Counter-Governance: Citizen Participation beyond Collaboration

The theory and practice of urban governance in recent years has undergone both a collaborative and participatory turn. The strong connection between collaboration and participation has meant that citizen participation in urban governance has been conceived in a very particular way: as varying levels of partnership between state actors and citizens. This over-focus on collaboration has led to: 1) a dearth of proposals in theory and practice for citizens to engage oppositionally with institutions; 2) the miscasting of agonistic opportunities for participation as forms of collaboration; 3) an inability to recognise the irruption of agonistic practices into participatory procedures. This article attempts to expand the conception of participatory urban governance by adapting Rosanvallon’s (2008) three democratic counter-powers—prevention, oversight and judgement—to consider options for institutionalising agonistic participatory practices. It argues that these countergovernance processes would more fully realise the inclusion agenda that underpins the participatory governance project.


The Collaborative Turn in Urban Governance
The theory and practice of urban governance in recent years has undergone both a collaborative and participatory turn (Bingham, 2006;Fung, 2004). In many ways the two are very much connected. Ideas of collaborative governance have become inextricably linked to hopes for increased citizen participation in policy-making (Bussu, in press). Though originally rooted in "neo-corporatism" (Osborne, 2006), a concern to bring "multiple stakeholders together in common forums with public agencies" (Ansell & Gash, 2008, p. 543), it is the citizen as stakeholder that has particularly captured the imagination. Incorporating citizens or local communities as partners in the policy-making process is intended to address democratic malaise by repairing the relationship between citizens and the state. It has been conceived as a response to the dual problems of citizens' declining interest and participation in democratic politics and the lack of re-sponsiveness of policy outcomes to citizen and community needs (Barnes, Newman, & Sullivan, 2007;Geissel & Newton, 2012). This new urban governance approach has therefore become bound with a concurrent deliberative and participatory democratic project. Governance theorists have made "empowered participation" a defining tenet of the new approach (Torfing & Triantafillou, 2013), and stressed how "collaborative governance is characterized by dialogue and deliberation" (Bingham, 2006, p. 817). Moreover, participatory and deliberative democrats interested in deepening citizen participation have made collaborative governance initiatives a prominent category in typologies of democratic innovations (Geissel & Newton, 2012;Smith, 2005).
This strong connection between collaborative and participatory urban governance has meant that citizen participation has been conceived in a very particular way: as varying levels of partnership between state actors and citizens. The urban space is, however, both a site for the formation of new collective solidarities and a site of deep and enduring conflicts, for instance; increasingly sharp conflicts over the possession of space between local residents and global capital (Sassen, 2011). "Immovable resistance" is a frequent issue for urban governors; one for which the institutional response is often vilification of resistance or an attempt to bypass it, in the process exacerbating the conflict (Inch et al., 2017). How to constructively engage with citizen resistance is thus an important problem for the practice of urban governance. Agonistic democracy, with its focus on transforming antagonistic relations between implacable adversaries into agonistic relations between legitimate opponents (Mouffe, 2000a), holds some promise in this regard. Nonetheless, participatory urban governance, with its focus on partnership through collaborative dialogue and deliberation, gives little guidance on how citizens can engage in agonistic practices in the face of enduring conflicts.
This article addresses this lacuna by adapting Rosanvallon's (2008) three democratic counter-powersprevention, oversight and judgement-to consider options for institutionalising agonistic participatory practices in urban governance. In elaborating these forms of counter-governance it is demonstrated how participatory urban governance's over-focus on collaboration has led to: 1) a dearth of proposals in theory and practice for citizens to engage oppositionally with institutions; 2) the miscasting of agonistic opportunities for participation as forms of collaboration; 3) an inability to recognise the irruption of agonistic practices into participatory procedures. It is further argued that this has had the effect of curtailing the inclusion agenda that underpins most participatory urban governance initiatives, since they have failed to consider unequal influence after the point of the decision and force citizens into a particular relation to institutions, privileging those who want to collaborate over those who want to contest. Institutionalising additional, counter-governance opportunities would mitigate these defects and broaden inclusion.

The Agonist Challenge to Collaborative Governance
"Collaborative governance" is used as a shorthand throughout this article to refer to a family of approaches that have been variously termed collaborative governance (Ansell & Gash, 2008;Bingham, 2006), cogovernance (Smith, 2005;Talpin, 2012), New Public Governance (Osborne, 2006), and co-production (Bovaird, 2007;Durose & Richardson, 2016). This family of approaches is often presented in paradigmatic terms, as possessing its own distinct values and practices. The New Public Governance has been situated as an alternative political-administrative system to Classical Public Administration and the New Public Management (Osborne, 2006;Torfing & Triantafillou, 2013). Similarly, collaborative governance is viewed as a replacement for older adversarial and managerialist forms (Ansell & Gash, 2008;Bingham, 2006), and Archon Fung has described partici-patory urban governance as "a third path of reform that takes its inspiration from the traditions of civic engagement and participatory democracy rather than publicmanagement techniques or competitive markets" (2004, p. 9). So what characterises this new approach?
The common core of these approaches is the way in which the social relations between actors in governance processes are conceived. Collaborative governance, as the name would suggest, is intended to be collaborative. Ansell and Gash's influential definition stipulates that it is characterised by collective and deliberative processes of two-way communication with the aim of arriving at consensus decisions (2008, p. 546). Citizens and officials come together in processes of collective problem-solving and mutual learning, in which discreet but additive knowledge and expertise results in winwin solutions (Durose & Richardson, 2016). Relations between them are not adversarial or competitive; they are based on interdependence, trust, reciprocity and nondomination (Durose & Richardson, 2016;Torfing & Triantafillou, 2013). Opportunities for citizens to participate in governance are therefore conceived and structured in a particular way: citizens are discursive partners, both with each other and with public officials, in a solidaristic search for shared solutions to shared problems.
The elevation of collaboration, characterised as shared endeavour through trust, reciprocity and nondomination, to a paradigmatic value means that collaborative governance has a complicated, often confused, relation to conflict. Collaboration is presented as an alternative to adversarialism and competition. Its advocates do not deny the existence of conflicts and competing societal interests, but these are something to be overcome through collaboration. Collaborative governance is, for instance, in the odd position of claiming it is a remedy for declining trust in public institutions-a way to rebuild the relationship between citizens and state-whilst it also "demands respect for different forms of expertise, nuanced facilitation and relationships of trusts" (Durose & Richardson, 2016, p. 200). As such, it is dependent upon the thing it is intended to produce. It is possible that institutionalising processes that demand trust will generate trust; that behaviour will transform attitudes; and that such processes will act as exemplars, radiating out trust to the rest of the political system. However, it is just as likely that, in the absence of trust, trust-based procedures will founder upon underlying conflicts. This is particularly problematic in a "society of generalized distrust" (Rosanvallon, 2008, p. 11), where reported levels of trust in institutions and other citizens is low and declining, and where increasing diversity is multiplying the potential bases for conflicts.
The implicit assumption that collaboration can overcome any conflict is indicated by the fact the collaborative governance literature gives little serious analysis to how to proceed when irreducible conflicts break out. Durose and Richardson do briefly discuss the issue, retreating from consensus to Dryzek and Niemeyer's idea of "meta-consensus" (2016, pp. 188-190). Nonetheless, this is unsatisfactory for co-governance processes that claim to be about shared decision-making. Metaconsensus is not a decision, but simply an agreement on the basis of disagreement. Nor does it necessarily bring us closer to a decision; elucidating the basis of a conflict may in fact sharpen the conflict rather than foster its resolution. Meta-consensus is not then an alternative coordination mechanism to consensus, it is no coordination mechanism at all. New participatory governance approaches thus seem to be beset by the perennial problem of radical participatory organisation: in the face of irreducible conflicts there is no means of coordination beyond splintering into new sub-groups.
Agonists have a very different conception of social relations. A democratic space is one in which actors face each other as opponents, not as partners. For agonists, "politics is the continuation of war by other means" (Foucault, 2004, p. 15). The democratic project is not a search for shared solutions to shared problems but one of turning enemies into adversaries who mutually recognise each other's legitimacy to inhabit the political space (Mouffe, 2000b(Mouffe, , 2013. Conflict and power are constituent elements of a political relation, not something that can be overcome through politics. The idea that the pursuit of consensus untainted by coercion should be a regulative ideal for assessing the legitimacy of decisionmaking is also rejected, "we have to accept that every consensus exists as a temporary result of a provisional hegemony, as a stabilization of power, and that it always entails some form of exclusion." (Mouffe, 2000a, p. 17). Every decision, even those arrived at through deliberation, necessarily results in the exclusion of some interests, values and identities in favour of others. Any decision is foreclosure of other possibilities. These exclusions are then continually renegotiated, so that a decision is only a temporary cessation of ongoing processes confrontation. If we acknowledge this conception of decision-making as a process of exclusion, moreover, a process of exclusion structured by power, then it becomes paramount to consider the institutional avenues by which those exclusions can be challenged. Whereas collaborative governance has primarily focussed on making decision processes more inclusive, the agonistic perspective points us to the need to also democratise avenues for challenging decisions.
Agonists have recognised the importance of creating these institutional avenues, "to make room for dissent and to foster the institutions in which it can be manifested is vital for a pluralist democracy" (Mouffe, 2000a, p. 17). Nevertheless, agonism has been described as having an institutional deficit (Lowndes & Paxton, 2018). Unlike deliberative democrats, who have invented a range of procedures intended to manifest deliberative democratic principles, agonists have been much more circumspect in proposing agonistic institutional arrangements. This is partly a result of the poststructuralist tradition of deconstruction. Agonist democracy is "deconstructive, rather than constructive. It focuses on the shortcomings and limitations of other approaches and is suggestive of alternatives but refrains from specifying them in any but the most abstract form" (Norval, 2014, p. 77). In addition, in the instances where radical pluralists have touched on the empirical, this analysis has tended to focus on the extra-institutional: the counter-power of civil society (Rosanvallon, 2008); strategies for insurgent political parties (Mouffe & Errejón, 2016); or citizens' informal agonistic practices in the spaces between institutions (Wagenaar, 2014). This appears to be a function of radical pluralists' belief that political science and philosophy has a tendency to focus too narrowly on government institutions and "to exclude and ignore all those wider relations of governance through which individuals and groups are subjected and constituted as actors and political agents" (Griggs, Norval, & Wagenaar, 2014, p. 30).
Recent work has challenged the idea that agonism is incompatible with institutionalisation. Lowndes and Paxton (2018) attribute this to two false premises based on an outdated conception of political institutions. The first is that institutions are a fixed and stable expression of shared values. The second, a corollary of the first, is that to institutionalise is necessarily to reify and universalize those values. The critical institutionalist conception of institutions as contingent means: The key tenets of agonism actually resonate with contemporary developments in institutionalist thinking; indeed, they provide theoretical resources with which to extend such developments further. Rather than a paradox, we find a productive tension. (Lowndes & Paxton, 2018) The next section of this article employs the theoretical resources of agonism to go beyond the dominance of the collaborative governance approach to theorising participation in urban governance. Using Rosanvallon's (2008) three democratic counter-powers as a framework it explores new ways for citizens to adopt an agonistic relation to institutional actors and engage in counter-governance. Moreover, it suggests some ways that existing practices of participation should be recast in agonistic terms.

Three Options for Institutionalised Participatory Agonism
Agonistic practices have been designed into democratic institutions since their foundation. Modern democracies often arose out of circumstances of extreme conflict. Institutional design was sensitive to the twin fears of mob-rule and autocratic monarchy. James Madison, describing the federal institutions of the nascent US, explicitly justifies the separation of power between executive, legislature and judiciary in agonistic terms. The intention was to arrange "opposite and rival interests" in constant tension, "by so contriving the interior structure of the government as that its several constituent parts may, by their mutual relations, be the means of keeping each other in their proper places" (Madison, 1788). Even in the UK, where democratic institutions developed through gradualist reform, agonistic practices still abound. The weekly ritual of Prime Minister's Questions is a prominent example: the leader of the opposition, confronts the government in an adversarial rhetorical contest. Legal trials are another example of widespread institutionalised agonism: prosecution and defence faceoff as opponents in a structured contest over truth. All of these instances of enduring agonistic institutions are, however, examples of elite contestation. They show that institutionalising agonistic practices is possible, yet tell us little about the prospects for institutionalising a more participatory agonism.
This article is concerned with participatory agonism; when actors within a participatory space recognise each other as opponents, thus their relations are oppositional rather than collaborative. This agonistic relation can be constituted in multiple ways. The above are examples of horizontal agonism between elites, and one can similarly find horizontal agonism between citizens. The specific focus of this article is, however, a certain type of vertical agonism: citizen opposition to institutional power. Rosanvallon (2008) proposes three counter-powers that can be wielded against institutional power, which he terms oversight, prevention and judgement. Each posits the people in an agonistic relation with institutional actors, so that counter-democracy broadens the minimal democratic conception of the people as electors to encompass "the people as watchdogs, the people as veto-wielders, and the people as judges" (Rosanvallon, 2008, p. 17).
Urban governance is a good starting point for exploring these popular counter-powers since their historical manifestation was predominantly at this level. They were often directly institutionalised in Ancient city states and medieval towns, whereas the rise of the modern democratic nation state either subsumed such powers into parliaments or dispersed them into an informal social sphere (Rosanvallon, 2008, Chapter 3). In medieval European towns ordinary citizens were tasked with monitoring and constraining the abuse of administrative power through auditing town accounts, and this became a core component of notions of citizenship (Rosanvallon, 2008, p. 77). The practice has long roots stretching back to Ancient Greek cities, which for Aristotle were democratic to the extent that citizens scrutinised the work of magistrates through positions as overseers, auditors, supervisors and ombudsmen (Rosanvallon, 2008, p. 85). These forms of public audit typify the kind of institutionalised participatory agonism that this article defines as countergovernance. Counter-governance concerns mechanisms for citizen opposition or contestation constituted with a direct and formal relation to power, such that these become an explicit organising principle for the coordination of state activity. This differentiates it from Rosanvallon's conception of counter-democracy, which encompasses informal power such as protest.

Prevention
Prevention is the power to obstruct. If government is the positive power to decide upon collective projects, prevention is its negative counterpart. Its most visible manifestation in democratic institutions is the power of the second legislative chamber to block legislation. The notion of popular prevention through the right to resist, founded in the people's capacity for insurrection, predates citizens' rights to vote or participate in government. Early democratic theorists attempted to devise an institutional alternative for insurrection based on a complex notion of popular sovereignty that combined both positive and negative elements (Rosanvallon, 2008). There is a resonance between this negative power of prevention and agonistic democracy as the institutionalisation of dissent (Mouffe, 2000a;Norval, 2014) The agonistic perspective on decision-making provides a clear rationale for the importance of avenues to obstruct or oppose decisions. If decision-making is an active process of creating inclusions and exclusions, then it is impossible to determine ex ante the significant cleavages upon which conflicts will be founded. It is therefore important for those who recognise their exclusion ex post to have means to seek redress. Nonetheless, Rosanvallon sees popular prevention disappearing from democratic institutions as liberal conceptions of democracy began to predominate, so that it is now exercised primarily through industrial strikes, civil society protest and parliamentary opposition.
There have been few calls by advocates of participatory urban governance to strengthen citizen participation in these ex post forms of prevention, despite the fact that institutionalised popular prevention is virtually non-existent in urban governance. It is instructive, for instance, to examine Bingham's (2006) outline of new urban governance processes: citizen participation abounds in "upstream" processes of will-formation and decision-making, but is absent from "downstream" processes of disputation. Overview and Scrutiny Committees are the main accountability body responsible for investigating the policies and implementation of English city governments. The Local Government Act 2000 divided city councils into two functions: Executive and Overview and Scrutiny. Overview and Scrutiny Committees thus comprise of the city councillors who do not form the council executive. They hold the primary mechanism of prevention, the power to call-in executive decisions, delaying implementation whilst a decision is scrutinised and a recommendation on whether it should pass, be amended or withdrawn is made. Whilst it is constitutionally possible for call-in to be triggered by the public, only 2% of English local authorities allow the public to perform this role (Cave, 2014). Even this weak prevention power-the executive is not obliged to follow any recommendation-thus mostly remains an elite rather than a participatory mechanism.
Given that call-in is the nuclear option in city government, only activated when there are serious con-cerns of impropriety, it is necessary to consider means for citizens to object to milder infractions. One potential means to obstruct is through petition. The internet has facilitated the exponential growth of reactive petitions through a plethora of new petitioning platforms. There have even been moves to institutionalise these processes by connecting them directly to national legislatures, for instance, in the UK and Finland. Nonetheless, this has obscured the original conception of petitioning, which was as a means for those wronged by institutional power to present their claim for redress. There was no requirement that an individual must demonstrate political support with a welter of signatures. A single signature was enough (as it still is in Scotland). As such, in early democracies, petitions provided a mechanism for minorities and the unenfranchised to access lawmakers and seek redress (McKinley, 2016). Popular prevention is both underutilised and underpowered in today's urban governance. An agonistic perspective directs us towards strengthening these powers. A city-level right to petition and expanding citizen participation in call-in are two ways those who find themselves excluded by a decision could challenge it.

Oversight
Oversight is the surveillance of power to prevent its abuse. Rosanvallon (2008) gives the people as watchdogs three principal tasks: vigilance, which is control through constant and comprehensive active attention of society to institutional action; denunciation, the identification and publicising of the violation of community norms; and evaluation, analysis of institutional competence through technical assessment of quality and efficiency. His account is of a mostly mediated form of surveillance, focussing on oversight through the media, social movements, non-government organisations and independent quasi-governmental agencies. It falls short of an account of direct citizen participation in institutionalised processes of oversight. Nevertheless, oversight is the one form of counter-governance where citizen participation is proliferating.
In the UK recently there have been a wide range of initiatives to try to directly involve citizens in the oversight of urban governance and public service institutions. The creation of Overview and Scrutiny was accompanied by mechanisms that draw on citizen vigilance. Citizens are able to propose topics for scrutiny as well as submit questions and evidence to scrutiny reviews (Dearling, 2010;Stoker, Gains, Greasley, John, & Rao, 2007). Vigilance has been accompanied by opportunities to become involved in evaluation as a "co-optee" on scrutiny committees. Cooption is not common, but Bristol City Council and the London Borough of Waltham Forest retain pools of citizens to act as co-optees (Dearling, 2010).
Citizen oversight of the executive has been matched by opportunities to oversee local public services. There has been an expansion of citizens' roles in technical evaluation, through direct involvement in audit and inspection. One of the core functions of Healthwatch, which draws its name from the "watchdog" metaphor, is enabling people to hold local services to account by monitoring and reviewing provision (Local Government Association & Healthwatch, 2013). It uses various tools to discharge these functions and understand quality of performance from local people's perspective, including "enter and view" inspections, "patient-led assessments of the care environment", and "15-step challenge visits" (Gilburt, Dunn, & Foot, 2015). Likewise, the Care Quality Commission (CQC) now advertises for "expertsby-experience" to assist inspections of health and social care services.
Countries with endemic corruption problems are experimenting with citizen denunciation, creating anonymous mechanisms for informing on corrupt city officials, such as Sierra Leone's Pay No Bribe and India's I Paid a Bribe platforms. 1 UK citizens have not been encouraged to inform on city officials, though they have been encouraged to denounce neighbours fraudulently claiming social security through the creation of a National Benefit Fraud Hotline.
The problem with oversight, unlike with prevention, is not then an absence of opportunities. The dominance of the collaborative governance approach for thinking about participation, however, has meant that in practice citizen oversight is often presented as another opportunity to collaborate. CQC's call for "experts-by-experience" employs just such a collaborative governance trope. The role of the citizen as a partner with professional inspectors is emphasised, rather than the agonistic relation to those who will be inspected. Even the inspectorinspectee relationship is now cast in partially collaborative terms through the notion of the "critical friend" (see Centre for Public Scrutiny, 2005;Gilburt et al., 2015), tasked both with holding to account and working constructively to improve effectiveness. This has had some interesting effects. Local Healthwatch organisations, for instance, have struggled to reconcile these opposing functions, instead opting to act as critic or friend, but not both (Gilburt et al., 2015). Oversight is not a friendly activity, it is agonistic. The best overseers are suspicious and forensic not collaborative. Recognising the difference between agonistic and collaborative practices could help to prevent the kind of dissonance experienced by Healthwatch. It could also point to a different imperative for citizen recruitment. Collaborative governance, with its concern for mutually respectful, collective preference formation, is often wary of involving "the usual suspects" with an axe to grind. Agonists instead advise us to mobilise the passions (Mouffe, 2000a). Citizens with a legitimate grievance against an institution may not be the best candidates for partnership working, but they may prove to be the best watchdogs.

Judgement
Popular judgement concerns the capacity of citizens to constrain institutional action by testing it against community norms of governing. In ancient Athens, for instance, tribunals of citizens, selected-by-lot, could strike down the decisions of the Assembly as unconstitutional. Rosanvallon (2008) sees two trends in modern democracies that have eroded this once central function of citizenship. The pronounced functional separation between the judicial and political means that the judiciary has absorbed these responsibilities. In addition, the liberal democratic association of democracy with the act of voting has erased the separation between judgement and authorisation. The vote becomes both an ex post judgement of the incumbent government and an ex ante authorisation of the new regime, without us ever knowing which predominates. Nonetheless, citizen judgement as members of a legal trial jury remains the most common form of popular participation in the business of the state, thus it is surprising that citizen judgement has not had a more prominent place in theories of citizen participation (Dean, 2017).
The only widespread citizen judgement in local governance is a direct democratic innovation that is little remarked upon within the collaborative and participatory governance literature: recall. Recall is a feature of urban administrations in a number of countries from Poland to the Philippines (Shah & Chaudhry, 2004), and the majority of the more than 100 recall elections that occur annually in the US affect city officials, predominantly city councillors and school board members. 2 As Rosanvallon notes, recall is one form of voting that does not conflate judgement and authorisation, and is closer to an indictment than an election. Citizens solely render a verdict on the behaviour of the representative, hence act more like judges than electors (Rosanvallon, 2008, p. 209).
The popularity of deliberative mini-publics occasionally leads to their somewhat unconventional use as quasijudicial processes of adjudication on controversial issues (Dean, 2017). Two such urban governance cases that have been thoroughly documented are: the use of a citizens jury to breakthrough a deadlock resulting from local opposition to the proposed restructuring of health services in the English city of Leicester (Parkinson, 2004), and a citizens assembly launched in response to a stalemate when residents of Vancouver mobilised in opposition to a City Council neighbourhood plan that they viewed as unduly influenced by property developers (Beauvais & Warren, 2015). Nevertheless, as deliberative democratic innovations, these cases are analysed in deliberative democratic terms and their potential agonistic function is underappreciated. Despite noting their roots in local opposition to institutional action neither Parkinson nor Beauvais and Warren consider that their cases could be fulfilling a need for popular judgement. Beauvais and Warren, for instance, instead view their case as an attempt to fix a broken communication link within the local administration.
The irruption of agonistic practices is seen as a democratic threat. Parkinson argues the use of mini-publics for adjudication by randomly selected citizens is a threat to deliberative democratic norms since it excludes active citizens from processes of reflective preference transformation. An agonist might share Parkinson's concern about this exclusion of active citizens, though for a different reason. The randomly-selected citizens of a citizens jury usually also play the roles of prosecution and defence, expected to quiz expert witnesses from different angles of the debate. There is a risk that, given these citizens are specifically selected because they are not active in the agonistic confrontation, this process abstracts too far from the conflict, and thus the jury will make proposals that are unacceptable to both sides. It is notable that this tension is manifest in the practice of Beauvais and Warren's case: local partisans who had been instrumental in opposing the rejected neighbourhood plan forced the organisers to compromise on random selection for the citizens' assembly to draft the new plan. Again, the authors view this as a threat to democratic norms of inclusion. These examples indicate two potential benefits of a greater attention to an agonistic perspective. It would provide the conceptual tools to appreciate when participatory processes take on an agonistic dimension and situate such practices in competing democratic norms, rather than viewing them simply as a democratic threat. In addition, it can assist the design of more appropriate institutional innovations, so that minipublics are not used for rendering popular judgement despite question marks over their suitability for this task from deliberative and agonistic perspectives.

Conclusion
There have been few proposals in urban governance theory or practice for democratic innovations that enable citizens to contest institutional power. Citizens' powers of popular prevention and judgement are virtually nonexistence. The agonistic character of oversight has been neglected, instead miscast as another opportunity for collaboration. The notion of counter-governance, characterised as institutionalised participatory agonism, provides a lens for redressing these dysfunctions of the dominance of collaboration in conceiving of participatory urban governance. However, this article should not be read as a proposal to replace collaborative governance with a new counter-governance paradigm. Citizens and city officials will always be potential partners and potential adversaries. Collaborative governance will thus in many cases be wholly appropriate, but its claim to paradigm status as a comprehensive mode of governing occludes certain other organisational possibilities. It is far from clear that collaboration is always the most appropriate mode of interaction between citizens and city officials, particularly when their interests diverge. We should remember that historically the term also has its negative connotations: collaboration can mean working together as equal partners but also acquiescence in oppressive power.
The over-focus on collaboration has restricted in two ways the laudable inclusion agenda that underpins many efforts at participatory urban governance. The first is that there has been a proliferation of attempts to include citizens in "upstream" deliberations pre-decision (Bingham, 2006) in order to diversify influences over public decisions, however; agonists' insight that the decision itself creates exclusions, means we need to also consider unequal influence after the decision. The institutionalisation of means for ex post citizen contestation of institutional action would go some way to broadening inclusion. It could help to improve a long-standing problem of collaborative governance initiatives: that they often result in frustration for participants as competing institutional imperatives thwart implementation (Lowndes, Pratchett, & Stoker, 2001;Newman, Barnes, Sullivan, & Knops, 2004) or authorities cherry-pick proposals that fit existing agendas (Font, Smith, Galais, & Alarcon, 2017). Properly constituted forms of citizen prevention, oversight and judgement could reduce officials' discretion to side-line citizens' priorities.
Constructing almost all participation opportunities in collaborative terms forces those who want to participate into a particular relationship with the state: they must accept the state as a partner for collaboration. This may at times be inappropriate, as it neglects the violence that institutions sometimes do to citizens, particularly the poorest citizens. It is not reasonable to ask someone on the verge of being ousted from the neighbourhood where they grew-up by a gentrifying regeneration programme to accept the officials pushing the plan as partners for collaboration. Such experiences often create an energy amongst those affected to become involved to prevent government failing them and others like them. Collaborative initiatives are unlikely to prove very attractive to those who have a deep grievance. Agonistic processes that "mobilise the passions" (Mouffe, 2000a) and enable them to enter into an oppositional relation with institutions would be more likely to harness this energy to improve governance for those whose needs are often overlooked. This is the second way that expanding agonistic participation in urban governance could make it more inclusive.
This article has provided three directions for thinking about how counter-governance could expand our conception of participatory urban governance, but it is only a starting point. Institutionalising participatory agonism is unlikely to be straightforward, and requires a great deal of further theoretical development and empirical research. Prevention, oversight and judgement hold promise for constructive engagement in the face of the conflicts that often characterise contemporary urban governance, but whether they do in fact work is an empir-ical question. Participation is often proposed as a remedy for the failures of representative processes (Fung, 2006), however; counter-governance mechanisms will also create new tensions with representative modes of governing. If poorly configured, their introduction could result in a sclerotic urban governance in which it is impossible to achieve anything. Mobilising the passions, for instance, may be functional for processes of oversight, but less so for prevention and judgement if it blocks all decisions, including those made in good faith.
It must also be remembered that not all citizen resistance to institutional power is laudable. In the US, for example, there are instances of white families opposing school redistricting reforms that aim to increase socioeconomic and racial diversity (Inch et al., 2017). Understanding how to prevent elite domination of forms of counter-governance, particularly their potential for undermining legitimate decisions arrived at through representative democratic processes, will be a key question. Some ideas can be borrowed from the more mature participatory governance literatures, for example; selection of participants by sortition has been a common feature of deliberative innovations in order to prevent interest group domination and improve inclusion. Interestingly there have been proposals to appoint the aforementioned co-optees to Overview and Scrutiny Committees by lot (Centre for Public Scrutiny, 2005). New thinking will also be needed. One proposal by Lucie Laurian is that planning should have an inbuilt bias towards those with least choices, giving them a veto over plans that violate sites they hold sacred (Inch et al., 2017). This would be one way to orient forms of prevention towards inclusion.
There is also potential for expanding institutionalised participatory agonism beyond prevention, oversight and judgement. As aforementioned, this article is limited to citizen-state conflicts. Yet conflicts can be articulated along many dimensions, for instance; they can be between citizens, within institutions, or multi-faceted rather than dichotomous. As such there is a need to elaborate how agonistic processes can deal with different kinds of conflicts. How collaborative and agonistic practices may usefully be combined in order to complement each other also remains a topic for further theoretical and empirical exploration. It would seem unlikely that there will be a single model for all seasons; institutional design needs to take account of context. In circumstances of high trust and a shared vision between citizens and officials then a collaborative partnership may be possible and agonistic checks unnecessary. When distrust and conflict prevail the presence of robust mechanisms for oversight and prevention may be a prerequisite of carving out a space for collaboration. Such combinations are likely to be contingent, negotiated amongst the actors involved, still; it is useful to understand which types of design might achieve which ends. The pursuit of these questions and the development of a fuller understanding of how to institutionalise participatory agonism will furnish a more comprehensive set of tools for revi-talising urban governance to negotiate the conflicts that pervade the 21 st century city.