Mini-Publics as an innovation in spatial governance

Mini-publics—deliberative fora made up of randomly selected, representative groups of citizens—have attracted considerable interest as a means of resolving perceived weaknesses in existing forms of governance. In this paper, we consider the use of a mini-public or citizens’ assembly to constitute an ad hoc governance space based on the Travel to Work Area of Cambridge in the United Kingdom rather than working within the existing local government boundaries within which transport infrastructure is usually governed. Through this case study, we explore the question of embedding mini-publics in the wider processes of policy and decision-making. More specifically this is the question of the extent to which they ought to be permitted to inform and even assume responsibility for local-level transport policy decisions. We argue that, if they are to become more widely used, then it will be necessary to understand the practices associated with such democratic experiments and their potential to transform existing governance networks in contested areas of spatial policy.


Introduction
The use of deliberative fora to govern can be traced back to the Athenian city-state.However, this approach to governance has also experienced a more recent renaissance.Ever since Robert Dahl's call in the 1970s for their reintroduction to city governance (cited in Smith and Setälä, 2018: p. 2), a "deliberative wave" has ebbed and flowed with numerous experiments in their use, often in areas like urban and infrastructure planning and often at the local level (OECD, 2020;Bussu et al., 2022).Despite this, recent academic interest has concentrated less on questions that are explicitly or directly spatial.Work by political scientists has focussed instead upon the use of such fora to address constitutional or national issues (Ryan and Smith, 2014;Lafont, 2015;Smith and Setälä, 2018;Farrell and Stone, 2020;Smith, 2021).Examples of their use in addressing infrastructural and spatial questions still tend to focus upon the issues of interest to political science such as legitimacy and quality of deliberation (Niemeyer, 2011).Nevertheless, an important question that emerges from this literature is the nature of the connection between small-scale experiments in deliberative democracy, decision-making structures and political institutions (Hendriks, 2016).If what are generally referred to as mini-publics improve the quality of decision-making through deliberation that enhances legitimacy and makes up for some of the deficits in representative democracy (Mackenzie and Warren, 2012;Kuyper and Wolkenstein, 2019), then there is a strong argument that they ought to play a greater role in these processes.Yet, despite this, the nature of such embedding has until recently been poorly defined (Bussu et al., 2022).Clearly there is a question of the strength of connection between decision-makers and civil society as well as between debate and activism in the public sphere and the specific deliberation that takes place in mini-publics constituted to address infrastructural questions (Hendriks, 2006(Hendriks, , 2016)).Beyond these connections to the wider "deliberative system" (Parkinson and Mansbridge, 2012), we concur with Bussu et al. (2022) that there remain important and productive questions about the "relationship between participatory processes and the practices of both civil society and public administration, within different spaces" and that this is something best achieved by defining embeddedness in relation to larger social structures.
In this paper, we address the question of embedding through the use of a mini-public in one such space, namely, the polarised debates around transport infrastructure.With road space a finite resource contested by different users and the expansion of road space generating further conflicts over the social and environmental impacts of such action, transport planning is inherently agonistic and "inherently political" (Legacy, 2016).Furthermore, knowledge is often contested (McArthur, 2019), and the presence of powerful, technocratic, pro-growth and pro-car (Walks, 2014) discourses has produced considerable scepticism among the general public towards attempts to contain the politics of transport within participatory exercises that are primarily intended to achieve consensus (Legacy, 2017).Moves to inject deliberative democracy into such a contested realm are by no means new.For example, there is evidence of such approaches being applied to the planning of major infrastructure in France (Leheis, 2012;Marshall, 2016).Grounds for doing so may be more instrumental, based upon the assumption-not always borne out-that it is a tool for more sustainable forms of mobility planning (Hajer and Kesselring, 2007).They may also rest on less defined normative assumptions about deliberation as improving decision-making quality and building trust in democratic institutions (Bloomfield et al., 2001).
Here, we examine the case of the Greater Cambridge Citizens' Assembly (GCCA), one of three experiments in using mini-publics at the local level as part of a national government programme.The assembly was deliberately constructed to bridge the rural/urban divide and reach beyond political boundaries by being based on the travel-to-work area (TTWA) of Cambridge in the United Kingdom.What links this initiative to current debates in democratic theory and practice is the use of a citizens' assembly or mini-public based on sortition (the stratified, randomised invitation of participants).As this was conducted within a predetermined geography, it saw the experimental creation of a new public, one apparently distinct from existing publics and the networks of diverse interests formed around transport planning.This has clear implications were the practice to become more widespread.First, there is the "designability" of mini-publics (Kuyper and Wolkenstein, 2019), which may be used both to target acknowledged or perceived deficits in democratic institutions and include normally under-represented groups (Smith, 2009;Farrell and Stone, 2020).They thus offer the potential to actively convene a group that may differ from the type of selfselecting publics and entrenched relationships that normally develop around such issues.Second, in selecting the TTWA as the scale at which this experiment in governance is conducted, it allows the scope to develop institutions that better align with the issues of infrastructure and its management than the bounded political institutions of local government through which they are traditionally managed.Our findings, however, show that relationships-between the national and local levels, between the local government and civil society as well as between the elected and non-elected sides of local government-remain central to the benefits and challenges that have been identified not only by interviewees involved in the process of governance by mini-publics but also advocates of their embedding in decision-making at this level.
In the following section, we begin by drawing out the features of mini-publics relevant to their potential as a means of informing or possibly making spatial policy decisions.We also use this to identify both normative and instrumental arguments for greater embedding.We then show how a key feature of mini-publics, sortition, has been used in this case to constitute a novel governance unit outside of the boundaries of local government.We discuss a critique of such "soft spaces" and within this the argument that consensus is neither possible nor desirable (Allmendinger and Haughton, 2012;Haughton, et al., 2013).In response, we argue that there is a need to evaluate the transformative potential of such experiments (Savini and Bertolini, 2019) in disrupting the relational networks through which transport is typically governed.In the section on methods, we explain our access to the case study as well as the small group of civil servants and local politicians involved in experiments with governance by mini-publics.We also acknowledge our own normative position and the measures we have taken to correct for any bias this may inject.The case study section sets out the governance context of transport in the Greater Cambridge area and explains how a mini-public was convened as a means of overcoming a degree of political stasis.Our findings then illustrate the way those involved in the process of embedding perceive the relational networks of local government that would be critical in shaping any wider adoption.In our final discussion, we examine the evidence of the case study and different scenarios for future embedding.Our conclusion is that if the use of mini-publics in the governance of such issues as transport is purely instrumental, then critiques of their use to arrive at consensus carry some weight.If, however, the politics they generate proves harder to contain, critical analysis of their ability to facilitate meaningful transformations is still required.

Embedding mini-publics
A number of definitions of mini-publics exist.The more expansive include various types of participatory decision-making (Fung, 2003;Pateman, 2012) from Citizens' Juries and Planning Cells (Dienel and Renn, 1995) to citizens' assemblies.More recently, researchers have identified their two principal characteristics: structured deliberation and the use of representative samples to constitute the sub-group of the population that is invited to participate.Alongside these a third, implied characteristic is that they offer an alignment of policy-making with the considered views of citizens (Ryan and Smith, 2014).Claims that representative sampling through sortition produces a greater breadth of democratic engagement and gives structured deliberation more depth (Curato and Böker, 2016) underpin normative arguments for the greater embedding of mini-publics in policy and decision-making processes.A spectrum of opinion nevertheless exists on the extent to which this is desirable.At one end mini-publics are perceived as consultative and little more than focus groups (Setälä, 2017), albeit consultations that provide high-quality information on public opinion owing to the time spent in deliberation.At the other end of the spectrum is the view that they offer an alternative to representative politics, either replacing it completely (Hennig, 2017) or by "peeling off" policy areas-for example, climate change-and entrusting them to deliberative bodies (Smith, 2021).This is because mini-publics are seen either as being less prone to the short-term considerations generated by electoral cycles (Smith, 2021) or as a means of getting beyond polarised debates within the political or wider public realm that can result in deadlock and thereby inhibit policy-making on contentious issues (Lo et al., 2013;Niemeyer, 2014;Setälä, 2017).
Whilst the question of the extent to which mini-publics are vested with decision-making authority remains an important one, within this there is a more granular set of questions of how they may and do interact with different elements of the polity and existing institutional architecture.The extent to which they are designed-whether passively or more actively-to "couple" with parts of a system such as legislative bodies and activist or elite groupings raises important questions about the risks and benefits this generates for all parties (Hendriks, 2006(Hendriks, , 2016)).These include co-option and the instrumental use of mini-publics to solve problems defined in the narrow terms set by existing institutions, a risk that is mitigated somewhat by the capacity of activist groups to "decouple" and actively engage in protest and criticism of elites or legislative bodies (Hendriks, 2016).The concept of coupling clearly goes hand in hand with concerns about the connections between civil society and the structures and practices of decision-making as seen in the recent discussion of embedding (Bussu et al., 2022).The latter concept, however, casts a somewhat wider net than basic institutional design.In defining embedding in Polanyian terms, the concept raises questions of political economy that shape many of the discourses within which debates on transport infrastructure are conducted.Furthermore, Bussu et al. (2022) draw attention to the spaces in which the embedding of democratic innovations takes place, including their temporality and, crucially for our argument, the practices associated with them.

Sortition and 'soft spaces'
One such space was created by the sortition process for the GCCA, which targeted 10,000 households within the Greater Cambridge TTWA in search of potential participants.During the second stage of the process, these were then stratified to develop a sample that was broadly representative-in terms of age, gender and ethnicity-of the demographic profile of the area.More specifically, participants were selected on the grounds of their (rural or urban) residential location, their work-related travel behaviour (into or out of Cambridge) and their main mode of travel.The selection of the TTWA itself was a response to the challenges that transport poses to administrative geography.Movement of people and goods invariably confounds any attempt at delineation.The arbitrary boundaries of local authorities in England simply accentuate this issue, which helps to explain the increasing role of 'combined authorities' designed around city-regions, of which Cambridgeshire and Peterborough are one.TTWAs emerged as a technocratic solution to this problem in that their boundaries are defined by a rule stipulating a working population of over 3,500, three-quarters of whom live and work in the area and are seen as "looking to" a particular centre (ONS, 2015).In commissioning the GCCA, the Greater Cambridge Partnership elected to recruit participants not from Greater Cambridge alone but from this much wider area.Thus, some participants would not normally be able to influence democratic decisions over these issues as a result of their living in different local authority areas.
As the use of mini-publics in the UK is still at an early stage, it makes sense to see their use as the manifestation of recent, politicised trends towards urban experimentation (Savini and Bertolini, 2019).In treating the GCCA as an example of urban experimentation, it is important not only to assess their performance against the usual claims of enhanced legitimacy and improved decisionmaking that are advanced in favour of mini-publics.The relevance of these experiments would appear to be their capacity to break with and reconfigure existing, stuck governance structures (Hoffman, 2013).Thus it is vital to ask whether they break out of the niche in which they occur and challenge existing relational and institutional networks or whether they die there or are co-opted to the point that they produce little in the way of transformative urban practice (Savini and Bertolini, 2019).
The creation of a geographically defined public through sortition and empowering participants with (as we will discuss) an expectation that their conclusions will influence decisions potentially brings a new spatial governance institution into being.Although the GCCA was unique, a growing number of examples of the permanent institutionalisation of mini-publics at the regional and city level (Smith, 2021) suggests that through such experimental actions the necessary power to transform urban governance could grow (Hoffman, 2013).Thus, it is conceivable that such bodies could play a greater role in areas such as transport planning in the very near future.Were such practices to become widely adopted, publics of this type could be seen to "exist both beyond and in parallel to the statutory scales of government … involving the creation of a new territorial entity which sits alongside and potentially challenges existing territorial arrangements or the dominance of particular scales of governance."(Haughton, et al., 2013, p. 218) If this were to be the case, the nature of embedding becomes critical as the challenges posed increase in tandem with decision-making authority.Therefore, we argue that the critical analysis of such "soft spaces" (Haughton, et al., 2013;Allmendinger et al., 2015) of governance has value in informing questions of how and whether mini-publics ought to be further embedded in decisionmaking processes.This is significant as the experimental nature of soft spaces is seen as having potential value in addressing sensitive, cross-boundary issues (Allmendinger et al., 2015).This is reflective of similar claims that mini-publics can address complex or sensitive issues where existing institutions have failed.
It has also been argued that there are democratic risks to the establishment of such soft spaces, which can obscure accountability and disrupt transparency (Haughton et al, 2013).At one level the practices of sortition, the public nature of the recommendations and, to an extent, the deliberations that underpin them would appear to counter the risk that decisions made by mini-publics are opaque.Questions of accountability are, however, harder to answer (Lafont, 2015).Time-limited membership may have benefits in other areas such as removing short-term political considerations (Smith, 2021), but it makes it difficult to hold members responsible for past decisions.There is insufficient space to explore the issue fully here, yet it is important to acknowledge that for the advocates of mini-publics such problems are resolvable through institutional design.Nevertheless, it must be noted that as one looks beyond the process itself the issue of obscurity returns in the translation of decisions into specific policies.This is pertinent to the question of embedding as it goes to the heart of the related question of the extent to which existing institutions of government are influenced by and compelled to act upon the recommendations that are made by mini-publics.
Central to the critique of soft spaces is the notion of consensus.Criticism of consensus as an objective has been central to a critique (Allmendinger and Haughton, 2012) of an earlier wave of interest in deliberative measures to address the conflicts inherent in the planned use of space (Healey, 1992).The contention is that the pursuit of consensus excludes the agonism inherent within spatial politics and that the creation of new governance spaces is the manifestation of a less transparent, post-political approach that ultimately displaces such conflicts (Allmendinger and Haughton, 2012).Whilst the ability to achieve consensus is not the only claim made by advocates of a greater role for mini-publics, it is, however, still a significant part of their perceived capacity to engender and legitimise innovation.
The critique of soft spaces sits within a much wider debate that has established the binary of agonism versus consensus.Adherents tend to reject the notion of the latter achieved under neoliberal forms of governance as being shorn of inherently agonistic 'politics' (for a fuller discussion, see Legacy, 2016;Legacy et al., 2019).Yet, in contrast to tightly defined mini-publics, the spaces that have been analysed within which consensus is sought are much wider.The deliberation amongst representatives and interest groups identified by researchers in this area (Bickerstaff and Walker, 2005;Allmendinger et al., 2015) appears in marked contrast to the deliberation amongst a representative sample of a public seen as excluding organised interests (Hendriks, 2016).This casts a spotlight on the practices associated with the design of mini-publics, sortition in particular.Is this an innovation that introduces previously excluded forms of politics into experiments in urban governance, or are the spaces in which it occurs so bounded as to contain it within a niche and thus dull its transformative potential?

Methods
The research analysed in this paper draws on a combination of ethnographic and interview data collected by the authors.It follows an iterative approach in three stages beginning with separate ethnographic observations of the GCCA.These provided both insider and outsider perspectives on the process.The former was achieved through one author's participation in the advisory group on the design of the assembly and thus gave access to the senior individuals involved, including their intentions and motivations.The latter was more of the "announced observer" (Gray, 2004) used in studies of infrastructure planning (Rydin et al., 2018), with the second author responding to the offer for interested parties to attend in this capacity.The fieldnotes and informal interviews with participants in organising the GCCA alongside a review of the online video material and the final report formed the basis of the following two sections on the governance of transport in Cambridge and the structure of the assembly.This also informed the questions on what the embedding of the process more widely might look like.These questions were posed during the second phase of the research as a series of ten semi-structured interviews with senior practitioners.Such practitioners were either responsible for delivering the GCCA or local government officers and local politicians involved both in transport planning and from the small group of civil servants and civil society organisations involved in establishing citizens' assemblies in the UK.As these interviews were more formal than the discussions in phase one, they were recorded and transcribed.They were then analysed first individually and then jointly; through this process, a series of categories emerged in a more grounded way.The strongest of these categories-in the sense that they were the most frequently discussed-are described in the findings section below.With one of the authors being a Cambridge City resident and with both authors engaged in transport research in the city, this approach benefited from a degree of ethnographic immersion in and sensitivity towards local transport systems and the politics associated with them (Cresswell, 2012).This informed the final phase of the research, which returned to the wider context of observing both the implementation of the recommendations of the GCCA and its impact on broader debates around transport in Cambridge.
We have been conscious of and sought to control for the risk of speaking only to advocates for mini-publics.The formal interviewees did not include individuals who were openly opposed to or sceptical of the process itself as we were seeking their views on embedding.However, the ethnographic component of the study balanced this through access to groups that were suspicious of the process and to the wider political discourse surrounding transport in and around Cambridge.In addition, there were a range of positions within the sample of interviewees, from advocates and enthusiasts within authorities that were forging a path in the wider use of mini-publics, to officers for whom it offered an interesting experiment, albeit one led by the demands of elected representatives.Of the interviewees directly related to the GCCA, by no means were all of them advocates of the method, with some being involved more on account of their expertise in transport planning.This latter approach acted as a further control owing to the focus on the quality and potential of the outputs produced rather than simply the value of the process.Finally, we have chosen to use the more analytical term mini-publics for these initial sections and return to it for our final discussion and conclusions.However, in the more empirical sections, we have referred to the specific form of mini-publics used in this case-namely, citizens' assemblies-as this was the term exclusively used by interviewees.

Governing transport in the Greater Cambridge area
The Greater Cambridge Citizens' Assembly was one of three mini-publics convened as part of the UK government's Innovation in Democracy Programme (IiDP), which was commissioned jointly by the Department for Digital, Culture, Media & Sport and the Ministry of Housing, Communities and Local Government (DCMS/MHCLG, 2019).This has been just one part of a growing interest in the use of mini-publics by policy-makers and local government in the UK (Hughs, 2019).At the national level, the process had been used around such issues as social care and climate change and also with some interest in their use as a means of resolving the issues generated by the decision to leave the European Union.The IiDP assemblies were distinct in that they invited bids from local government.The result was that the three successful authorities all sought to address issues that were directly and inherently spatial: transport planning in Cambridge along with town centre regeneration in Dudley in the West Midlands and Romsey in the South East of England.
As well as the focus on transport, two further features distinguish Cambridge from the other participants in the IiDP.The first was the responsible authority.The Greater Cambridge Partnership (GCP) is not directly elected (Dudley and Romsey citizens' assemblies were both led by local authorities); rather, the GCP is the local City Deal delivery body established in 2013 to administer this stream of central government funding for a combination of infrastructure and skills development.Its own governance is carried out through a mix of political bodies including the city council, the surrounding district of South Cambridgeshire and the county council-all represented alongside major employers such as Cambridge University and other business organisations.The second distinction was the selection of participants from the Cambridge TTWA rather than the local authority area.This stretches beyond Cambridgeshire (the county level is traditionally responsible for transport) to take in parts of the surrounding counties of Essex, Bedfordshire, Hertfordshire and Suffolk.
The nature and scale of growth centred on Cambridge and its booming economy has long seen considerable and interconnected stress placed upon transport and housing through high demand and soaring prices (While et al., 2004).Cambridge City Council and the surrounding authority of South Cambridgeshire have a history of cooperation to accommodate the housing demand for which the city itself lacks space.More recently, this has been formalised through joint planning.Attempts to deliver sufficient housing have been through a combination of urban extensions and the new communities of Cambourne and Northstowe in South Cambridgeshire.High levels of employment in the city's knowledge economy have also driven the expansion of surrounding settlements in other counties resulting in an expanded TTWA.This housing stress manifests itself within the transport network through high levels of inward commuting from this large hinterland.Inward traffic flows come up against the limited scope to expand road infrastructure, particularly in the city's historic core (Gilligan, 2018;Durrant, et al., 2020).Congestion charging had been previously been proposed twice, once in the 1990s and once in the 2000s, but failed to secure political support.The GCP is the latest in a line of similar partnerships and development organisations that have sought to overcome the mismatch between the scale of growth the city and surrounding areas are experiencing and the administrative boundaries.Such bodies have always received strong support from central government owing to the outsized role Cambridge plays in the UK economy.The latest iteration of this relationship between the local and national level, the City Deals programme, is highly transactional.Central government infrastructure funding is received in return for successful 'bids' deemed in accordance with the (then Coalition, 2010-2015) government objectives on growth and job creation (O'Brien and Pike, 2015;O'Brien and Pike, 2019).
Despite a history of cooperation, governance arrangements remain complicated.At the time of this research, the Cambridge City Council, South Cambridgeshire and the 'Combined Authority' of Cambridgeshire and former unitary authority Peterborough were respectively Labour, Liberal Democrat and Conservative controlled.The 2021 local elections altered this picture with the Combined Authority coming under Labour control.Nevertheless, this was a body that did not contribute to the GCCA and was (under a Conservative elected mayor) pursuing its own transport agenda with poorly aligned outline proposals for a 'Cambridgeshire Autonomous Metro'.This offered a technical fix to congestion through the construction of underground routes serving the historic centre.GCP's own proposals, which include the widening of arterial routes outside of the centre, have attracted considerable opposition.There is also an ongoing campaign against an extension to the Cambridge Guided Busway (a small bus rapid transit system that connects Northstowe and two major employment areas, Addenbrookes Hospital and the Cambridge Science Park, on the edge of the city to the main rail station) to serve Cambourne.Here opponents object to new infrastructure running through greenfield sites and close to villages as opposed to reallocating space for it (and away from private cars) on existing roads.Finally, the city council itself has intermittently sought to manage demand through road closures with considerable opposition and limited success.Proposals were abandoned in 2016 following a concerted campaign by local traders (Gilligan, 2018).More recently, similar, often heated opposition has formed over the temporary closure of Mill Road, a street of largely independent small shops surrounded by residential areas that forms a key route into the city.

The Greater Cambridge Citizens' assembly (GCCA)
It was against this backdrop that a group of 53 participants were selected through the sortition process discussed above.Over two weekends in September and October of 2019, they were tasked with developing a set of policy recommendations in response to the question (posed by the GCP as the commissioning body): How do we reduce congestion, improve air quality and provide better public transport in Greater Cambridge?
The deliberations themselves remained private; these were facilitated by a team from the NGO Involve, which specialises in running participatory events, alongside staff from the GCP and South Cambridgeshire and Cambridge City Councils.There were a range of invited speakers with transport academics from around the UK being heavily represented.In addition to this, there were speakers bringing who had local authority expertise on public health, two individuals representing local business interests and a number of local civil society groups interested in either transport or the environment.These speakers were allowed to present and take questions but were not involved in the deliberation amongst assembly members.Alongside the team from Involve that was facilitating and recording these individual small-group discussions, there was a facilitator on behalf of the independent expert panel established to ensure technical information was accurate and unbiased as well as facilitators from the GCP.
Following sessions in which participants received presentations from-and were able to question-the speakers, the event moved towards structured deliberation of the issues set by the facilitation team.Ultimately, there was a series of ballots in which participants were invited to rank different policy measures and supporting interventions and to provide additional recommendations in written form.The results of these were a very strong consensus around the need to reallocate road space away from private cars to enable public transport.This was manifest both in the level of support and the lack of opposition, with other still popular measures proving more controversial.The second most popular measure by some distance was the establishment of a clean-air zone followed by a range of road-charging options with a flexible charge, to be reinvested in public transport, being the most popular and parking charges the least popular.

Findings
The findings set out in this section illustrate the main categories that emerged from the formal interviews.We begin with the national context as this was by far the strongest theme and was mentioned by every interviewee.We then highlight another area where again there was considerable agreement on the instrumental value of mini-publics as a means of overcoming stasis where polarisation inhibits policies from resolving issues like the reallocation of road space.We then discuss the breadth of opinion on embedding as this underpinned a key question put to all interviewees on the extent to which mini-publics ought to have decision-making authority.Here, as with all the quotes, we have selected those from interviewees that most clearly articulated or developed a key position either within a spectrum of views or shared by all interviewees.In addition, we have removed any possible identifying information.Finally, as part of our contribution to the debate on the relational nature of embedding, we analyse these key quotes alongside data gathered from the ethnographic observations of the process and debates over transport planning in Cambridge.
National political context.Unprompted, all interviewees mentioned the value of mini-publics as a solution to the limitations of representative politics and particularly the polarised nature of political discourse.This is perhaps understandable in light of recent political developments within the country.The Innovation in Democracy Programme was launched during a febrile period in British politics between the EU Referendum and the Johnson administration in 2019-2022.At the national level, this saw the UK struggling to reconcile the constitutional implications of one form of direct democracy with the representative political system.The form of deliberation offered by citizens' assemblies was regularly held up as an alternative form of politics.Yet, owing to the centralised nature of British politics, these national-level relationships also presented an existential threat to the programme.One interviewee who operated at this level was dismissive of attempts by a group of MPs to constitute a citizens' assembly in order to resolve the divisive deadlock over Brexit as it risked providing the "one bad example" that they felt could lead to either the abandonment of the IiDP or failure to further develop the beyond the three experimental case studies.Such fears illustrate the way such innovations can be vulnerable to caprice at the national level.Within the GCCA process itself, however, the issue was clearly considered too divisive and was explicitly 'banned' (albeit in a relatively light-hearted way) by the facilitation team.
The risks of the method becoming associated with a single divisive issue or one political party were perceived differently at the local level.In terms of the structure of the IiDP, two of the three experiments were led by a Conservative authority.In the case of Cambridge, party politics were partially displaced given that the sponsoring body was not a local authority-although one benefit that interviewees identified was the way the approach generated support across party divides.One local government officer described successful attempts by their authority to bring councillors from opposition parties into the process despite this being politically unnecessary owing to the size of the governing majority: "[I]n terms of the numbers, you don't need the [opposition] group, but there is something very much in terms of their cooperation in order to give [the citizens' assembly] legitimacy across the city." This suggests that the process is perceived as having a dual value in legitimating decisions and transcending party-political divisions.
Political stasis.There was also a high degree of consensus amongst the interviewees regarding the type of issues citizens' assemblies were appropriate to address and their instrumental role in overcoming political deadlock.Notwithstanding citizens' assemblies' limitations in addressing a highly polarised political impasse at a national level, many interviewees made reference to early experiences with a citizens' assembly on payment for social care-another political impasse, albeit less polarised-where recommendations had little impact on policy.All interviewees felt that at the local level this was exactly how they ought to be used.In reference to this case study, many of those involved in citizens' assemblies saw transport as an ideal subject for the GCCA.Much of this relates to the information-gathering role of the assembly discussed below.However, this also contains a tacit acknowledgement that existing means of gathering information and resolving issues are inadequate.One interviewee cited the evidence of public protest over climate change as an indication of a level of public concern where it was still hard to gauge what specific policy responses were desired or acceptable.The following description provides a useful illustration of this perspective: "[T]he place that these things are used best is when you take something that you just can't really come to a position on, you know, that's when they're used in the most effective way to help you make a decision on things which are really, where there is stasis or stagnation." The IiDP itself was described as being able to "go to a local authority and say, Give us an issue that you have to deal with that's really hard that you want to work through with your residents, and we will help you do that." This indicates that the assumption citizens' assemblies function as a tool for problem resolution was built into and promoted as a key benefit of the programme.
The question of embedding.In response to questions of how and whether mini-publics ought to be embedded, the same interviewee described the tortuous process of securing ministerial approval for the IiDP.As the interviewee explained: "I was told very strongly, very early on, that if I want to get the Secretary of State to sign up for the programme, I needed to stop talking about democratic engagement in the round." This was significant because it meant that the programme had to be presented purely in terms of its instrumental benefits-or as the interviewee put it, "because it works".Furthermore, any attempt to fully embed the process, in the sense that the outcomes of the citizens' assemblies were to be seen as binding on elected representatives, was explicitly rejected.Another interviewee confirmed the perception that, at the local level, elected representatives' support for citizens' assemblies would largely be seen in instrumental terms: "If [citizens' assemblies] were to be brought into the council processes in our case, whoever the administration may be, I would expect, [elected representatives] would want to exercise some form of control over [it], in particular if they are paying for it, as a mechanism, and they're unlikely … to readily support something which is liable to disagree with [their] political view…" In one sense it could appear that the question of embedding, insofar as it was manifest in this particular case, was largely resolved with the IiDP assemblies, GCCA included, being seen as purely advisory.This view is confirmed by an initial appraisal report (Brammall and Sisya, 2020).Certainly, the majority of the interviewees perceived them in this way.Almost all saw the value of the process largely in terms of information gathering, albeit in a superior form to conventional polling, the idea being that it offers access to "local knowledge" (Yannow, 2003) and "better answers" than those reached within a local authority.
Another interviewee developed this theme by suggesting a deeper understanding of how opinions are formed: "It tells you how people think about something.It doesn't necessarily tell you that, having thought about it, you should do what they then conclude, but you should understand how they came to their conclusion." The two interviewees who were the most enthusiastic advocates of citizens' assemblies also stressed the civic value of the process.They saw them as "transformative for how we do democracy" and as particularly appropriate to bridging perceived gaps in involvement in local authority decision-making.It was this practice of bringing in a wider range of voices that generated the only indication that the nature of sortition creates its own form of legitimacy.Here, one interviewee described "representativeness [as] baked into the process" and went on to refer to the way diverse and seldom heard groups who may not normally engage could be brought in.At the other end of the spectrum, there were more critical comments questioning the legitimacy of basing decisions on such a small sample: "[I]t's only 50 people, right?So why do 50 people get a say over all this?"This suggests that there is nevertheless a long way to go before Fishkin's argument that sortition generates theoretical representativeness (Fishkin, 2020) gains any traction against the belief that representative politics provides the main, possibly only source of legitimacy.
Whilst the view that legitimacy was derived from representative politics was strong across the board, further evidence makes it hard to simply relegate citizens' assemblies to an advisory role.The way the question of the mandate generated by assemblies re-emerged was through the extent to which the binding nature of decisions worked in practice.This was described as a strong commitment to the outcome coupled "with a presumption in favour of implementing".Another interviewee went further still, describing this as "not a formal commitment but there is a real political will and commitment to do it" and ultimately: "I think we are not ready to say this is binding, but in practice that is how it ends up working." This commitment from elected members to respond to the findings of the assembly was also a feature of the GCCA.This was a point made by electoral representatives at both the beginning and end of the process-the only points where local politicians in their role as elected representatives were permitted to attend-and was clearly presented to participants as an expectation that their conclusions would influence policy.

The relational nature of embedding and resistance
In terms of the relationships challenged by the embedding of citizens' assemblies, the most prominent, cited by elected and non-elected interviewees alike, was the risk that resistance would arise from the way they appear to challenge the connection between representatives and the electorate.In Cambridgeshire at least, one elected representative from a surrounding authority directly questioned the legitimacy of the process and any conclusions reached by it.Furthermore, a number of interviewees cited a perception among more sceptical elected representatives that even the information-gathering function of citizens' assemblies might not be attractive.One interviewee paraphrased this as the belief that Others felt that, as with the comments above, electoral concerns would always dominate and that, for example, the breadth of opinion polling would ultimately be preferable to the narrow but deep information on public preferences gained from citizens' assemblies, a notion supported by the belief that the former offers a better insight into voting behaviour.This scepticism was not a rejection of citizens' assemblies as a whole but more in line with the view that they work best in certain situations of stasis where the information is inconclusive.
The second key set of relationships disrupted by the formation of alternative publics through citizens' assemblies was between the authority and civil society.This was acknowledged as potentially problematic in that such groups had built up relationship with their authorities over many years.A number of interviewees were conscious of the sensitivity of this matter.By way of example, one interviewee explained how one elected representative had explicitly seen a citizens' assembly as an opportunity to bypass these relationships, quoting them as saying "we don't want the usual suspects"-a term often used (sometimes in a pejorative sense, although never in this way by any of our interviewees) to describe groups or individuals who can be relied upon to engage with specific issues.Their response, from an officer's perspective, was more sceptical of the value of doing this: "Well, actually, maybe you do.There is nothing wrong with the usual suspects because these people do have capacity.They do understand the issues … and the history." This interviewee went on to speculate on possible mechanisms for bringing this form of local knowledge into the process.
This highlights the complexity of the relationships between civil society and local authorities, some of which have been built up over a long time.The way publics form in response to unpopular policies means that relationships can be highly adversarial, but they can also be collaborative as the inclusion of some groups within the GCCA shows.They can also change depending on the issue or over time.Indeed, in this case the most critical voices of the process itself came from members of groups already in conflict with the GCP and local authorities over proposed transport infrastructure such as the Guided Busway Extension.Whilst some groups lobbying for better cycling facilities, for example, were included as expert witnesses, those lobbying against particular schemes were not.They had argued they ought to be allowed to participate as members of the assembly.In this case, their participation would have to have come about through sortition and was highly unlikely given that the selection criteria were geographic and socio-demographic rather than opinion-based.Even though the process produced recommendations that were possibly closer to their perspective in that demand management was seen as preferable to expanding the city's infrastructure, they were still highly critical of the GCCA and their perceived exclusion from it.This was against a background of scepticism towards the GCP as an unelected and-as they perceive it-opaque body.

Discussion: The transformative potential of mini-publics
Whilst interviewees accepted the politics of transport infrastructure as a space in which mini-publics had value, they often acknowledged the tension that can be generated between civil society and the authority if they are designed with a view to sidestepping unwelcome information that comes from more adversarial positions.This was an issue for the GCCA, which in a third relational dimensionone that interviewees may have been uncomfortable in voicing yet is clear in the construction of the GCCA-permits officers to bypass relationships between elected representatives, civil society and the electorate in defining and constituting an alternative public and controlling the framing of the deliberations.On the question of the extent to which the recommendations of the assembly have proved binding, evidence does point towards an acceptance of the need to act upon them.The board has made a formal response (Greater Cambridge Partnership, 2020), and requests for regular progress reports have been received favourably.The GCP is pressing ahead with a programme of experimental road closures that began in January 2021.One reading of this is that the exercise has revealed there is "significant appetite" for this type of "bold measure" (Greater Cambridge Partnership, 2020).Our own conclusions, based on the framing of the question, the selection of expert opinions and the overall construction of the process, are that this was largely the answer the GCCA was intended to produce.This was confirmed in follow-up discussions with interviewees.The opportunity afforded by the IiDP was used instrumentally by the GCP as a means of re-opening and legitimising a range of policy initiatives that had previously been attempted and then abandoned in the face of public opposition.
Under such a reading, the findings of the experiment have been co-opted.And criticism of the consensus reached by the GCCA around the desirability of demand-management solutionsspecifically, that it is an exercise in obscuring rather than engaging with the agonistic nature of transport politics-appears well founded.Yet, before dismissing the consensus generated by this experiment, it is necessary to consider the practices and their relation to the wider context over time.It is too early to draw such conclusions.Responses to the GCCA are still ongoing, policy related to it is still being implemented and it may be through the continued engagement with the assembly through reporting mechanisms that it will continue to exert an influence.Further transformational impact is possible given that it has revealed a public that are, under the right conditions, supportive of demand-management measures.This should in itself not be underestimated as the consensus generated may be sufficient to enable political action.Vocal pro-car lobbies are more able to assert themselves in more polarised public debates than in deliberative spaces.Were elected representatives to be empowered by the assembly to then activate this public and develop an electoral mandate around demand-management measures, there is potential even in this instance for beneficial outcomes to emerge from reconfiguring the relationships between representatives, civil society and the electorate.Thus, the experiment itself may die in its niche, but its findings could still go on to achieve some form of transformative impact.
A second alternative could see the experiment retained yet still co-opted.Were mini-publics to be used less at the higher level of agenda-setting and move towards trying to resolve the polarised debates generated by specific road closures or new infrastructure, there could be considerable potential for the approach to transform transport decision-making.As our findings show, this would require a means of including the interest groups and representing a spectrum of positions that the practice of sortition seeks to exclude.Furthermore, this would be dependent on the flexibility of the existing governance networks to admit new relationships and potentially a source of alternative power.In this case, the potential instrumental benefits from generating sufficient consensus to act comes with a catch: Opposing groups would have no incentive to participate in a process, even less so to commit themselves to be bound by any conclusions without a commitment for those conclusions to be equally binding on decision-makers.
The final scenario where transformational impact may be achieved depends upon the claims that introducing deliberation into a representative system benefits the system as a whole.The GCP board meeting of 19 February 2020 contained a section on the formal response to the GCCA.The meeting itself was disrupted by activists from Extinction Rebellion (a group calling for action on the climate crisis, with a citizens' assembly to actually decide upon such action being one of their key demands).
The protestors were critical of the slowness to act against climate change and posed a direct challenge to the current political economy of transport infrastructure with particular emphasis on the disproportionate focus on economic growth in the GCP's attempts to decarbonise transport.Cambridge is a comparatively small town and as a result a number of participants in the assembly were known to the authors.Therefore, it is not especially surprising that one of the demonstrators was also among the group randomly selected to participate in the GCCA.This provides some insight into the way one assembly member perceived attempts by the coordinators to use their agendasetting position to filter out issues that were a cause of some discomfort to the GCP (e.g., the relationship between economic growth and climate change) as largely ineffective.This same assembly member also criticised the process as being insufficiently binding (Gardner and Scialom, 2020), which suggests that even in this case the politics of transport might not be so easy to contain.
Despite attempts to co-opt this experiment in deliberative governance, transformative potential could be achieved were groups outside of the more rigid governance networks to 'de-couple' the adoption of either the methodology or the findings and offer their own interpretation.In one sense, the above example suggests scope for a more radical role for mini-publics as a means to inject a new set of relationships into the process and disrupt the existing distribution of power within these networks.Other groups might well also (legitimately) point to the GCP's selective interpretation of the GCCA's call for demand-management measures in the city centre whilst expanding and constructing new public transport infrastructure in preference to reallocating road space elsewhere.

Conclusion
Any critique of consensus that guards against co-option is clearly of great value, perhaps even more so when it is applied to experimental spaces for deliberation.Thus theory provides a valuable starting point for any conclusion.However, if it becomes an end point, if theoretical positions extend to a wholesale rejection of the role of consensus within a certain governance space in shifting public debate, thereby enabling political action or legitimating more radical interpretations of shared positions, then it risks limiting any analysis to a binary, either consensus or agonism.Seeking to resolve such questions purely by recourse to theory denies the opportunity for insight into the dialectical relationships between the two that develop as people involved in the GCCA use both consensual and agonistic strategies to navigate policy implementation, limits and the politics of transport infrastructure.We would argue any analysis of how these relationships are played out in a particular context calls for the full range of methods that take a wider perspective on the practices, spaces and temporalities associated with the embedding of structured deliberation within the complex network of governance relationships.Indeed we would argue methods that look beyond a specific instance of deliberation are essential in tracing the ripples that emanate from such instances through ongoing debates and conflicts.Here we would suggest critical analysis might be better directed towards a more pressing question of whether consensus achieved through deliberation within mini-publics is or is not a vehicle for overcoming political inertia (a situation that often favours entrenched interests and the status quo) and for transforming policy-making in areas where the status quo is untenable.

Declaration of conflicting interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

"
[i]f I [as an elected representative] want to understand what people want, I'll just go and knock on doors and talk to my constituents.I know what they want; they voted me in."