Public values in the socio-technical construction of autonomous vehicle futures

ABSTRACT

This study argues that policymaking guided by public values can help to secure public benefit in technology implementation, for two reasons. Firstly, the emergence and deliberation over plural values can result in more publicly beneficial outcomes; secondly, public values establish the meaningful link between public interest and policy goals. We explore the value bases of arguments around the societal implementation of autonomous vehicles through interviews with professionals. Mapping values on a conceptual model with two dimensions of recognition-neglect and consensus-dissensus, we find that a limited set of values receives recognition, while two distinct sets are contested or neglected almost entirely.

KEYWORDS:

• Public values
• deliberative democratic theories
• public policy
• dissensus
• autonomous vehicles

Introduction

The advent of the autonomous vehicle has often been anticipated as transformative for access (Jonas et al. 2017) and freedom in mobility (Stayton, Cefkin, and Jingyi Zhang 2017). Yet, its potential negative implications on urban sprawl (Jinping et al. 2021), car-centric development and transport inequities (Emory, Douma, and Cao 2022), amongst others, have gained increasing attention in the literature. The current process of autonomous vehicle (AV) introduction is shaped by corporate players, interest groups, academia and governments acting in an uncertain environment. It is unclear how much the technology can realistically achieve; the purposes, ways, and extent of regulating the technology; and how to harness the technology for public benefit amidst diverse and conflicting societal values and goals (Cohen et al. 2020). Against this backdrop, public values (Bozeman 2002, 2007) can be used to strategically and actively guide this potential mobility transition to serve cross-cutting public interest and socio-urban goals (Guo and Marietta 2015).

The public values literature focuses policy attention towards public benefit by harnessing democratic processes for effective public action and safeguarding the public interest. While some touch on conflicts in public values (Jørgensen and Bozeman 2007), the literature tends to define value consensus as a goal (Bozeman 2007; Huijbregts, George, and Bekkers 2021) and does not adequately problematize notions of dissensus, value conflicts (Thacher and Rein 2004), multiple publics and value pluralism (Sarah, Brown, and Cherney 2021). To address this, we draw from deliberative democratic theories to apply the notions of consensus and dissensus in public values.

We posit that the recognition and inclusion of public values in societal deliberations on the ramifications of new technologies are essential for shaping their implementation for public benefit. For this reason, this study seeks to understand if and how public values are used by professionals in the societal implementation of AVs. We set up the Recognition-Consensus Model as a conceptual framework to analyse the AV discourse in Israel, which is one of the global hotspots in AV technology development, according to KPMG’s (2020) industry research.

Building on the argument that policymaking based on public values helps to secure public benefit, we advance two lines of reasoning. The first is that public values can be made more useful in societal deliberations not so much by making consensus a goal, but by giving space for the plurality of values to emerge in the search for publicly beneficial outcomes. Section 2 expands on the theoretical basis of this argument. Secondly, public values can create the meaningful link between broad public interest and specific policy goals, and therefore should be more explicitly and consistently applied in policymaking. Furthermore, in the absence of clear policy goals and directions (which is common in new policy fields such as AV implementation), public values can serve as the building blocks and basis for creating policies aligned with public interest. Drawing from the interview results, Section 6 further develops this second argument, elaborating on the theoretical and practical roles of public values.

The next section presents a literature review on public values and deliberative democratic theories on consensus and dissensus. Section 3 introduces the Recognition-Consensus Model. Section 4 describes the research methodology. Section 5 covers the results and discussion on the value mapping and analysis of AV implementation in Israel. Section 6 offers the conceptual and policy contributions of public values to societal deliberations and decision-making. The final section covers the conclusion.

Literature review

The concept of public values

The concepts of public value, public values, the public interest, and the public sphere are related with some distinctions (Bryson, Crosby, and Bloomberg 2015; Benington 2015). Benington (2015, 30) differentiates between the notions of ‘public value’ and ‘public values’, whereby the former relates to the end goal of deepening democracy through an informed and empowered public that takes responsibility for decision-making for public benefit. The latter, public values – the focus of this paper – relates to the virtues, capabilities and principles such as justice, fairness, transparency, and accountability necessary to advance the democratic praxis towards enhancing public interest (Benington 2015).

Bozeman (2007, 132) defines public values as those providing normative consensus about (a) the rights, benefits, and prerogatives to which citizens should (and should not) be entitled; (b) the obligations of citizens to society, the state, and one another; and (c) the principles on which governments and policies should be based. For Veeneman and Koppenjan (2010, 224) public values, in abstract or operationalized form, are those that ‘we collectively expect governments to secure in our society’.

Moore (2014, 468) delineates public values as those that are ‘articulated and embraced by a polity working through the (more or less satisfactory) processes of democratic deliberation’. He further sets out that ‘the appropriate arbiter of public value is the collectively defined values of a “public” [working through the] imperfect processes of democratic governance’ (Moore 2014, 465). Specifying the constituent of this ‘public’, Bryson et al. (2016) call for multi-actor involvement: the public, private and third sectors working with varied logics at different levels and action arenas, in the co-creation of public values.

Guo and Marietta (2015) argue that public values, rather than policy preferences, can be a better basis for democratic representation, especially where citizens’ attitudes on policy are not formulated or not clearly grasped:

The recognition of public values requires an explicit emphasis on the beliefs that the public holds (values) rather than the usual emphasis on what they do not hold (policy preferences). […] While many citizens do not have informed attitudes, […] they do possess political values that are consistent across time and shape the policy attitudes that they do develop. […] Political values allow for an important but infrequently recognized form of representation grounded in value congruence rather than a specific agreement on public policy

(Guo and Marietta, 2015, 68).

The emphasis on public values over policy preference is especially relevant for novel and emerging issues with multiple potentialities and possible widespread societal implications. Few but an increasing number of works address the role of public values in technological innovations. Through the approach of value-sensitive design, Taebi et al. (2014) call for the inclusion of public values towards responsible innovation. Van Dijck, Poell, and De Waal (2018, 5) argue that the platform society created by Big Tech companies brings about ‘confrontations between different value systems, contesting the balance between private and public interests’ and undermines institutions defending broad values such as democratic control. The authors call for greater articulation of public values as a basis to redesign platform systems aligned to social goals. Examining urban experimentations of new smart city technologies, Kronsell and Mukhtar-Landgren (2020) note that in the collaborative ethos of experimental governance, municipalities, while being regulators, take on potentially contradictory roles as facilitators for technology trials with the private sector. Here, values such as transparency and accountability in decision-making need to be clarified and safeguarded. Similarly, Eneqvist and Karvonen (2021) highlight the role of governments as the moral backbone of securing ‘collective values such as democracy, legality, impartiality, transparency, and rule of law’ (ibid., 186) in experimental governance.

Public values: a project towards consensus?

Consensus as a goal appears to take priority among public value scholars such as Bozeman (2007, 132) who frames the purpose of public values as a basis for ‘normative consensus’. Furthermore, the alignment of public values to a neoliberal framework (especially under Moore’s 1995 public management paradigm) and their amenability to public-private collaborations can lead to a kind of ‘managed democracy’ that impinges upon ‘open-ended democratic contestation’ (Dahl and Soss 2014, 497). This potentially leaves socio-political inequalities and power disparities unaltered, and diminishes the role of the state as a countervailing force to market forces (Stoker 2006; Rhodes and Wanna 2007).

Consensus purportedly brings about a greater sense of participation and stake in decision-making processes, supports the sharing and refinement of ideas and choices for decision-making, and enhances collaboration and a sense of shared purpose. Landemore and Page (2015) note that in the problem-solving paradigm, normative consensus retains an appeal. Its grounding is in the Habermasian ideal of the rational consensus (Habermas 1991, 1996), supported by early deliberative theorists such as Cohen (1989) and Elster (1986) who pursue consensus as the goal of deliberation, where through reasoning the better argument prevails.

Yet, the framing of consensus as a goal of deliberation has been subject to criticisms (Little 2007). Firstly, using consensus-seeking as a strategy and goal may put participants under undue pressure for compromise to reach an agreement. It may tend towards conformism, which can circumscribe the diversity of opinions and the search for better solutions (Friberg-Fernros and Karlsson Schaffer 2014). The deliberative arena may become vulnerable to power dynamics where influential persons or groups drive the agenda. Secondly, there are broader socio-political difficulties and ramifications of seeking consensus as the main aim. Contemporary societies are often characterized by value pluralism, multifaceted societal conflicts, and incommensurable differences, which throws the attainability of consensus into question (Friberg-Fernros et al. 2019). Consensus may also be an undesirable goal when striving towards it may serve to marginalize, suppress and exclude minority voices, interests and perspectives (Young, 1996). Sagoe (2018, 327), referencing Swyngedouw (2009), relates consensus-based governance relations with the ‘post-political condition’ which ‘comprised the supplanting of antagonistic governance relations with more consensus-based governance relations in support of neoliberal governance agendas’. This resonates with Rancière’s (2004) concept of ‘post democracy’ where political discourse centres on consensus building that serves the establishment and growth of liberal economies while eroding dissensual debates for the staging of its alternatives (Sagoe 2018).

Incorporation of dissensus in imagining alternative futures

To incorporate dissensus in imagining different futures, it is useful to identify conflictual (and neglected) values to place marginalized concerns onto the public values and public policy purview of awareness and consideration. This resonates with Guo and Marietta’s (2015) assertion of the need to address conflictual values as necessary public value questions.

A starting point can be what Mouffe (2004, 46) identifies as the nature of consensus: a ‘temporary result of a provisional hegemony, as a stabilization of power, and that it always entails some form of exclusion’. Upon this, there is opportunity to reshape the democratic public sphere that ‘makes room for the expression of conflicting interests and values’ (Ibid). In a similar vein, deliberative theories have evolved to accommodate conflicts and dissensus, and question the need for consensus as a goal (Landemore and Page 2015; Van Bouwel and Van Oudheusden 2017; Friberg-Fernros, Karlsson Schaffer, and Holst 2019).

Dryzek and Niemeyer’s (2006) meta-consensus can be a useful concept through which parties can find mutual recognition of values in the discussion arena, without having to agree on the specifics of what is to be done. Normative meta-consensus is defined as ‘the extent that there is agreement on recognition of the legitimacy of a value, though not extending to an agreement on which of two or more values ought to receive priority in a given decision’ (Dryzek and Niemeyer 2006, 639). Normative meta-consensus implies engaging the attitudes of reciprocal understanding and respect for basic values – however different they may be – of other groups, to enable the continuing and creative search for outcomes (Dryzek and Niemeyer 2006). Gutmann and Thompson’s (1996) deliberative disagreement seeks to address differences through reasoning in a mutually justifiable way while allowing parties to continue to disagree over core moral principles. Landemore and Page’s (2015) concept of ‘positive dissensus’ does not necessitate resolution by consensus, but an aggregation of a range of alternative ideas that generate epistemically better solutions. This can be particularly relevant for policymaking for the societal implementation of emerging technologies, where in uncertainty and instability, diversity in idea and solution generation can provide greater rigour and responsiveness (Page 2010). Bouwel and Oudheusden (2017, 497) explore deliberative theories employing the constructive use of conflict for ‘mutual learning, the articulation of disagreement, and democratic modulation’ in scientific assessment and practice. Brown and Tregidga’s (Brown and Tregidga 2017, 2) study on social and environmental accounting through the Rancièrian lens finds benefits of dissensus in ‘opening new social realities’ in the ‘creative staging of alternatives’ and in ‘addressing social injustices and ecological unsustainability’.

In the following, we present a conceptual framework for analysing public values in the discourse on the societal introduction of AVs in Israel.

The recognition-consensus conceptual model

The conceptual model (Figure 1) comprises the two dimensions of recognition and consensus. The recognition dimension maps the degree to which values are highly emphasized (frequently raised) on one end, to those that are mostly neglected or possibly misrecognized on the other. Highly recognized values may be a proxy for their perceived importance. Conversely, values of low recognition are those either not frequently raised by participants, or those explicitly mentioned as receiving little recognition in the discourse. As such, the recognition dimension can reflect the relative prioritization of values. The consensus-dissensus dimension maps the extent to which values receive social agreement or are contentious. This study distinguishes two types of dissensus. Dissensus at a meta-level addresses disagreements over the importance and priority of a value in decision-making. Dissensus at the operative level refers to disagreements on specific approaches or interventions towards achieving the value.

Figure 1. The recognition-consensus model.

When the recognition and consensus dimensions are jointly considered, that is, with the two axes juxtaposed, they provide the backdrop for mapping a broad spectrum of relevant values in relation to their social accord. The four value typologies are organized and described below.

Values of high social accord

These can be considered shared values, receiving high recognition and high consensus. These values are drawn from issues considered dominant and receiving social agreement. The high social accord of these values means that perspectives or actions that address these values are generally legitimized, given attention and resources, and can proceed in the societal agenda with fewer obstacles. Yet, there can be a risk that perspectives building on such values can potentially dominate the discourse and crowd out other important ones.

Values of high contention

These are values of high recognition and high dissensus, and are usually drawn from known and highly contested issues. Being highly contested, issues drawing from such values may result in conflicts, antagonism or even stalemate. Given their high recognition, they can draw attention, resources and societal debate needed to clarify and address the nature of dissensus.

Values of low social accord

These are values of low recognition and low consensus, which may be drawn from poorly identified and conceptualized issues. This can mean that issues surrounding these values are only understood among certain groups or that their implications have yet to be well-formulated. There can be a lack of available information attributing to fragmentary collective knowledge on those subjects. These may signal unidentified problems such as uncharted policy areas, blind spots in conceptualization, or even issue evasion. These may also be socially and politically difficult issues subject to splintered opinions.

Subordinate values

These are values of low recognition and low dissensus, drawn from issues which are not well-examined and may be susceptible to misrecognition. Some of these may be ‘expert problems’ requiring certain competencies to understand. As such, any apparent consensus about them may be a veneer consensus that obscures untested assumptions, unexamined problems or contextual implications. These issues may thus be prematurely determined (by a lack of dissensus) to be less or not important. This dovetails with what Guo and Marietta (2015) describe as values portrayed to be consensus values but with different underlying definitions or interpretations. This makes them in essence conflictual values if different groups ascribe different meanings to these values and seek conflicting policy outcomes.

This framework helps to identify dominant values and to understand whether and why particular issues are more likely to advance in the societal agenda. It surfaces highly contested values and issues – how dissensus brings out value contradictions within complex issues, and the value contestations and trade-offs in decision-making. It brings to awareness values that receive insufficient attention or are misrecognized, stemming from issues susceptible to conceptual flaws or narrow/distorted problem formulation.

In this study, the model is applied as a research tool to discern dominant narratives and areas of dissensus, and identify gaps in problem formulation and policies. We will briefly discuss its potential application within a policymaking context in the concluding section.

Methodology

In this study, we use the model to structure and analyse the values at play among AV professionals in Israel. We identified values through the AV issues raised in the interviews, including their accompanying arguments, rationales and value bases. This study focuses on professionals because the nature of AV implementation requires expertise and knowledge from various sectors and disciplines.

The methodological sequence is as follows. We interviewed AV professionals to identify the values they used in their work and to understand which of these values are recognized or less recognized, in consensus or dissensus. We used thematic analysis to parse the transcripts of each interview to identify the values. Subsequently, all interview transcripts were analysed jointly to plot the identified values on the conceptual model. The results are presented and discussed according to key values, and organized around the four quadrants of the model.

In-depth interviews

The interview questions were structured into three parts (see Appendix). The first part consists of questions on interviewees’ professional opinions on the societal implementation of AVs. The second part focuses on issue identification and problem formulations of AV implementation that the interviewees encounter in their work. This part also includes a question on the implications of a hypothetical experiment for an exclusive AV zone in central Tel Aviv. Thus, the first two parts of the interview did not directly inquire about values but were crafted to draw out how respondents used values (whether consciously or not) in their professional work – in their tasks, problem formulation, decision-making, and discussions with others on AV issues. Only in the third part we posed explicit questions on values – value consensus, dissensus and neglect. As such, relevant values emerged from interviews; they were not prescribed.

Respondents

The 20 interviewees were AV or transport-related professionals in Israel across different sectors (government, private sector, NGOs and academia) and disciplines (transport planning, transport engineering, mobility technology development, etc). Some respondents held multiple roles across sectors, but they are sorted into one sector based on their main role (Table 1).

Table 1. Respondents’ backgrounds.

Sector	Number of respondents	Where respondents are from (As some respondents held multiple roles, numbers in the brackets may not add up to the total)
Public sector	4	Transport-related government organizations at national (2 persons) and city (2) levels.
Private sector	5	Industries such as urban planning consultancy (2), AV technology development and vehicle manufacturer (1), shared-mobility providers (1), transport infrastructure (1).
Academia	7	Universities or research institutes including urban design (1), transport planning and policy (1), transport modelling (3), social implications of transport (1), travel behaviour (1), AV law and regulations (1).
NGOs	4	Transport-related NGOs examining transport issues, including future mobility and AVs, such as social and policy aspects of transport (2) and transport technology accelerators (2).
Total	20

Analysis of findings

The interviews were carried out in English. They were audio-recorded, transcribed and tagged for keywords. We conducted a thematic analysis of the interviews to identify a set of relevant values (the themes correspond to the values themselves). Sometimes, values (‘equity’, ‘governance’ or ‘transparency’) were directly identified by the respondents. Yet at other times, respondents may not use these or other like terms, but their references to these values were extracted from the content of responses. The identified values were then mapped onto the Recognition-Consensus model.

Figure 2 shows sixteen values distilled from interview responses and mapped onto the four value quadrants in relation to their recognition/neglect and consensus/dissensus. The values were mapped based on the extent to which each value is raised – not just the number of mentions but the degree to which respondents drew on a particular value to support their arguments. Each value was positioned based on its relative ‘ranking’ vis-à-vis other values on the two axes. For instance, ‘safety’ is ranked slightly lower in recognition but much higher in consensus than ‘cross-sector collaboration’.

Figure 2. Value mapping for AV Implementation in Israel.

*Note that only the relative positions of the word bubbles in relation to the two axes are relevant (and not their sizes).

Results and discussion: value mapping and analysis of AV implementation in israel

High recognition and high consensus: values of high social accord

Values of high recognition and high consensus were those frequently raised and received a high level of agreement, and are thus values of high social accord. These values pertaining to potential AV implementation among respondents are ‘safety and harm avoidance’, ‘economic growth and entrepreneurship’ and ‘innovation’. The following presents how respondents framed these issues, and discusses the beneficial and problematic implications of values of high social accord.

Safety and harm avoidance is a value in this study and in the general literature (such as Kalra and Groves 2017) often deemed as straightforward and of high priority, requiring little clarification. That is, without achieving a baseline level of safety, AVs would simply not be implemented. Conversely, the other line of argument is such that if AVs prove to be safer than human-driven cars, there is then little reason not to implement AVs. In this study, respondents across sectors cited safety as a high priority and potentially one of the most important contributions of AV implementation. An AV technology developer noted that it can help to ‘reduce accidents from human error and aggressive behaviour in driving’, and a private sector consultant to the government suggested that it could support the move towards ‘Vision Zero’ [from a current level of] a few hundred deaths a year’.

Another frequently raised value is innovation. Respondents identified a spectrum of innovations that AVs can bring about in transport, technology, urban planning, and business. Responses fall into two broad groups: those that associated innovation mainly with the optimistic rhetoric of technological transformation, and those that further probed the public value dimension of innovation and questioned how to extract this value. The former group tended to relate innovation to the smart mobility revolution: the re-conceptualization of how transport can be organized, provided, and consumed, and the reshaping of its functional and social meaning. As one public sector respondent explained, ‘we are looking at the smart mobility revolution, comprising the four characteristics of shared, autonomous, connected and electric mobility, which would give people more opportunities in a cheaper way’.

Some others were cognizant that innovations do not in themselves bring about public benefits but have problematic aspects. A public sector respondent spoke of how public benefit can be at risk in decisions on how and where to deploy AV ride-hailing and sharing: ‘Robo-taxis may not be the favourable type of transport [in Tel Aviv] especially when it does not compliment public transport, because the companies want to provide services especially in the city centre, right where it is less needed’. A handful identified the public value angle of innovation such as improving public transport and opening up data for societal learning in AV technology adoption. Yet innovations for public benefit are not automatically delivered, but need to be negotiated and secured. As such, the public value lens focuses efforts on extracting public benefit from innovations. For AVs, these can include: societal learning and knowledge building; making data collected by service providers come under the public domain for use in planning; directing public resources invested to support AV implementation for public benefit, such as how the private sector technology may be harnessed to improve public transport; and enabling innovations in AV implementation to serve the larger goal of public space transformation.

High recognition and high consensus values can tend towards becoming dominant values with strong rationales serving as ‘overwhelming argument(s)’ that push the agenda, sometimes at the expense of other important values (Soh and Martens 2022), as seen in arguments for safety. For the value of innovation, the learning experiences shared in this study show that public values need to be deliberately incorporated and negotiated to direct technological implementation for public benefit.

High recognition, high dissensus: values of high contention

Values of high contention were those frequently raised by respondents but receive contested viewpoints and approaches towards possible solutions. In this study, values in high dissension – ‘efficiency’, ‘governance’ and ‘cross-sector collaborations’ – appeared to gain attention towards better framing and understanding of the issues surrounding these values rather than reaching impasses in discussions.

The value of efficiency, predominantly transport efficiency, was raised most extensively by respondents. Efficiency can be considered as an end in itself or an instrument to achieve other values such as accessibility, availability of choices, and environmental sustainability. Respondents often engaged the value of efficiency through questioning the practical viability of AVs as a potential transport solution. This included harnessing the technological efficiency of AVs to improve traffic flows, transport capacity and levels of service, enhance access to more destinations, and provide more choices in routes and modes. The key area of dissensus was whether those efficiency claims of AVs can be actualized in reality. A respondent from an NGO noted, ‘There was a major difference of opinion that AVs can be solutions to private car congestion. I did some research if AV technology holds the promise of solving congestion. I found it is impossible to provide the capacities currently being provided by the bus system in Tel Aviv with AVs.’ A private sector respondent gave a different viewpoint, ‘Personally, I believe in the great power, the great advantage of autonomous shared transportation vehicle services, although today I understand more the disadvantages of its implementation in certain locations. That is due to new research that I was exposed to recently, and to my discussions with other colleagues’. Several respondents noted that the operative dissensus on efficiency has shifted noticeably towards consensus (indicated by the arrow). A greater agreement has emerged on the disbenefits – from an efficiency perspective – of over-dependency on shared vehicles at the expense of mass public transportation, and the need for suitable policies for shared and privately owned AVs.

We observed that both proponents and opponents of AVs used arguments of efficiency to support their perspectives. This can be due to factual indeterminacy, such as the over- or underestimated efficiency claims of AVs. Furthermore, the multiple possible pathways of AV adoption based on different assumptions (such as AV technological capability, the level of autonomous driving, and transition stages of mixed human-driven to full AV traffic) result in very different scenarios. As such, dissensus has not arisen because people disagree on the value of efficiency and its relevance for AV implementation, but due to different conceptualizations of efficiency. In this case, the value of ‘efficiency’ has benefitted from deepening discussions from surface disagreements towards contextualized problem framing.

The value of good governance was frequently raised. A common observation was governance lag – how policies and regulations trail behind the pace of AV technology development and its push for implementation. This situation is not unique to Israel as other countries and cities also prepare for future mobility by shifting to urban experimentations. Such experimentations refer not only to technological trials of AV vehicles, but in a broader sense of urban, social, and regulatory experiments. AVs bring about new challenges for governance to be more responsive and flexible, with tolerance for errors and the willingness to reconsider the validity of old rules. For instance, there would be a need to review regulations that preclude the possibility of AV services. Current Israeli regulations prohibit those without a taxi licence to offer ride services in deference to the demands of taxi drivers’ unions. As one respondent from academia explained, ‘in Israel, taxi drivers have a very strong lobby. That’s why we don’t have Uber here. Here, you need a licence to transport people for payment. There is no political incentive [to change regulations that would affect the] taxi drivers and the bus drivers … that is 20,000 jobs.’

Another issue was the allocation of governance responsibility and ownership in the complex nature of AV implementation: infrastructural investment and coordination, standards and regulations, equitable access, and data management, etc. An NGO respondent pointed to the ‘constant struggle between the local authorities and the national government [on ownership of responsibility]. The national government wants to dictate things, but they’re disconnected from what’s going on on the ground’. A city official noted that ‘when it comes to transport services, we don’t have authority – not to plan, not to regulate, not to finance. But we know that these AVs first will affect especially the big cities. So, this disruptive technology will affect us very much, but we do not have the authority to regulate it’. Respondents who spoke of dealing with AV futures as a part of a broader mobility transition in Israel revealed deep-seated issues such as a lack of strategic and long-term urban and transport planning, and difficulties in coordinating across sectors.

In summary, values of high recognition and high dissension can sometimes lead to antagonism and conflicts in the deliberation process but also positive results. An outcome unfolded thus far in Israel’s case is that less desirable alternatives – an industry-driven approach for accelerating AV implementation now take deference to a stronger focus on strategies and policies. The value of efficiency in AV implementation has moved towards better problem framing as professionals are exposed to a range of efficiency rationales held by others. In governance, a shift has begun in aligning issues of AV implementation at a strategic level – even as dissensus still abounds in approaches – to define goals, and formulate a common language and a practicable work plan.

Low recognition, low consensus: values of low social accord

Values of low social accord were those receiving low recognition and low consensus. In this study, these values related to the twin facets of equity: ‘spatial equity and access’ and ‘social equity and inclusiveness’. The two other values ‘power parity’ and ‘freedom and choice’ also have equity implications.

Spatial equity and access are related to ‘social equity and inclusiveness’ but focus on the spatial dimension. Some respondents saw AVs as an opportunity to address the urban-periphery divide and patchy public transport coverage in Israel. Through simulations studies (by themselves or others), several respondents found that if AV deployment through ride-hailing and sharing were left to market forces without active policy direction, service deserts would be perpetuated and created. A simultaneous retreat of public transport provision would exacerbate spatial inequalities. This prompted some respondents to question the need for policy intervention to direct AV service provision to peripheral or underserved areas. A couple of public sector respondents described the learning curve of regulating a different mobility mode – shared electric scooters, which may be deemed as a precursor to future mobility modes such as the AV. Tel Aviv had gained a sudden and unexpected popularity of shared electric scooters provided by private companies. The Tel Aviv municipality experienced overt spatial injustice where these scooters were configured to simply stop working when they reached their ‘service boundary’ which often excluded the ‘bad neighbourhoods’. This raised a red flag to the service regulators who outlawed such practices and used this as a learning case on spatial discrimination for future mobility modes. A public sector respondent described, ‘When these electric scooter schemes came up, they did not operate in specific areas of the city. The operators were afraid of vandalism, theft and damage. Once you passed a specific neighbourhood, the scooter would shut down. The first thing we talked about [as policymakers] is that this cannot happen. If something operates in the city, it either operates in the entire city or not at all’. This learning experience propelled the value of spatial equity and access into greater recognition (as indicated by the arrow) in transport policymaking.

Social equity and inclusiveness pertain to questions of equity for social groups which may be excluded from a transport system, irrespective of residential location. Respondents perceived a general lack of discussion to direct AV implementation to improve accessibility for groups such as the disabled, elderly, and non-drivers, which AVs were purported to help. A respondent from an NGO remarked,

In Israel, 30% of families do not own a single car. Over 40% do not have access to cars in Tel Aviv. However, when there’s a discussion about transportation, when the focus is not on cars but on people, the people being focused on are the car owners as opposed to looking at the full spectrum of people trying to enjoy life in the city. So, it is a very lop-sided discussion. The issues are too focused on capacity and travel timesaving. And that is just looking at those who travel more and for whom travel time savings are of higher value.

The few respondents who raised this topic commented on the lack of recognition and effort in addressing potential exclusions in various demographic groups. These include those more likely to face technological barriers such as the strictly-Orthodox Jews, or those reliant on public transport but have restricted travel options when there is little or no public transport tionwide on Fridays and Saturdays.

Another issue of equity and inclusiveness raised is how AV implementation can augment public transport or detract from it. An important social function of public transport is public benefit. One respondent from a government-owned enterprise opined that transport is a public right requiring a ‘minimum guaranteed service standard’. She argued that there is a need for ‘a greater recognition of mobility as “up there” with health and education’, suggesting that the value of equity is currently not accorded due recognition (a proxy for importance or priority). This reflects a dissensual viewpoint at a deeper level regarding the right societal prioritization of a value.

The value of power parity was raised by a few respondents observing that AV implementation was often between unequal partners across different sectors, partly due to differentials in technological expertise and influence. These respondents raised caution against an AV implementation predominantly driven by market forces, and when the government retreats from its role to secure public interest in guiding this implementation. A government official noted that even though AV implementation in mass transit would be more beneficial, the large majority of AV companies approached the municipality to test the typical 5-seater vehicles:

I think most of the major players are focusing on the private car. That’s what they drive every day and that’s how they see it and I think that’s the big problem. The people who develop or invest in it don’t use public transportation, or rarely do. They drive the car. And when they dream of an AV, they dream of a Mercedes which is autonomous. They don’t dream of a bus driving [autonomously].

In this sense, the envisioning of AV futures has already been subject to the kind of solution formulation by AV companies which does not deviate much from existing car-dependency trajectories of the 5-seater sedans or ride-hailing models. This has perhaps already insidiously foreclosed the imagination of other possibilities and practical implementation that can pervade the entire AV industry in designing components (sensors, batteries, etc) and urban infrastructure (charging locations) configured to suit private automobiles. This broaches the question of how society can have a say in the kind of technology it selects.

In summary, values of low social accord at an early stage of the discourse are less likely to be sufficiently addressed to pre-empt negative implementation pathways. These may be discovered during actual implementations – a stage of more entrenched technological lock-in when it would be more difficult to make changes. This may point to the increasing relevance of urban experimentations of novel technologies on a limited scale, in parallel to robust societal deliberations. Most values of low social accord in this study come under the umbrella of social justice, signalling a key area of possible neglect. This is notwithstanding a few respondents from the NGO and public sectors who ascribed strongly to these values and had deep knowledge of the issues relating to these values. These values are deemed ‘low consensus’ as there appeared to be little agreement on how to prioritize them. This kind of dissensus – the disagreement in assigning importance to different values – potentially goes deeper than the operative dissensus discussed earlier.

Low recognition, low dissensus: subordinate values

Values of low recognition and low dissensus are subordinate values and were the least frequently mentioned and were not perceived as disputed values. In this study, ‘technological accountability and transparency’, ‘citizenship and democracy’ and ‘public realm and civic life’ fall within this group. These may be considered as low dissensus values rather than values that have achieved consensus per se.

The value of technological accountability and transparency was addressed either at the level of narratives (rhetoric about the technology) or at the operative level of governing, regulating the algorithms, and attributing the liability of AV technology. On the former, several respondents found the overly-optimistic narratives of AV technology as a panacea for transport problems as ‘misleading’, ‘dangerous’, and based on ‘false arguments’. A public official pointed to the challenge of the technological black box facing non-technologists, particularly for regulators to verify and validate the claims of AV manufacturers, and to navigate the legal and liability issues.

Not many substantive issues were brought up on the topic of citizenship and democracy, such as what AV implementation means for democracy, or enhancing societal involvement in shaping a more democratically-oriented AV implementation. Issues raised revolved around public engagement. Several respondents remarked that the general public does not yet have sufficient exposure to the technology, and it was still too early to get a meaningful sense of public opinion on AV implementation. One commented, ‘If you are just a civilian thinking about AV, you just think that it is a car that drives for you. It sounds like the future, you don’t know if you want it … So right now, we may not get a quality discussion with the public.’ This may point to the need for the framing of AV issues into topics relevant to the public to draw them into discussions. Similarly, other issues in this quadrant – ‘public realm and civic life’, ‘environmental sustainability’ and ‘privacy’ – may need contextualized framing to better understand their implications. There were limited critical angles raised on issues stemming from these values. This could mean that respondents have not drawn strong connections between the implications of AV implementation and these (apparently) abstract values lacking an immediate relation to transport issues.

Public values: the missing link between public interest and policy goals?

Values are often implicit in discourses and require probing to define what they are. With rare exceptions, there is a lack of consistent societal discourse at the level of public values. Drawing from the above conceptual development and empirical findings, this section presents arguments to enhance the conceptualization of public values. Towards concept operationalization, we elaborate on how public values may take on a stronger role in societal deliberations and policymaking.

The first conceptual argument is that public values are the important but often missing level of specificity that makes the logical and substantive bridge between public interest and policy goals. Public interest can be considered the amalgamation of broad and multi-faceted public welfare or the common good. Public interest is often a broad goal of governance and public service (Davidoff, Davidoff, and Newton Gold 1970), which is however vague and ambiguous (Lewis 2006). Policy goals, in contrast, are part of the specific and practical aspects of policymaking. Yet there is a conceptual gap between the construction of policy goals and how it addresses public interest. To strengthen the conceptualization of public value, our first assertion is that public values, being conceptually prior to policymaking, are helpful in the societal deliberation and formulation of policy goals.

The second argument is that the set of relevant public values and the priority (level of importance) of each value within the set are context-dependent and determined through societal deliberations. We propose that public values can be specified as a set of context-relevant principles, ideals and beliefs that safeguard and enhance public benefit, attending especially to (potentially) marginalized groups. The third argument is that constructive dissensus is useful or even instrumental in reaching substantively better and politically more legitimate solutions in societal and policy deliberations.

From these conceptual arguments, we delineate the benefits of public values in public policy practice:

1. Public values clarify the principles, ideals and beliefs important for the securing of public benefit on societal issues.

2. Public values are building blocks for creating policy goals amid policy, societal, and technological uncertainty.

3. Public values build and clarify linkages between the general ‘public interest’ and actual policy goals.

4. Public values accommodate diversity through attention to dissensus. This ensures attention to the values of minority, and possibly marginalized, groups.

5. Public values enhance politically legitimate and inclusive processes.

Especially in conflictual, complex or unprecedented circumstances, there is a need for public value-based policy goals constructed through democratically reasoned societal deliberations. If public interest can be seen as a monolithic concept that does not differentiate between heterogeneous people and needs, and even prioritize the majority at the expense of minority groups, focusing instead on public values can bring advantages. Public values are more specific and contextual than the concept of public interest. They are socially constructed from each time, place, and socio-technological context by stakeholders, public representatives, and civil society. Yet public values are not overly specific for application in new and shifting pre-policymaking landscapes that accommodate changes in the sociotechnical realm. Public values are conceptually prior to policymaking. That is, these values do not function at the level of detail of specific policies but at a higher level where they can guide policymaking. When public values are specified, they become the crucial and traceable linkages between ‘public interest’ and policy goals. This serves to strengthen commitment, transparency and accountability of public policymaking to public welfare.

Conclusion

In this paper, we have argued that public values should be integral in social deliberation and decision-making because they help align public policy with important societal values. We contend, firstly, that societal deliberations that encourage a diversity of values, including conflictual values (rather than seek consensus as a goal), better contribute to publicly beneficial outcomes. Secondly, we posit that public values create the meaningful link between broad public interest and specific policy goals. In a policy vacuum, public values can serve as a basis for creating policies aligned with public interest.

We subsequently presented the Recognition-Consensus Model as a conceptual means to assess how public values are recognized or neglected, and receive consensus or dissensus in a particular case. Applying the model to Israel’s case of AV development and implementation, we find that issues undergirded by the values of safety and efficiency are dominant and may crowd out other relevant values, foreclosing deliberation processes and alternative possibilities. Some key values such as equity and inclusiveness, citizenship and democracy, and environmental sustainability are currently in low recognition in Israel. An awareness of possible (societal and policy) neglect of these values amid (newly) developing deliberation and policymaking processes can help redirect attention to societal goals and aspirations. Regarding dissensus values – such as efficiency, governance, and spatial equity, which emerged in Israel’s case – we argue that accommodating dissensus enlarges the spectrum of ideas and perspectives while potentially eliminating inferior ones.

A limitation of the analysis through this model is that (like other qualitative methods) it is reliant on the researcher’s interpretation of how values are classified. Sometimes it may not be obvious whether a value is sufficiently recognized. For instance, equity issues may be highly recognized amongst social scientists, but to a lesser extent in other communities. There is a need to not assume a homogenous professional community.

While the theoretical model is applied here as an academic study, it can potentially also be used in policymaking. Stakeholders from the public sector could use the tool to critically explore whether certain values dominate the ongoing decision-making process, while others are neglected. Such an analysis could be carried out in various stages of the decision-making process, including problem identification, goal-setting, or problem-solving. It could also be employed to analyse processes or outcomes of a decision-making process, such as focus group discussions, policy documents, or white papers. Future studies could test the usefulness of the model in public sector policymaking. While the model could be adopted by civil servants, we surmise a stronger potential for its use by those who can take a disinterested position – academics, NGOs and think tanks – in policy critiques.

Acknowledgements

We thank the interview participants for sharing their insights in this study, and WISE-ACT COST Action CA16222 on Wider Impacts and Scenario Evaluation of Autonomous and Connected Transport for facilitating fruitful discussions. We are grateful to the two anonymous reviewers whose comments helped to sharpen the arguments of this paper.

Disclosure statement

No potential conflict of interest was reported by the authors.

Supplementary material

Supplemental data for this article can be accessed at https://doi.org/10.1080/14719037.2023.2171094

Additional information

Funding

This study has been partially funded by the Israeli Smart Transportation Research Center.