Introduction

New technologies may become publicly contested when they fail to take important societal values into account. The acceptability of a technology can be compromised when it (imposes a risk that) negatively affect(s) values that are deemed important by a substantial part of society (Roeser 2011). The framework of value sensitive design (VSD) is geared towards the inclusion of values in technological artefacts and systems. VSD aims to anticipate value conflicts by incorporating societal and ethical aspects in the development of technology (e.g. Friedman et al. 2008). The approach reflects the notion that technologies are not neutral but produce and reproduce values that may be undesirable (Akrich 1992; Ihde 1990; Latour 1992; Oudshoorn et al. 2002; Verbeek 2005; Wolsink 2000; Waelbers 2009; Winner 1980). By identifying relevant public values, and by adequately considering these values in the technological design, the social acceptability of a new technology can be increased, potentially.

Two important prerequisites for embedding values in technological design are, firstly, to identify values at stake and, secondly, to determine which values are conflicting and cannot be accommodated in the same design. A typical example is the design of a car in which at least two key values are embodied, namely safety and sustainability (fuel efficiency). A trade-off between the two values seems to be inevitable; i.e. the lighter the car, the more sustainable it will be in terms of fuel consumption, but the less safe it will become for the driver and the passenger (Van Gorp and Van de Poel 2001). Alternative production methods or the use of other materials could help bypass this conflict to some extent, but eventually a choice needs to be made regarding what level of safety and sustainability we regard as acceptable. Balancing these two values requires addressing the underlying normative question as to which one of the two values is most important, and why.

We argue that values are not only at stake in the design of technologies, but also in the process of their deployment, especially when we deal with major technological projects with significant societal impacts. This means that not only the design of the technological artefact is to be considered, but also the institutions in which these artefacts are to be embedded (Correljé et al. 2015; Wüstenhagen et al. 2007). Institutions can be defined as “humanly devised constraints that structure political, economic and social interaction” (North 1991, p. 97). They involve informal institutions, such as taboos, customs, traditions, and codes of conduct, which often inspire formal institutions such as laws, property rights, regulations, and procedures (see also Commons 1931).

Hence, institutions not only ‘prescribe’ technological characteristics and design requirements. They also shape the environment in which the various stakeholders behave and interact and in which social contestations takes place. Participation of stakeholders in policy-making and planning is often suggested as a means to create more democratic and socially supported technologies and projects (Fiorino 1990). Public values are issued in the interaction process between stakeholders with different backgrounds, interests, expectations, and attitudes towards the technology (Walker et al. 2011). This implies that a VSD approach should include stakeholder participation.

This paper focuses on the crucial methodological question that follows from this view on VSD, namely how to identify the relevant and the potentially conflicting values. The relevance of addressing this methodological question has been emphasized earlier (e.g. Manders-Huits 2011; Pesch 2015). This paper takes a first step in answering this question. We argue that the rich public debate is an important source for finding relevant public values. We present an explorative method for a structured identification of public values and their potential conflicts using the data of the public debate. We do this based on a debate regarding shale gas exploration and exploitation in the Netherlands. We believe, however, that the presented methodology has broader applicability for identifying public values at stake in technological and infrastructural projects.

“The Dutch Shale Gas Debate” section of this paper will briefly sketch the Dutch debate on shale gas. “Methods” section presents the method by which we identify the values at stake in this debate. “Values in the Dutch Shale Gas Debate” section presents the values we have found and introduces the difference between substantive and procedural values. “Discussion” section reflects on these findings in a discussion. In “Conclusion” section we conclude by summarizing the implications of this approach to VSD for science and engineering.

The Dutch Shale Gas Debate

Shale gas is natural gas which is produced from shale layers deep underground. The Netherlands has a long tradition of natural gas exploration, exploitation and consumption providing much technological and institutional experience in dealing with the societal issues at stake in traditional gas exploitation (Correljé et al. 2003). Anticipating the gradual decline of the Groningen field, the largest European natural gas field, the Dutch government placed renewed emphasis on exploiting smaller fields. Shale gas exploitation is consistent with this policy (Ministry of Economic Affairs 2013a).

When burned, natural gas causes considerably lower CO2 emissions, as compared to petroleum and coal. Hence proponents argue that shale gas development, by expanding the gas reserves, has the potential to contribute substantially to a reduction of carbon emissions (Oudeman 2011). Natural gas-fired power plants are then seen as ideal back-up facilities to compensate for the fluctuations in energy supply associated with renewable energy sources like wind and solar.

At the same time, however, there are serious concerns about the negative effects of shale gas exploitation on health, safety and the environment. These involve the risk of leakage of fracking chemicals into drinking water reservoirs, the possibility of earth tremors and local nuisance (e.g. Ministry of Economic Affairs, Agriculture and Innovation 2012). Moreover, it is argued that an expansion of the producible fossil energy reserves with large amounts of previously un-exploitable shale gas will delay the transition to a sustainable energy system. This leads to the paradoxical situation that shale gas is both positioned as a means to enhance welfare and sustainability and as a fuel that obstructs the transition towards a sustainable energy system.

This contested nature became clearly visible in the Dutch shale gas debate. In September 2008, the British company, Cuadrilla Resources Ltd. requested an exploration permit for shale gas in the Dutch province of Noord-Brabant in the South of the Netherlands. Initially, the application proceeded without many difficulties and, by October 2010, the municipalities of Boxtel and Haaren agreed to adjust their zoning plans temporarily to allow for the exploration activities. Yet, when the first permits for exploratory drilling were awarded, local public opinion turned against this initiative. Rather quickly, the local resistance evolved into a broad, nation-wide, anti-shale gas movement, supported by NGOs and several political parties (Metze 2014). There were many important events that led to this resistance. In May 2011, there was an Earthquake in Blackpool (UK) that may have been the result of fracking activities. Additionally, in September 2011, the anti-shale gas film Gasland was broadcasted on Dutch national television. This increased the visibility of the film that had already been online since 2009 (Metze 2013). The image of flames from a tap shown in the film and positioned to be related to shale gas extraction had considerable impact (Metze 2014).

The sudden resistance towards shale gas emerged somewhat unexpectedly given the fact that The Netherlands has a long history of easy public acceptance of natural gas exploitation and a tradition of consensus-based policy-making in which industry, public authorities, citizens, and the local population generally manage to agree (Correljé  et al. 2003; Correljé 2011). This makes an analysis of the debate worthwhile both from a methodological and an empirical point of view.

The Dutch government instituted a temporary moratorium and commissioned an investigation of the risks of shale gas exploitation in October 2011. A public consultation, commissioned by the EU in the period between 20 December 2012 and 23 March 2013, indicated a low support of shale gas in The Netherlands. From the Dutch individual respondents, 72 % indicated that no shale gas should be developed in Europe, 17 % indicated that shale gas development should only happen if proper safeguards regarding health and safety are in place, and only 11 % were in favor (Bio Intelligence Services 2013).

In August 2013, the research consortium of three engineering firms concluded that the technological risk would be manageable (Witteveen et al. 2013). This conclusion was communicated to Parliament (Ministry of Economic affairs 2013a). The opposition in Parliament and a number of NGOs heavily criticized the findings of the report, particularly regarding the objectivity and the independence of the engineering firms. Moreover, the report was condemned for being too narrow in focus; excluding the discussion on the need and necessity of shale gas exploitation, and ignoring important environmental effects (Commissie m.e.r. 2013). As this criticism received ample media attention, the study fuelled the controversy rather than resolving it. In September 2013, the government decided to extend the moratorium and commissioned additional investigations regarding location-specific aspects. The results of this study are expected to serve as input for a general framework for decision-making on shale gas exploration (Ministry of Economic Affairs 2013b, 2014a). Until this study is completed at the end of 2015, the moratorium remains in place (Ministry of Economic Affairs 2014b).

Methods

We observed a rapidly evolving debate extending from the local to the national level, in which the various parties brought in an increasingly wide range of arguments for and against shale gas exploitation. We also observed that in the arguments in the public debate, values play an important role. For example, the EU study indicates the values of health and safety that impact the positioning of stakeholders (Bio Intelligence Services 2013). In addition to these two values, there are also a number of other values at stake; this was vivid in the public debate. Our analysis is based on the arguments that were put forward by various stakeholders in the Dutch shale gas debate. We analyzed key documents from the National Government, NGOs, Industry, and the Dutch Energy Council,Footnote 1 published between May 2010 and November 2013. Drawing upon the sources referred to in these documents, a snowballing procedure was performed which continued until repetition of the arguments indicated that a sufficiently rich overview had been reached. In total, 97 arguments were collected, of which 41 could be identified as being in favor of shale gas exploration and 56 against (see “Appendix” section).

In the next step we analyzed the arguments using Van de Poel’s concept of a “value hierarchy” (Van de Poel 2014). The value hierarchy comprises three levels. At the top, there are a limited number of values, such as safety and resource durability. At the middle level, we find norms that identify actions that support those values. Norms may include objectives (such as “maximize safety”, “safeguard the environment” or “minimize costs” without a specific target), goals that specify a more tangible target, and/or constraints that set boundaries or minimum conditions. At the bottom of the hierarchy, there are design requirements. Design requirements are very specific and detailed and they form the core of (engineering) design (Fig. 1).

Fig. 1
figure 1

Value hierarchy and arguments (adapted from Van de Poel 2014)

Full size image

We observed that the public debate mainly addressed the mid-level of norms. Norms can be made explicit and expressed in the form of arguments, which are put forward in the public debate. Such arguments comprise normative statements about how the world should be. From the arguments used in the public debate we gained insight into the values at stake and the associated norms. Figure 1 shows an adapted version of Van de Poel’s value hierarchy, showing the position of the arguments examined in this study.Footnote 2

We then identified the values to which the arguments referred. This was an iterative process. We derived values from the arguments and refined these drawing on the literature on applied ethics and ethics of technology (Friedman and Kahn 2003; Beauchamp and Childress 2009; Taebi and Kloosterman 2014; Friedman 1996; Friedman et al. 2002, 2006; Borning and Muller 2012). This process resulted in a list of 20 values.

If the expressed argument did not clearly refer to a specific value, other documents from the stakeholder were used to determine the precise meaning of that argument taking into account its context. For example, the argument: “Shale gas exploration consumes large quantities of drinking water” is an argument that relates to ‘resource durability’. However, because this argument has also been voiced in the context of the regional drinking water supply and the equal distribution of (the risks of) industrial activities, ‘distributive justice’ was also attributed to this argument. The analysis was validated by cross-verification. The values were coupled to the arguments by one member of the research team, then discussed in a broader group of researchers from the team, until consensus was reached. A second step in the verification involved stakeholders from industry, NGOs, researchers, and policy-makers who participated in expert workshops. In these meetings we presented the definitions of the values to the participants and asked them to link these values to a selection of 47 aggregated arguments. On average 38 arguments were coupled with at least one value similar to our labeling. This provided an indication that our coupling of arguments and values was recognized as valid.

Values in the Dutch Shale Gas Debate

In the analysis of the Dutch shale gas debate we identified two different types of value hierarchies. Firstly, there was a hierarchy of substantive values, or values that relate to the technology of shale gas and the effects of the project(s), which consisted of three main values: sustainability, security of supply, and affordability. Secondly, a hierarchy of procedural values was found. Procedural values relate to the nature of the rules and regulations and the procedures that constitute the decision-making on the exploration and exploitation of shale gas. These latter issues concern aspects such as the limitations of current legislative frameworks, access of stakeholders to the decision-making process, justice and transparency (see Fig. 2; Table 2).

Fig. 2
figure 2

Substantive values in the Dutch shale gas debate

Full size image

Substantive Values

The value hierarchy was constructed based on the key documents mentioned earlier. From these documents we specified the three main values of sustainability, security of supply, and affordability. These values refer directly to the three objectives of Dutch energy policy (Ministry of Economic Affairs, Agriculture and Innovation 2011) and EU energy policy [the reduction of GHG emissions, security of supply and the competitiveness of the European economy (European Commissions 2006)].

We then analyzed which norms stem from which values (and likewise, which design criteria stem from which norms). Conversely, we have found norms in these documents, that we have connected to the higher level values (and likewise, design criteria that have been connected to norms).

This process resulted in a list of six substantive values, each of which relate to one of the key values of energy policy (see Fig. 2).

Table 1 presents the definitions of each of the substantive values. For example, the argument “Shale gas can increase the desired geographical diversity of the gas supply” refers to the value ‘international stability’ and the argument “Shale gas might have adverse environmental impact on flora and fauna” refers to the value of ‘environmental friendliness’.

Table 1 Definitions of substantive values and examples of arguments
Full size table

Arguments may also refer to two values at the same time. For example, the argument “To realize a carbon extensive economy in 2050, a balance between fossil and renewable energy is required”, links to both ‘resource durability’ and ‘environmental friendliness’. We will further elaborate on this issue below in the discussion section.

Each of the arguments was assessed on the basis of the themes it addresses. This gives an overview of the thematic scope of the values in the debate (Table 2).

Table 2 Substantive values and themes in the Dutch shale gas debate
Full size table

Procedural Values

Three procedural values were identified: distributive justice, procedural justice, and accountability. Each of these values relate to different aspects of the procedure surrounding shale gas exploration and exploitation (see Table 3 for definitions).

Table 3 Definitions of procedural values
Full size table

While these procedural values were present in the debate, the distinction between substantive values and procedural values might not be as clear-cut as suggested in Figs. 1 and 2. The interpretation of a value can also determine whether it is a procedural or a substantive value. For instance, when the value ‘distributive justice’ refers to (questions regarding) patterns of distribution of an important good or commodity, ‘distributive justice’ is a procedural value. When it focuses on the unit of distribution; i.e. what is the entity or unit of benefit, good or commodity that is to be distributed, it is a substantive value. This later interpretation would, for example, be present in a (hypothetical) discussion regarding the fairness of distribution of two goods with comparable (though not the same) characteristics in which one of the goods is expected to run out. The question that can then arise is whether or not the second good is a fair substitute. In the analysis of the Dutch shale gas debate, ‘distributive justice’ is a procedural value, since we have only observed the procedural interpretation of this value in the debate.

Distributive justice was often used to indicate the (un)fair distribution of costs and benefits among groups and between generations such as the distribution of adverse effects (e.g. pollution or the risks of accidents over time and over locations) or the benefits (e.g. economic gains) (Fig. 3).

Fig. 3
figure 3

Procedural values in the Dutch shale gas debate

Full size image

Table 3 presents the definitions of each of these values that were used in the debate for linking the values to arguments. For example, the argument “The prospective drilling site of the municipality of Boxtel is located near a railway. It is undesirable to locate two activities with risk together” links to the procedural value of ‘distributive justice’ and the substantive value of ‘health & safety’.

The procedural values were also assessed in terms of their thematic scope (Table 4). The timing theme in Table 4 relates to both the timely disclosure of information and to the timely involvement of stakeholders. These concerns pertain to the desire for recognition and the ability to influence the process of decision-making.

Table 4 Procedural values and themes in the Dutch shale gas debate
Full size table

Discussion

Our analysis of the Dutch shale gas debate resulted in two interesting findings. First, we observed that all values, both procedural and substantive, are shared amongst different stakeholder groups, both proponents and opponents. This shared mobilization of similar values holds for both procedural and substantive values. However, this does not mean that all stakeholders agree on the interpretation of the values. For example, an advocate’s supportive argument involving ‘environmental friendliness’ is: “Shale gas offers the opportunity to realize cheap and relatively quick CO2 reduction,” whereas an anti-shale gas stakeholder refers to the same value in the following argument: “Shale gas is a fossil fuel and cannot be sustainable.” While the value of environmental friendliness was shared, this shared value also brought up the discussion topic of whether or not some fossil fuels can be more or less environmentally friendly.

Also, procedural values were endorsed by stakeholders. In this case, proponents generally seemed satisfied with the existing institutional frameworks while opponents demanded additional norms and restricting criteria. As regards the value ‘distributive justice’, however, both proponents and opponents identified the need to come up with new norms and criteria for the improvement of current institutional structures. Various arguments criticized the limited possibilities for the compensation of local communities. The issue of compensation was, however, a source of controversy in itself. Compensation is for example being criticized because it can be seen as bribery (consistent with e.g. Ter Mors et al. 2012; Hannis and Rawles 2013).

The second observation is that there are many shared values but conflicts arise about how different stakeholders translated or operationalized a shared value. This finding contrasts with empirical studies on cases in which two or more values conflict; e.g. usability and privacy can conflict in software design processes; safety and sustainability seem to oppose each other in car design, and the safest nuclear reactor is not necessarily the most secure one (e.g. Friedman, et al. 2008; Van Gorp 2007; Taebi and Kloosterman 2014; Oosterlaken 2014). What all these cases have in common is that trade-offs between these values are inevitable. Our shale gas study showed that, rather than an inter-value conflict that necessitates a value trade-off, there seems to be an intra-value conflict which pertains to different understanding of how a particular value could best be served. This suggests that VSD should be expanded by looking beyond value trade-offs in order to develop a more nuanced understanding of the conflicting operationalization of values and how to manage value conflicts when developing design requirements. Here we would like to point to John Rawls’ distinction between concept and conception; i.e. while virtually everybody would endorse the notion that we need a just society, there are various opinions about of what justice exactly entails. Some might argue that justice is understood as the fair distribution of basic goods while others see justice as the respect of fundamental human rights, and again another group considers justice the course of action that maximizes the utility for the greatest number of people. To put it in other words, there seems to be a consensus about the concept of justice whereas there are drastically different conceptions of justice (Rawls 1971; see also Hart 1961). This theoretical difference has also been observed empirically by Doorn (2012); i.e. while everybody would endorse the notion of responsibility, stakeholders’ rationales for distributing responsibility in an R&D network (or their conceptions of responsibility) could differ. In line with Doorn (2012), we have empirically shown that at the concept-level there is a great degree of consensus regarding the central values that society upholds for the development of a technology. Even in the fierce shale gas debate with all its controversies, this consensus could be found. It is therefore important to understand public values at the level of conception, since controversies seem to arise from a different conception, or different operationalization of the same value.

To reveal conceptions we need to understand how a certain value is being operationalized into a norm. For instance, the value of safety in shale gas drilling could relate to groundwater contamination as a result of drilling but it could also relate to the potential pollution from accidents with trucks that supply chemicals for fracking. In order to take such different conceptions of safety into consideration, we have to analyze the public debate. Each of these conceptions will eventually result in different norms and, consequently, different design criteria. Conceptions reveal themselves in the translation steps we make from the top level of value to the mid-level of norms in a value hierarchy (see also Van de Poel forthcoming).

Value sensitive design holds the potential to reconcile value conflicts and thus to lead to more responsible and socially-supported technologies, institutions and procedures. Our results show however, that a consensus on values may appear a fake consensus, since it can hide conflict at the level of norms. This means that value sensitive design needs to account for the divergent and dynamic ways in which a value is translated into different norms, i.e. the divergent conceptions of a value. Value sensitive design will benefit from dialogue between stakeholders, in which the two levels of values and norms are made explicit and the relation between those is jointly explored. This helps stakeholders to reflect on their own (conceptions of) values and relate these to the values of other stakeholders, which will contribute to the mutual understanding of different positions in the debate. Ideally, it leads to shared norms, or shared conceptions, that can overcome previous conflicts.

Conclusion

VSD has originally only focused on the design of artefacts. We argue that VSD is also relevant for the design of institutions as well as participation procedures (for an elaborate discussion see Correljé et al. 2015; Taebi et al. 2014; Van den Hoven et al. 2014). We focus on one of the methodological challenges of VSD: the identification of values and value conflicts prior to and during the developments of a new technology by drawing on the arguments expressed in the public debate. We identified two hierarchies of values from the debate: substantive and procedural. Both types of values are important to opponents and proponents of shale gas because all stakeholders operationalized the same values (both procedural and substantive) in the debate. Thus we conclude that contestation can be characterized not as inter-value conflicts, but rather as intra-value conflicts in which stakeholders interpret the same relevant values differently.

In sum, we argue that VSD needs to start by identifying values associated with contested technologies. One way to do so is by examining the arguments expressed in a public debate and analyzing their underlying values. Insight into stakeholders positions can help in facilitating a timely and iterative discussion about values, norms and design criteria. To create legitimacy, such a process should be open for new information and new insights to allow (re)positioning and enrichment of viewpoints (Pesch 2014). Deliberate design for values is not a process in which the majority rules because minority opinion may include values that are relevant (Taebi et al. 2014; Dunn 2001). In other words, to create and implement a technological design, we must look beyond technology itself and iteratively include institutions and stakeholder interactions to acknowledge the complex and dynamic embedding of a (new) technology in a societal context. Such an approach may not prevent contestation to arise, but it may present a deadlock as a consequence of such contestation.