved analysing the hypothesized structural model. The items that composed each scale formed the indicators in the measurement and structural models. The only exception was the social identity variable, for which items from each of the two aspects of the variable (group membership and value similarity) were averaged to form two indicators, using an approach detailed by Kishton and Widaman (1994). The goodness of fit of both the measurement and structural models was assessed using four practical fit indices: The comparative fit index (CFI), the TuckereLewis Index (TLI), the root-meansquare error of approximation (RMSEA) and the standardized root mean residual (SRMR). Based on the recommendations of Hu and Bentler (1999) the cut off values employed for these indices were CFI >.90, TLI > .95, RMSEA < .06 for a well-fitting model,
Contents lists available at ScienceDirect
Journal of Environmental Management journal homepage: www.elsevier.com/locate/jenvman
Social trust, risk perceptions and public acceptance of recycled water: Testing a social-psychological model Victoria L. Ross a, *, Kelly S. Fielding b,1, Winnifred R. Louis c, 2 a
Smart Water Research Centre, Griffith University, Gold Coast Campus, Edmund Rice Drive, Southport, QLD 4222, Australia Institute for Social Science Research, The University of Queensland, Level 4, GPN3 (Building 39A), Campbell Road, St Lucia, QLD 4072, Australia c School of Psychology, The University of Queensland, McElwain Building, St Lucia, QLD 4072, Australia b
a r t i c l e i n f o
a b s t r a c t
Article history: Received 6 August 2013 Received in revised form 11 January 2014 Accepted 17 January 2014 Available online 4 March 2014
Faced with a severe drought, the residents of the regional city of Toowoomba, in South East Queensland, Australia were asked to consider a potable wastewater reuse scheme to supplement drinking water supplies. As public risk perceptions and trust have been shown to be key factors in acceptance of potable reuse projects, this research developed and tested a social-psychological model of trust, risk perceptions and acceptance. Participants (N ¼ 380) were surveyed a few weeks before a referendum was held in which residents voted against the controversial scheme. Analysis using structural equation modelling showed that the more community members perceived that the water authority used fair procedures (e.g., consulting with the community and providing accurate information), the greater their sense of shared identity with the water authority. Shared social identity in turn influenced trust via increased source credibility, that is, perceptions that the water authority is competent and has the community’s interest at heart. The findings also support past research showing that higher levels of trust in the water authority were associated with lower perceptions of risk, which in turn were associated with higher levels of acceptance, and vice versa. The findings have a practical application for improving public acceptance of potable recycled water schemes. Ó 2014 Elsevier Ltd. All rights reserved.
Keywords: Recycled water Risk Fairness Identification Trust
1. Introduction Positive public perceptions and acceptance of water reuse are now recognized as key factors for the successful introduction of wastewater reuse projects (Dolnicar and Hurlimann, 2011; Dolnicar et al., 2011, 2010; Friedler and Lahav, 2006; Nancarrow et al., 2008). Regardless of the strength of the scientific evidence, public opposition can cause wastewater reuse projects to fail at any stage of their implementation (Friedler and Lahav, 2006; Uhlmann and Head, 2011). There is ample evidence of this from potable water (i.e., for drinking water purposes) reuse projects that have failed to be implemented in the U.S. and Australia. For example, in the 1990s the San Diego County Water Authority’s plans to mix recycled water with the drinking water supply was subject to negative
* Corresponding author. Tel.: þ61 7 5552 7808; þ61 407 583 054 (mobile). E-mail addresses: victoria.ross@griffith.edu.au (V.L. Ross), k.fi[email protected] (K.S. Fielding), [email protected] (W.R. Louis). 1 Tel.: þ61 7 3346 8725. 2 Tel.: þ61 7 3346 9515. http://dx.doi.org/10.1016/j.jenvman.2014.01.039 0301-4797/Ó 2014 Elsevier Ltd. All rights reserved.
public campaigns highlighting the health risks associated with the project and dubbing it “Toilet to Tap”(Po et al., 2003). Despite significant investment the project was abandoned (Hurlimann and McKay, 2004). In South East Queensland, Australia, in particular, where the combination of rapid population growth and prolonged drought placed pressure on the government to seek out rainfallindependent water supplies, there have been a number of noteworthy examples of proposed potable wastewater reuse schemes that have failed to be implemented. During the mid-1990s, potable recycled water projects proposed by the Maroochy and Caloundra councils in Queensland were both strongly opposed by their communities and ultimately rejected (Uhlmann and Head, 2011). Public opposition to the projects was fuelled by campaigns from opposition groups such as CADS (“citizens against drinking sewage”) who warned of alleged health risks associated with drinking recycled water. In 2006, in the droughtstricken regional city of Toowoomba, Queensland, a referendum was held for a proposed potable reuse scheme to augment dam supplies. The proposed scheme created intense opposition and publicity from the CADS group, who cited health risks and a negative impact on Toowoomba’s image as their opposition to the
62
V.L. Ross et al. / Journal of Environmental Management 137 (2014) 61e68
project. The project became politicized, and as noted by the mayor of Toowoomba at the time, moved the focus from water to politics and vested interests (Hurlimann and Dolinicar, 2010). Even with severe water shortages and dam levels dropping to record lows of only 23%, the scheme was abandoned after 62% of residents voted against it (Hurlimann and Dolnicar, 2010). More recently, the Western Corridor Recycled Water Project in South East Queensland (the largest water project to be built in the Southern Hemisphere), was put on hold, despite completion of $2.5 billion of infrastructure. Extensive media speculation regarding possible health risks impacted significantly on community confidence (Roberts, 2008) and in response to the drop in public confidence (as well as unexpected rainfall which restored dam levels) the government changed its policy so that recycled water would only be introduced to drinking water supplies in an emergency, that is, when combined dam levels in the region drop to below 40% (Queensland Water Commission, 2009). Given the critical importance of public perceptions to the implementation of potable reuse schemes, the current study examines the factors that are related to public risk perceptions and acceptance of recycled water. Specifically, the study develops a model that investigates the role of trust in predicting risk perceptions and acceptance and the factors that help to promote trust. In doing this the study makes an important theoretical contribution to the literature on the social dimensions of recycled water management and also contributes more broadly to the study of environmental management and to the social psychological literature on trust.
management, it has been argued that in order for water management schemes to be successful, the community requires trust in the responsible authorities to deliver them a safe water supply (Hurlimann and McKay, 2004; Marks and Jadoroznyj, 2005). The relationship between trust, risk and acceptance has been empirically explored in the risk communication literature in the context of new technologies (Eiser et al., 2002; Pavlou, 2003; Siegrist, 2000; Siegrist et al., 2007), and more specifically in relation to public acceptance of potable reuse (Hurlimann et al., 2008; Nancarrow et al., 2009). Thus, past research and theory provide a strong foundation for the hypothesis that greater trust in the water authority to deliver safe recycled water will be associated with lower risk perceptions and greater acceptance of the recycled water scheme. What past research has not explored though, are the factors that underpin trust in a water authority. This is a critical question given that water resources are likely to become more vulnerable in the future and water authorities will need to explore alternative water sources, including recycled water (Dolnicar and Hurlimann, 2011; Hurlimann et al., 2009). Understanding how trust can be developed will be essential information to aid in this process. In the present model, we explore three inter-related antecedents: perceived procedural fairness, social identity, and source credibility. We focus on these variables as, theoretically, they have been linked to trust, although these linkages have not previously been tested in relation to recycled water.
2. The hypothesized model
According to the procedural justice literature, the belief that one has been treated fairly by authorities enhances acceptance of legal decisions, obedience to laws, and evaluations of public policies; while the belief that one has been treated unfairly prompts protest behaviour (Lind, 2001; Tyler, 2001; Van den Bos and Lind, 2002). Within the context of water management, there is evidence that fair procedures are a major predictor of acceptability of the schemes and compliance with urban and rural water management and policy issues (Hurlimann et al., 2008; Nancarrow et al., 2002; Syme et al., 1999), including community intentions to drink water from potable reuse schemes (Nancarrow et al., 2009). Hence, past research suggests that the more community members consider a recycled water scheme to be fair, the greater their acceptance of the scheme. The relational model of authority (Tyler and Lind, 1992) provides a theoretical explanation for these findings, proposing that people care more about how decisions are made than they do about the actual decisions because procedural treatment provides them with important information about their relationship with authorities (Skitka and Mullen, 2002). Tyler and Degoey (1996) suggest that the way people are treated by authorities provides them with information about whether they are respected members of the group and whether they should feel pride in the group as a whole. In other words, fair procedures indicate to community members whether and how much they share an identity with authorities. Shared social identity in turn increases the likelihood that group members trust an authority and accept their decisions (Tyler and Degoey, 1996; Williams, 2001). Hence, according to the relational model of authority, shared social identity is a key mechanism by which fair procedures influence trust in authorities. It was therefore hypothesized that perceiving that the water authority treats community members fairly predicts a greater sense of shared identity between community members and the water authority. In the current study shared social identity is reflected by the extent to which people see the water authority as a member of their group (i.e., the community)
Below we outline the theoretical and empirical basis for relationships proposed in the hypothesized model. 2.1. Risk perceptions, trust and acceptance A strong body of research has shown the centrality of social factors in water management (Dolnicar et al., 2010; Dolnicar and Schäfer, 2009; Hurlimann et al., 2009; Lienert et al., 2013). In the failed reuse cases discussed above, the public perceived the possible health risks associated with recycled water as unacceptable despite the reassurances provided by authorities and by scientists (Dolnicar et al., 2010; Hurlimann and Dolnicar, 2010; Uhlmann and Head, 2011). The conclusion that perceived risk is a key predictor of acceptance of recycled water management schemes accords with the broader risk literature that shows a clear relationship between risk perceptions and acceptance (Eiser et al., 2002; Mankad and Tapsuwan, 2011; Robinson et al., 2012; Siegrist et al., 2007). In turn, in the risk communication literature, it is generally agreed that trust in authorities to manage risk is a critical factor in the perception and acceptance of risks (Earle et al., 2007; Lofstedt and Cvetkovich, 2008). Trust is a multidimensional, complex construct (Fischhoff, 1999; Grabner-Kräuter and Kaluscha, 2003), but in the current study we use a specific operationalization drawn from the literature (Frewer et al., 1996; Lewicki et al., 2006; Rousseau et al., 1998; Siegrist et al., 2000): a psychological state comprising the intention to accept vulnerability based upon positive expectations of the intentions or behaviour of the authority responsible for the recycled water scheme. Siegrist et al. (2000) have argued that many individuals lack the resources such as knowledge, time and interest to make decisions and take action in relation to science and technology and therefore they rely on trust in the relevant authorities or government agencies to make decisions. In the context of urban water
2.2. Predictors of trust: fairness, identity, credibility
V.L. Ross et al. / Journal of Environmental Management 137 (2014) 61e68
and whether they perceive a sense of shared values with the authority, both of which are important dimensions of social identity (Hogg and Abrams, 1998; Tajfel and Turner, 1986; Tyler and Degoey, 1996). Past research has also conceptualized value similarity as a reflection of social bonds and has shown, consistent with our reasoning that it is significantly related to trust (Siegrist et al., 2000, 2003; Vaske et al., 2007). This relational framing of trust accords with social identity theory which posits that social groups provide members with a social identity: a definition of who one is and evaluation of what that entails (Hogg and Abrams, 1998; Tajfel and Turner, 1986). Importantly, research from the social identity perspective has shown that people who belong to our groups (i.e., ingroup members) are perceived in more desirable ways (e.g., more trustworthy or honest) than outgroup members and that shared social identity reduces uncertainty and produces in-group trust (Brewer, 2007; Hogg, 2003; Hogg and Abrams, 1998). Drawing on social identity theory we further reason that in the context of acceptance of recycled water schemes, the relationship between shared social identity and trust will be mediated by source credibility, that is, the extent to which the water authority is perceived to be competent and has the public’s interests rather than vested interests at heart (Frewer et al., 1996; Siegrist et al., 2003; Tyler and Degoey, 1996). Past research has demonstrated that sources are trusted more when they are perceived to be competent and credible (Frewer et al., 1996; Sutherland et al., 2013; Tyler and Degoey, 1996). Given the technical nature of water recycling processes, and the potential impacts on the health of the population if the process were to go wrong, this variable seems likely to be a critical predictor of trust. Our analysis though, does not see this variable as an independent predictor but rather, we believe that judgments of credibility will be influenced by social identity processes. There is compelling evidence that links shared social identity to more positive evaluations of ingroup members across a variety of group relevant dimensions (Hogg, 2007; Tanis and Postmes, 2005; Williams, 2001). We predict, therefore, that social identity would also influence judgments of the source credibility of water authorities and in turn, trust. We reason that the more that people perceive a shared social identity with the water authority (perceive the water authority as a member of their community who shares their values), the more people will perceive the water authority as a competent entity that has the interests of the group at heart. To summarize the model: We hypothesize that community members who perceive that the water authority uses fair procedures will feel a greater sense of shared social identity with the water authority and this will be related to greater trust in the water authority via increased perceptions of credibility. Greater trust in turn will predict lower risk perceptions and greater acceptance of recycled water. Interestingly, previous research has often drawn distinctions between instrumental and relational approaches to trust, contrasting competency judgments and social relational concerns as the main basis for trustworthiness (Saparito et al., 2004; Tyler and Degoey, 1996; Zaheer and Venkatraman, 1995). Our model, suggests, however, that instrumental and relational perspectives may be strongly inter-related, with social identity influencing credibility judgments. In this way our propositions are consistent with previous researchers who argue for the compatibility of instrumental and relational approaches to understanding trust (Cho, 2006; Edwards and Kidd, 2003; Kerkhof et al., 2003; Tyler and Degoey, 1996).
63
To our knowledge the current study is the first study to advance a theoretical model of the role of trust in authorities and its antecedents and how that relates to risk perceptions and acceptance of proposed recycled water management schemes. Moreover, as the research was conducted in a context in which residents were about to vote for or against a wastewater reuse scheme, the issue was not hypothetical but instead very salient to them. Past research seeking to understand public responses to potable recycled water schemes has tended to focus on the hypothetical introduction of recycled water (e.g., Dolnicar and Schäfer, 2009; Nancarrow et al., 2008). Hence, the research provides a unique opportunity to gain insight into the social psychological variables that are important drivers of the acceptance or rejection of recycled water projects. Through identifying the variables that are related to trust in water authorities, the research can provide critical information to water authorities about how they can develop greater public trust and thereby promote more positive responses to potable recycled water schemes. 3. Method 3.1. Study context The hypothesized model was tested in the context of a proposal to introduce a potable recycled water scheme in the regional city of Toowoomba. Known as “the Garden City” for its parks and gardens, Toowoomba is located approximately 127 km west of Queensland’s capital city of Brisbane. With an estimated population of 157,023 Toowoomba is Australia’s second most populous inland city. As described earlier in the introduction, after years of drought and dam levels dropping to 23%, the water authority proposed a potable wastewater reuse project to augment dwindling supplies. The scheme was strongly opposed by the community action group “CADS” and later rejected by the public in a referendum. The data for this research was collected only weeks before the referendum, so the issue was extremely salient to the community at the time (Hurlimann and Dolnicar, 2010; Price et al., 2012). At the same time, the present study asked participants a broader question than the referendum, which focused on support for adding recycled water to the city’s water supply. The present study examined willingness to use recycled water over a range of contexts (e.g., outdoors, washing, showering, cooking and drinking), allowing a broader analysis of predictors of support for recycled water. 3.2. Participants and procedure Four hundred and one residents from Toowoomba were interviewed by telephone several weeks before the referendum for the proposed recycled water project was held. An attempt was made to obtain equal numbers of males and females, resulting in 220 (55%) females and 181 (45%) males. Of the 923 people asked to participate, there were 522 refusals, indicating a 43% response rate. The age of respondents were spread relatively evenly across predetermined grouped age categories which ranged from “18e19 years”, “20e24 years”, “25e29 years”, “30e39 years”, “40e49 years” etc, up to and including “75 years and older”. Of the 401 respondents, two cases were excluded because of significant missing data, and 19 cases were excluded because they were multivariate outliers (i.e., cases with unusual combination of scores on two or more variables that differ from the range of observations exhibited for other cases). This meant that the final number of respondents included in the analysis was 380. The study was conducted by a Social Research Centre using Computer Assisted Telephone Interviewing (CATI), with each interview taking approximately 15 min to complete. The sample
64
V.L. Ross et al. / Journal of Environmental Management 137 (2014) 61e68
was randomly selected from telephone listings. Telephone interviewers explained that they were conducting an independent study on Toowoomba residents’ thoughts and beliefs about their water supply. Participants were initially asked if they were connected to Toowoomba’s main water supply (and would therefore be affected by the proposed potable reuse scheme). Interviews did not proceed for those not connected. Before being asked about the project, participants were provided with details about the proposed recycled water scheme including information that the wastewater would be treated to a standard higher than safe drinking standards and the process by which it would be released to the dam and then treated again prior to being added to the water supply. This explanation was followed by questions in a format which required participants to rate their level of agreement. Participants were then asked to provide demographic information (i.e., categories of age, income, education, and household type).
Table 1 Scale reliabilities of constructs. Construct
Items
Alpha
Source credibility
The water authority is competent enough to manage our water supply. The water authority provides the Toowoomba community with everything they want to know about their water supply. The water authority does not listen to concerns raised by people like me. The water authority acts in the public interest when it comes to water quality. The water authority makes fair decisions about water provision. The water authority makes an effort to treat everyone fairly. I believe the water authority is a good representative of the Toowoomba community. The water authority is an important member of the Toowoomba community. In relation to the Toowoomba community, I believe that the water authority is “one of us”. I believe that the water authority has the same opinion as me about how to provide good quality water to Toowoomba. In relation to providing a good quality water supply, the water authority has similar values to me. In general, the water authority has similar values to me. I have confidence that the water authority will deliver a good water supply. I think that the water authority has good intentions in managing Toowoomba’s water supply. I can depend on the water authority to provide a good quality water supply. I have complete trust in the water authority to provide me with good quality water supply. How likely do you think it is that there would be any possible problems or risks associated with the recycling scheme? How risky do you think using a water supply that contains the recycled water is to you personally? How risky do you think using a water supply that contains the recycled water would be to other members of the Toowoomba community? I would be willing to use the recycled water for outdoor use. I would be willing to use the recycled water for washing clothes and showering. I would be willing to use the recycled water for drinking and food preparation. How acceptable is the proposed water recycling scheme to you?
.80
Fair procedures
Group membership
3.3. Measures The questionnaire assessed the proposed model variables. In the research context, the authority responsible for administering the recycled water scheme was the local city council. Table 1 lists the items used to measure each of the constructs. The survey assessed the extent to which the water authority’s procedures were considered to be fair, the perceived group membership of the water authority and shared values with the water authority (to measure social identity), source credibility of the water authority, trust in the water authority, risk perceptions in relation to the proposed recycled water scheme and acceptance of the recycled water scheme. The scales for each of the trust model variables were developed as theoretically distinct constructs. The items comprising the procedural fairness, perceived group membership and shared values scales were drawn from the theoretical frameworks presented by Tyler and Degoey (1996), Siegrist et al. (2000), Poortinga and Pidgeon (2003), Frewer et al. (1996) and Metay (1999). The source credibility scale was developed to measure the technical competence of the agency, as well as the agency having the public’s interests, rather than vested interests at heart (Frewer et al., 1996; Lewis and Weigert, 1985). The items for procedural fairness, shared values and source credibility were adapted from Poortinga and Pidgeon (2003). The trust scale was designed to capture Rousseau’s (1998) theoretical construct of trust as the “intention to accept vulnerability” and “positive expectations of the intentions or behaviour of the other party”. Items comprising the risk perception scale was drawn from Hsee and Weber (1997) and Weinstein (1999), and included items about the perceived level of risk to participants personally, as well as to other members of their community. Acceptance of the recycled water scheme was developed in terms of willingness to use the recycled water across a range of uses, from outdoor use, to increasingly personal uses (McKay and Hurlimann, 2004). The perceived group membership of the water authority and value similarity with the water authority were used as indicators of the social identity variable. In addition, background questions and basic demographic information were assessed. All items relating to the trust model variables required participants to rate their degree of agreement with statements on a Likert scale (where 1 ¼ strongly disagree/extremely unacceptable; 2 ¼ disagree/unacceptable; 3 ¼ neutral; 4 ¼ agree/acceptable and 5 ¼ strongly agree/extremely acceptable). The scales for the model variables were created by taking the mean of the items that comprised them. The Cronbach’s alpha was used to assess the reliability of the scales. The individual items that comprise each construct and the scale reliability are shown in Table 1.
Shared values
Trust
Risk perceptions
Acceptance
.64
.75
.88
.89
.90
.82
4. Results 4.1. Descriptive statistics for trust model variables Table 2 shows the means and standard deviations and zeroorder correlations amongst the model variables in relation to the proposed recycled water scheme. Significant correlations were found between a number of the model variables, particularly between value similarity and the other variables. As proposed, the highly correlated variables of value similarity and group membership were reconceptualized into the overarching construct of social identity. Some high intercorrelations remain among the model variables, however, as will be explained in the discussion, there is theoretical and empirical support for conceptualizing these as distinct constructs and the items’ measurement has distinct face validity (see Table 1).
V.L. Ross et al. / Journal of Environmental Management 137 (2014) 61e68
The means for trust and acceptance of the scheme were 3.51 and 3.77 respectively, indicating that respondents fell between neutral and some agreement as to whether they would trust the water authority and would find the scheme acceptable. Means for fair procedures, group membership, value similarity and source credibility were all between “neither agree nor disagree” and “agree”, suggesting that respondents fell between neutral and some agreement/acceptance on these variables. Mean perceptions of risk, at 2.68, fell between the points representing “unlikely” and “neither likely nor unlikely”. A breakdown of the percentage of responses for each response option for each of the questions that comprise the scales can be found in Table 3. To illustrate the differences between participants who were high and low in acceptance of recycled water we show the mean responses to the model variables for participants who were low in acceptance (i.e., scored 3 or below on the scale, this constituted 26.8% of the sample) and those who were high in acceptance (i.e., scored 4.75 or above on the scale, this constituted 26.4% of the sample). As Table 3 shows, high accepters had significantly higher mean responses to all of the model variables than the low accepters, except for risk perceptions where they had lower mean risk perceptions. 4.2. Data analytic strategy To test the hypothesized relationships between the model variables, Structural Equation Modeling (SEM) was conducted using the AMOS 20 program. All variables had less than 5% missing data, and missing data were replaced with the variable mean. The data were analysed using the two-step approach to SEM (Anderson and Gerbing, 1988). The first step involved analysing the measurement model to establish the relationships between the latent variables and the observed indicators. The second step involved analysing the hypothesized structural model. The items that composed each scale formed the indicators in the measurement and structural models. The only exception was the social identity variable, for which items from each of the two aspects of the variable (group membership and value similarity) were averaged to form two indicators, using an approach detailed by Kishton and Widaman (1994). The goodness of fit of both the measurement and structural models was assessed using four practical fit indices: The comparative fit index (CFI), the TuckereLewis Index (TLI), the root-meansquare error of approximation (RMSEA) and the standardized root mean residual (SRMR). Based on the recommendations of Hu and Bentler (1999) the cut off values employed for these indices were CFI >.90, TLI > .95, RMSEA < .06 for a well-fitting model,
