Expanding the Southern urban critique: Elite politics, popular politics, and self-governance in the wards of Mandalay

This article explores everyday urban governance and politics in Mandalay, Myanmar. We examine this through a focus on state-society interactions within Mandalay’s ward offices, which are the lowest tier of the administrative backbone of the Myanmar state known as the General Administration Department. This reveals the existence of three intertwined forms of urban ‘politics’ in Mandalay: elite politics, which echo the practices of civil society in the sense of Partha Chatterjee; popular politics, which echo the practices of political society; and self-governance, which is an approach to politics culturally and historically situated in Theravada Buddhism and Myanmar’s authoritarian legacies. The situatedness of the case prompts us to argue in favor of expanding the southern urban critique beyond its conventional focus on liberal democratic metropolises of the global South, in order to enrich our understanding of what constitutes postcolonial urban politics. We suggest this could be achieved, as we attempt here, by adopting collaborative research methodologies and by extensively building on southern area scholarship in ways that mediate epistemic expropriation.


Introduction
Urban Studies increasingly acknowledge that cities of the global South are not "pathological and in need of development interventions" (Schindler, 2017: 47) but rather spaces from which we have much to learn. Over the last decades, southern cities have nevertheless remained less studied than their western counterparts (Robinson, 2006;Roy, 2011;Sheppard et al., 2013;Zimmer, 2015), which has led a growing number of scholars to attempt to overcome western-colonial biases in Urban Studies (Edensor and Jayne, 2012;Lawhon et al., 2014;Robinson, 2006Robinson, , 2016aRobinson, , 2016b. Read together, their ontological and epistemic propositions now form a heterogeneous and yet cogent 'southern urban critique'. At the core of this critique are two nested arguments that are central to the framing of this article. The first is that "southern cities are socially, materially, culturally, politically and/or historically different from northern cities" (Lawhon and Truelove, 2020: 5-6;Leitner and Sheppard, 2016;Schindler, 2017). The second is that southern urban realities consequently tend to challenge the western-born concepts conventionally applied to them (Chattopadhyay, 2012;Connell, 2007;Lawhon and Truelove, 2020;Roy, 2009;Watson, 2009). In this light, the southern urban critique constitutes an invitation to better consider the "contexts and practices shaping southern cities, and the locatedness of all theorymaking" (Lawhon and Truelove, 2020: 14).
This invitation, which is the starting point for this article, has already been answered by a wealth of research engaging with everyday urban governance across the global South. Collectively, these studies have highlighted the heterogenous nature of postcolonial urban governance and the complexity of how actors, interests, practices and processes coexist and interact to shape southern cities on an everyday basis (Cornea et al., 2017a;Gabriel, 2014;Le Meur and Lund, 2001;Truelove, 2020;Truelove and Cornea, 2020). Examples of this complexity range from the intertwining of chieftaincy and formal state institutions in Malawi (Eggen, 2011), to the overlapping of formal electoral politics and everyday negotiations in Delhi (Schindler, 2014), to the interplay of multiple forms of political contestation against shortcomings in municipal service delivery in Cape Town (McFarlane and Silver, 2017). In a recent review of the field, Truelove and Cornea (2020: 3) pointed how an analytical engagement with this heterogeneity constituted an opportunity to "expose hereto unseen forms for agency and resistance, cooperation and conflict, inclusion, and exclusion". In this article, we wholly subscribe to this claim, not least because doing so has repeatedly proven to be an efficient means to reimagine urban theory from a southern perspective (Lawhon et al., 2014;Lawhon and Truelove, 2020). A prime example of this reimagining lays in the work of Partha Chatterjee, who famously reconceptualized 'politics' (Chatterjee, 2004). For Chatterjee, politics in postcolonial contexts differ significantly from western politics to the extent that they not only involve interactions between states and an organized 'civil society' made of 'citizens' -who are, for him, at best an elite minority in the global South -but also the eclectic activities of what he calls 'political society', that is, a nebula of populations, state and non-state actors central to everyday political life in the global South.
Numerous works on everyday urban governance in the global South have elaborated on Chatterjee's conceptual take on politics, notably to show how interactions between state and society often tend to blur the analytical boundaries he established (B enit-Gbaffou and Oldfield, 2011;Chatterjee, 2008;Corbridge et al., 2005;Ranganathan, 2011). For example, the works of Truelove in Delhi (2020) and of B enit-Gbaffou and Oldfield in southern Africa (2011) showed how the heterogenous ways in which urbanites made their claims heard with regard to service delivery challenged conventional definitions of 'state' and 'state power' and turned the state-society interface into a "gray zone" (Truelove, 2019).
What these works, and more generally research on urban governance in the global South, have not extensively questioned, is whether everyday postcolonial politics necessarily involve an interaction between state and society, let alone one characterized by negotiation, contestation or conflict. This oversight is possibly due to the fact that research on everyday urban governance has so far predominantly focused on metropolises of the global South located in neoliberal democratic polities (Edensor and Jayne, 2012;Parnell and Robinson, 2012;Robinson, 2006;Ong, 2011), particularly South Africa (B enit-Gbaffou andOldfield, 2011;Lawhon et al., 2014Lawhon et al., , 2016Lemanski, 2014;Parnell and Robinson, 2012) and India (Chattopadhyay, 2012;Narayanan, 2019;Roy, 2011;Truelove, 2020;Zimmer, 2015). This spatial focus has portrayed 'cities of the global South' as spaces in which politics are primarily unfolding under relatively 'open' conditions (Koch, 2013) and thus often take the form of explicit contestation. However, an important conclusion of this literature is that governance in southern cities is shaped not only by structural forces, such as colonial legacies and global neoliberalism (Chatterjee, 2004;Roy and Ong, 2011), but also through the mediation of these forces by local agency and historically and culturally situated practices (Pihljak et al., 2019;Roy and Ong, 2011;Truelove and Cornea, 2020). Taking this conclusion seriously, our contribution to the conversation intends to show how everyday agency is exercised in the partially 'closed' context (Koch, 2013) of postcolonial Myanmar. 1 In so doing, we wish to emphasize the generally overlooked role played by forces beyond neoliberalism and colonial encounters -in our case religion (Lanz, 2014) and authoritarian legacies.
To elaborate our proposition, we turn our attention to everyday governance in Mandalay, the last royal capital of precolonial Burma. Today, Mandalay is seen as the country's cultural heartland and a stronghold to be defended against foreign (particularly, Chinese and western) encroachment (Mya Maung, 1994;Roberts, 2017Roberts, , 2018Sanchez, 2019). Our focus is on the ward offices of Myanmar's General Administration Department (GAD), the lowest tier of the country's pervasive administrative apparatus. Taking a cue from discussions of postcolonial politics, we approach ward offices as a 'gray zone': a political assemblage defying analytical dualisms, characterized by complex regimes of legitimacy, and in which power differences are embedded (Truelove, 2019). This allows us to grasp the complexity and heterogeneity (Truelove and Cornea, 2020) of ward-level institutions, actors and practices, which are central to everyday life and its politics in the city (Arnold and Saw, 2014;Kempel and Tun, 2016;UNDP, 2015b).
Our contribution shows that the Mandalay case strengthens the southern urban critique's arguments as it shows that urban politics in Myanmar echo that of many southern cities. Everyday governance in Mandalay is indeed often characterized by civil society-like, political society-like, or hybrid and overlapping conceptions and practices of politics (B enit-Gbaffou and Oldfield, 2011;Chatterjee, 2004;Corbridge et al., 2005;Ranganathan, 2011). However, the case also shows that our understanding of urban politics in the global South could be fruitfully expanded to include forms of politics that are not rooted in the colonial or neoliberal encounters (Cornea et al., 2017b;Parnell and Robinson, 2012), but in conceptions and practices of politics culturally and historically situated elsewhere. In Mandalay, everyday politics are largely influenced by Theravada Buddhism and the country's precolonial and postcolonial authoritarian experiences (Maung Maung Gyi, 1983;Thawnghmung, 2011;Walton, 2016).
The remainder of the article is structured as follow. First, we briefly trace the genealogy of ward governance and describe contemporary institutions, actors and processes, including ward elections. Second, we define and discuss three conceptions and practices of politics in this gray zone. We start with the normalized, elite practices of 'civil society' before moving to the eclectic practices of 'political society' (Chatterjee, 2004;McCarthy, 2016;Prasse-Freeman, 2016), to conclude on what we read as situated politics of merit and selfgovernance (Hsu, 2019;McCarthy, 2019;Prasse-Freeman, 2012). Before delving into the case, however, we describe our methodology.

Methodology
This article is the product of a collaboration between a western academic and a (former) research assistant who was not 'silenced' (Turner, 2013). As such, it constitutes a genuine attempt at mediating 'epistemic expropriation' (Halvorsen, 2018) and at theorizing across western and southern perspectives (Lemanski, 2014). This attempt is of course incomplete and we acknowledge that our contribution remains ethnocentric in its epistemology and its methodology.
The article was elaborated in the context of the first author's doctoral research, which explored urban sanitation in the city of Mandalay. For this research, over 120 in-depth qualitative interviews were conducted by the two authors in Myanmar between May 2016 and December 2018 with governmental, civil, private and international actors. This article draws material from a series of interviews conducted in Mandalay with the GAD at the district, township and ward levels. These interviews were led in Burmese by the second author, who also translated questions and answers to the first. The interviews were recorded by the two authors separately, in the form of field notes, respectively in Burmese and English. The notes were lengthily debriefed and aggregated immediately after interviews took place.
Our respondent sample included at least one representative of each actor type involved in ward governance. District and township officers of the GAD, ward administrators, household representatives, clerks, elders, and local volunteers were thus interviewed separately or in groups. The Mandalay district office, all six township offices, and twenty-seven of the ninety-six ward offices were visited and accepted to provide information. Reluctance to answer was sporadically encountered although most respondents engaged with remarkable hospitality after having been assured that their anonymity would be protected, and after having received clearance letters from the GAD. 2 The ward offices we visited were selected based on their location, with the aim of establishing a representative pattern across the city. Maximal diversity was achieved in socio-economic and morphological structure thanks to triangulation of geospatial data, preliminary field observations, and informal conversations with local residents. A certain consistency in results could be observed, hinting at a relative homogeneity in everyday political practices throughout the city. 3

Ward governance in Mandalay: A gray zone
In this article, we approach ward governance as a 'gray zone' (Thawnghmung, 2011;Truelove, 2019). At the core of this gray zone are the ward offices, institutions that stand on the threshold between state and society. They embody a form of 'political society' (Chatterjee, 2004) in the sense that they are "constructively engaged in providing links between 'government' and 'the public'" (Corbridge et al., 2005: 191).
Like political society elsewhere, and like most political processes in contemporary Myanmar, ward governance is influenced by historical legacies (Maung Maung Gyi, 1983;Thant Myint, 2019). In precolonial Burma, local headmen (thu gyi) were relatively autonomous authorities whose political legitimacy was bound to their moral (Theravada Buddhist) conduct (Walton, 2016). Their appointment was decided by the King, and in many cases hereditary. Headmen were tasked with socio-religious, criminal, and fiscal administration, and often supported by locally appointed elders (Kyan, 1969;Furnivall, 1958;Kempel and Tun, 2016;Maung Maung Gyi, 1983). The nomenclatures and practices characterizing the thu gyi system were diverse, place-dependent (Kyan, 1969), and not legally enshrined (Furnivall, 1958;Maung Maung Gyi, 1983). Importantly, however, the thu gyi system was the only scale at which a degree of popular political participation could occur, and thu gyi were the only officials considered to actually represent the interests of the population (Maung Maung Gyi, 1983). In precolonial Burmese political rationality, the government was indeed regarded as alien to society and governmental politics were not considered to be the laypersons' concern. Above all, engaging with state affairs (nain ngan ye) was regarded as inappropriate (Maung Maung Gyi, 1983).
Following colonization during the 19th century, the thu gyi system was 'governmentalized' (Foucault, 2004b) through the introduction of formal legal mechanisms (Maung Htin Aung, 1962;Maung Maung Gyi, 1983). Thu gyi and elders were gradually coopted into the colonial apparatus in order to normalize and facilitate everyday administration (Kyan, 1969;Kempel and Tun, 2016), but also in order to suppress uprisings (Kyan, 1969;Furnivall, 1958). Governmentalization also led to the creation of wards as administrative subdivisions. This rendered 'legible' (Scott, 1998) spaces that until then had only existed as informal social units (Furnivall, 1958). In 1907, the Towns Act gave a unified legal framework to ward governance and institutionalized the functions of thu gyi. The Act bestowed upon them 'biopolitical' duties (Foucault, 2004a), including local sanitation and census, in addition of their traditional 'disciplinary' functions such as law enforcement or population surveillance (Foucault, 2004b;Kempel and Tun, 2016). The Act also formalized the supportive function of elders in civil and criminal administration (Furnivall, 1958). Until 1907, thu gyi and elders were appointed by Deputy Commissioners, district-level European officers who were subsequently encouraged to consult with the population when deciding on appointments (Arnold and Saw, 2014;Furnivall, 1958;Saha, 2013). In 1908, the Secretariat Office Act took local governmental autonomy away by placing thu gyi under the authority of the centralized Secretariat Office, the administrative backbone of the colonial apparatus. This achieved to turn thu gyi into instruments of state rather than representatives of society (Arnold and Saw, 2014;Furnivall, 1958).
After Independence in 1947, ward level governance scantly evolved (Arnold and Saw, 2014). In 1953, a legal reform attempted to transfer the functions of thu gyi to locally elected councils, in the aim of 'democratizing' ward governance and bringing the state closer to society. Furnivall (1958: 84) wrote about this reform that it "should minimize, though it does not remove, the main obstacle to efficiency in local self-government in a country like Burma, where the people do not want things [like popular participation] that in the modern world they need and ought to want". The reform was never successfully implemented (Kempel and Tun, 2016).
Following General Ne Win's coup in 1962, Burma entered five decades of military rule and the importance of thu gyi and elders faded. Their duties were transferred to Security Committees appointed by the military and supported by civilian personnel from the GAD, the postcolonial successor to the Secretariat Office. The GAD was then a division of the military-headed Ministry of Home Affairs (Arnold and Saw, 2014;Kempel and Tun, 2016). After 1988, thu gyi and elders started to earn back their responsibilities under continued supervision of the GAD while military personnel gradually distanced themselves from local civilian affairs (Kyed et al., 2016). Thu gyi were then directly appointed by GAD personnel, who could consult local populations (Kempel and Tun, 2016;Kyed et al., 2016). Several respondents in Mandalay pointed that the GAD's influence had diminished after 2011, and even more after 2016 when Daw Aung San Suu Kyi's National League for Democracy (NLD) government took office. In 2018, the GAD was transferred from the Ministry of Home Affairs to the civilian-headed Ministry of the Office of the Union Government, marking the end of military control over ward governance and its passing unto NLD oversight.
Nonetheless, the current legal framework for ward governance dates back to 2012, when the Ward and Village Tract Administration Law was enacted by the government of President U Thein Sein, controlled by the military-backed Union Solidarity and Development Party (USDP).
This legal framework can be read as an attempt at further governmentalizing and 'democratizing' the thu gyi system. Governmentalization is semantically enshrined in the Law, where thu gyi were renamed 'ward administrators' (yet kwet ouq chouq ye hmu 4 ). It is further distilled throughout the framework, which was designed for the state to "secure legitimacy not by the participation of citizens in matters of state but by claiming to provide for the wellbeing of the population", as is typical of governmentalized regimes (Chatterjee, 2004: 34).
Because of this design, the 'democratic' element of the legal framework should be understood in Burmese rather than western terms, according to which 'democratic politics' are not antagonistic and do not imply universal participation in nain ngan ye or 'state affairs' (Lall, 2016;Lall and Hla Hla Win, 2012;Maung Maung Gyi, 1983;Walton, 2016). In Chatterjee's words, the goal of the Law is thus not to turn ward governance into a space for 'civil society' where 'citizens' could exercise universal voting rights and participate in popular sovereignty (Chatterjee, 2004: 27-34). Rather, it constitutes an attempt at increasing the state's political legitimacy by reinforcing the role of its local representatives in everyday social welfare provision (Hsu, 2019;McCarthy, 2016). So far, however, this attempt is incomplete and everyday ward governance remains a gray zone in which everyday practices defy formal institutions. The persisting grayness of everyday governance in Mandalay is particularly observable in the distribution of functions among the actors involved, and in ward election processes.
Ward offices in Mandalay typically host a nebula of actors fulfilling diverse roles and functions in everyday urban governance. At the center of this nebula are the ward administrators, the ward clerks, and the 10-household and 100-household representatives.
Under the current Law, ward administrators are both part of society and state representatives. On the one hand, they are elected by the population and do not have the status of civil servants. On the other, they are accountable to the GAD for engaging in practices such as accepting bribes, peddling influence, or using their position to establish patronage relationships. They also receive subsidies from the GAD that aim to discourage such practices (Kyed et al., 2016). However, according to several respondents in Mandalay, patronage mechanisms remain widespread.
On an everyday basis, ward administrators carry out multiple functions that have both disciplinary and biopolitical dimensions (Foucault, 2004a(Foucault, , 2004b. Besides, as the legal framework remains vague as to how their duties are to be exercised, interpretative behaviors are the norm (Kyed et al., 2016). Ward administrators in Mandalay reported focusing primarily on the maintenance of ward security and order. Many mentioned the mediation of neighborhood disputes and the handling of petty criminality as their principal activities. In a few wards, security was understood in a broader sense with administrators setting up patrols of 'volunteers' to prevent alcohol or drug related nuisances. In other cases, security was enforced in relation to encroachment, with administrators alternatively choosing to defend, tolerate or impede encroaching practices. One administrator for instance mentioned that he was requesting landlords to compensate squatters in cases of disputed evictions. Another explained how he was brokering deals with the police to avoid expulsions. Several reported on the contrary assisting the police with evictions.
Parallel to these activities, ward administrators explained taking part in all 'local events' in their constituency, giving examples ranging from disasters to ceremonies or state health campaigns and public construction works. In particular, they explained getting systematically involved in everyday administrative procedures for their constituents, certifying for example birth and death certificates, household registration lists, job applications and university enrollment files. Because of this, ward administrators generally have extensive knowledge of (and power over) their constituency. This also implies that ward administrators have pervasive population control means, exposing residents to 'systematic bureaucratic arbitrariness' (Gupta, 2012: 6-14).
Finally, ward administrators perform duties that relate to local infrastructural development. For this, they primarily make use of the constituency development funds introduced by the National Parliament in 2013 (Robertson et al., 2015). These funds amount to 100 million Kyats (about 70,000 USD) per township per fiscal year and are distributed among wards to fund small-scale infrastructures. Ward administrators are in charge of identifying needs within their constituency, applying for funds, and managing them once received. Most administrators in Mandalay mentioned relying upon local communities, typically committees composed of local elders, to do so. Furthermore, although allocation of funds is legally at the discretion of Members of Parliament (Robertson et al., 2015), ward administrators reported that decisions were made collegially by township-level assemblies composed of the local Member of Parliament, GAD personnel, and all ward administrators of the township.
In their everyday duties, ward administrators are supported by clerks who act as the secretaries of ward offices. Unlike ward administrators, clerks are directly appointed by the GAD, to which they are also affiliated and accountable. They are civil servants, and as such can be sanctioned or transferred arbitrarily. 5 Because of this, clerks are considered to be loyal to the state. In practice, however, several clerks pointed out that their autonomy was increasing.
Ward administrators are further supported by population representatives. The formal position of 'block elder', created in 1907, was suppressed by the 2012 Law without much effect: respected residents continued to informally provide support to ward administrators and to get involved in local elections. Their functions were re-institutionalized in 2016 by the Law's third amendment, which created the positions of 10-household and 100-household representatives. Unlike these designations suggest however, 10/100-household representatives in Mandalay are rarely in charge of exactly this many households. Rather, they tend to represent a street or a residential block. The formal duties of 10/100-household representatives include the handling of ward elections and 'support' to elected administrators. In practice, representatives focus on ensuring communication between administrators and the population. For example, they are often in charge of collecting data or disseminating instructions within their household clusters. They also typically maintain permanence in ward offices, sometimes according to planned schedules, and tend to be organized hierarchically, the eldest for example substituting the administrator in their absence.
Finally, ward elections processes throughout Myanmar are highly diverse, arguably because the GAD is now exercising limited control over them (Kyed et al., 2016). In Mandalay, ward elections vary significantly from one office to the next and generally reflect this distancing. In other words, ward elections are not a normalized exercise of citizenship (Chatterjee, 2004). In many cases, the election of ward administrators is in fact rather a selection process: participants tend to choose consensually among candidates and only rarely cast secret ballots. This hints at culturally situated conceptions and practices of politics in the gray zone that do not entirely resemble those of civil and political society.

Politics in the gray zone
Conceptions and practices of 'politics' in the gray zone are in fact layered, and these layers are intertwined. The first two layers, which we call elite politics and popular politics, respectively echo the practices of civil and political society elsewhere in the global South. The third layer, which we call merit politics, is more specific to Myanmar. Together, all three layers constitute a hybrid continuum of practices rather than bounded and exclusionary spaces (B enit-Gbaffou and Oldfield, 2011;Narayanan, 2019;Ranganathan, 2011). In this section, however, we zoom into each one of them discretely for the sake of clarity.
Elite politics, civil society, and nain ngan ye Following Chatterjee (2004: 34), democratic politics in a Western sense unfold within civil society, a space constituted of an elite minority of organized citizens who engage with the state through law-abiding practices. There is a word in Burmese for such 'capital P' politics: nain ngan ye, "a state-centric, elite-level practice", with which most Burmese do not engage, and which thus poorly reflects everyday political practices (Maung Maung Gyi, 1983;Prasse-Freeman, 2016;Walton, 2016: 65). By extension, nain ngan ye can refer to national party politics and, in the wake of military rule, this form of political action is largely considered a dangerous activity rather than a popular way of engaging the state 6 (Hsu, 2019;McCarthy, 2016;Walton, 2016). Against this background, the actors of ward governance routinely engage in myriads of mundane political performances that reveal ambivalent and 'gray' relationships to nain ngan ye. In so doing, they alternatively become elite citizens, or on the contrary actively distance themselves from 'civil society'.
Ward administrators routinely engage with nain ngan ye on behalf of their constituents in order to promote social welfare, in a context of lack and fear of state service delivery (Kempel and Nyien, 2014). In so doing, they become elite citizens who act as intermediaries between the population and state institutions such as the Mandalay municipality or governmental ministries. In the city, ward administrators often build on their privileged understanding of these institutions to relay the claims of their constituents, for example relating to garbage accumulation, drain clogging or water supply interruptions to the relevant departments. Alternatively, they can also formally monitor state agents, for example asking municipal waste truck workers to systematically sign passage sheets.
The other way around, ward administrators can also bring nain ngan ye to the population. For example, they regularly assist the Ministry of Immigration or the Ministry of Transport in the distribution of official identification documents, such as driving licenses. At the time of research, many administrators were further participating in a public health campaign conducted by the municipality and the Ministry of Health to promote safe water, sanitation and hygiene practices. Several administrators mentioned that their role in this campaign was to identify areas where intervention was needed, and to reassure their constituents about state intentions -particularly within encroaching settlements. While this clearly shows a persisting rift between nain ngan ye and the population, and the bridging role of administrators, it is worth noting that trust and expectations in the state's ability to deliver social welfare seem to be growing since 2012. In Mandalay, many administrators and household representatives insisted that municipal service delivery, and state agents more generally, were increasingly trustworthy and concerned with popular welfare (see Sanchez, 2020).
Despite this change, ward clerks sometimes continue to actively distance themselves from their identity of nain ngan ye representatives. In effect, they continue to operate under rigorous state supervision, regularly receiving orders from and report back to GAD line offices. Further, like all civil servants, they must wear uniforms in the (military green) colors of the GAD. However, in contrast with their line officers, very few of the ward clerks encountered were actually wearing these uniforms. Instead, several wore jackets labeled with their ward's name, an outfit sometimes fashioned by ward administrators to create a collective identity for their offices. This choice of attire can be interpreted as a political statement and an attempt to demonstrate the clerks' actual sense of belonging (Egreteau, 2019). Indeed, ward administrators expressed continued disaffection towards the GAD, for instance mentioning that township-level administrators were generally unknowledgeable of local challenges, because often transferred, or were disinterested listeners.
As far as ward elections are concerned, national party politics are traditionally held at a distance. First, they are legally not supposed to unfold at this level, as administrators are required not to be affiliated to political parties in order to be elected. Second, in any case, party politics are still largely seen as disconnected from local electoral practices, which are rather an administrative exercise (Kyed et al., 2016). For example, in December 2017, rumors circulated in Mandalay that extra-ordinary ward elections would be soon held. Once the rumors were confirmed, speculations went that Daw Aung San Suu Kyi's recently elected NLD government was trying to install NLD-backed administrators and sweep the ones elected under President U Thein Sein's USDP government. However, most interviewees displayed resigned indifference at this possible unfolding of nain ngan ye in their lives. Discourses such as "we are base level, we don't know the agenda of the national level" or "whatever the government says, we will implement it" were common.
One of the consequences of this active distancing from nain ngan ye is that 'popular sovereignty' remains a distant prospect in ward elections (Chatterjee, 2004;Kempel and Nyien, 2014). In fact, indirect election or vote-less selection continue to be widespread despite recent reforms incentivizing more direct practices. Arguably, for ward elections, consensual appointment or entrusting decision-making to respected elders remain preferred practices (Kempel and Nyien, 2014).
Nevertheless, it seems that party politics are today playing a growing part in ward governance. Ward officeholders notably reported that the NLD and the USDP were increasingly getting involved in their daily activities, endorsing administrators or on the contrary trying to replace some with party supporters. Furthermore, a few ward officeholders admitted being involved in nain ngan ye. For example, one administrator stated that he remained an NLD member "by heart" although he had formally surrendered his party membership to comply with legal requirements; a ward clerk boasted pictures of her family wearing the NLD's trademark jackets (Egreteau, 2019); and the NLD's peacocked flag or portraits of Daw Aung San Suu Kyi decorated several ward offices.
These performances, while clearly revealing the penetration of elite politics into everyday governance, can also be read as displays of support to the persona of Daw Aung San Suu Kyi, which in turn reveals the persistence of 'personality politics' in Myanmar (Lall, 2016: 107). Personality politics are a precolonial feature of (Theravada Buddhist) Burmese political culture and can be understood as resulting from the 'law of status', a belief that power (aanaa) is not attached to contractual laws or institutions. Rather, it manifests in an individual's social status, often as a result of their moral conduct and meritorious actions, or hpon (Maung Maung Gyi, 1983: 174;Walton, 2016). This sustains a patriarchal notion of power that fosters inequalities, particularly because hpon is widely perceived to be an inherently male attribute (Aye Nwe, 2010;Miedema et al., 2016).
Personality politics thus often imply the iconization of certain authoritative elites (lu gyi), most prominently Daw Aung San Suu Kyi (Z€ ollner, 2012), but at a more mundane level ward administrators or elders who can themselves become regarded as lu gyi. 7 This iconization effectively entrenches the perceived rift between the political agency of elite actors and that of laypersons (Prasse-Freeman, 2016). This not only impedes the emergence of 'civil society' but also creates space for 'lowercase p' politics -typically, patronage relationships between elites and laypersons -that are characteristic of political society in the sense of Chatterjee.
Popular politics, political society, and nar lai hmu For Chatterjee, political society represents state-society interactions "as [they] actually happen" (2004: 41) in postcolonial contexts. On the ground, political society is composed of an eclectic nebula of populations, rather than of organized citizens. These populations engage with the state through myriads of political practices that are rarely institutionalized, sometimes not legal, and yet not always un-democratic (Maung Maung Gyi, 1983). In the wards of Mandalay, such practices often continue to prevail over 'capital P' politics, and shape both everyday governance and local election politics.
Personality politics, in the sense described above, continue to play a much more critical role in ward elections than party or elite politics. In other words, personal reputation remains the main electoral tool of administrators rather than political campaigning. As one respondent put it: "in this area there are no politics [nain ngan ye]. People don't look at politics, just at the person." Another administrator further explained how, during the last ward elections, the NLD had failed to install their supporter in the ward, only succeeding in mobilizing the residents against that candidate and having them vote massively for himself instead. According to the administrator, this happened because he himself was not a party member (engaged in nain ngan ye), and because he had been in charge of the ward previously. This conferred him a greater understanding of local needs than the parachuted NLD candidate. Very often, ward administrators are in fact local personalities whose legitimacy is primarily secured by an outstanding moral reputation according to local socio-religious standards (Chambers and Cheesman, 2019;Walton, 2016). It is also common to encounter administrator dynasties having survived the military era, especially among locally influential business families who tend to be perceived as less corrupt and better representatives of society than state agents (Kyed et al., 2016).
On an everyday basis, personality politics further lead to the establishment of patronage relationships between ward administrators and their constituents. These relationships make the daily bread of Mandalay's political society and often become opportunities for both residents and ward officeholders to practice 'lowercase p' politics while engaging with the state (B enit-Gbaffou, 2011). The narrative of one local businessman-turned-administrator illustrates the importance of patronage to ward governance. This administrator used to provide bamboo housing materials to low-income residents, particularly encroachers. In so doing, he acquired a good reputation that led to his appointment as ward headman in 2006. After that, he started using his status and influence to broker deals between encroachers and the police, offering to mitigate forced evictions to the former and petty criminality to the latter. He was then systematically reelected without serious competition.
The deals referred to in this narrative have a name in Burmese: nar lai hmu, "informal and tacit agreements struck with authorities [. . .] to overcome constraints, whether natural or institutional, in order to [. . .] fulfill individual and collective needs" (Thawnghmung, 2011: 646). Such deals are typically built through long-term, culturally acceptable social interactions, which means that they are often indirect, non-imposing, and non-confrontational (Roberts, 2015). Nar lai hmu (literally "understanding") deals are however never formally institutionalized or made legible to outsiders, and can thus be understood as political society-like practices that make access to everyday state welfare uncertain (Roberts, 2015;Thawnghmung, 2011).
In Mandalay, nar lai hmu, brokering and other forms of patronage are yet relied upon on a daily basis by ward officeholders. Many respondents mentioned that residents often requested them to intervene in their name, typically with the municipality, not only because of their enhanced understanding of nain ngan ye as described previously, but also because the deals they could establish thanks to their status ensured better service delivery. One administrator thus explained how he had secured the goodwill and preferential service of municipal drain sweepers by offering them meals whenever they conducted works in his ward. Another kept at hand the phone number of a municipal truck driver who accepted to empty septic tanks on short notice and outside of official schedules, provided that clients agreed to pay an extra fee. Yet another administrator explained how he had reconciled two households arguing over an overflowing septic tank. He helped the household whose tank was faulty contacting the municipal desludging service, but they failed to meet their costs. The administrator thus ordered them to save until they could afford the service, and in the meantime instructed their neighbors to endure the inconvenience (so as not to become imposing).
Such political society-like practices, and ward governance institutions more generally, are increasingly criticized by Burmese non-governmental organizations, who challenge not only the obvious limits of these institutions in incentivizing equal rights democracy, but also the legitimacy mechanisms they sustain (Action Committee for Democracy Development & Progressive Voice, 2018). This, in turn, hints at the emergence of a civil society concerned with nain ngan ye and right-based approaches, despite their inherent political risks and cultural limits (Maung Maung Gyi, 1983;Lall, 2016;Prasse-Freeman, 2012).
Nonetheless, several administrators mentioned that their ability to secure social welfare and state service delivery through nar lai hmu remained in fact a factor of their own political legitimacy (Hsu, 2019;McCarthy, 2016), and sometimes an incentive for their constituents to get involved in local affairs, including ward elections. In that sense, nar lai hmu constitutes a paradoxical space of power inequalities, hope, and actual welfare delivery at the core of Mandalay's political society (B enit-Gbaffou and Oldfield, 2011;Buire, 2011;Roberts, 2015;Rubin, 2011). Besides, nar lai hmu very often intertwines with lu hmu ye, another form of direct social welfare delivery by ward officeholders that avoids direct state engagement (Hsu, 2019;McCarthy, 2016). Because of this avoidance, lu hmu ye is a practice sometimes considered "apolitical" by those who engage in it. It is however as decisive to the distribution of everyday welfare in Mandalay as elite or popular politics are, and eminently political in nature although in a culturally singular and non-confrontational way (Hsu, 2019;Prasse-Freeman, 2012Thawnghmung, 2011).

Self-governance, merit politics, and lu hmu ye
Chatterjee pointed at the existence of population groups existing beyond political society, but argued that they were so located because they were "unable to gain access to [its] mechanisms" (Chatterjee, 2008: 61, emphasis added). In Myanmar, we would rather argue that populations have long been uninterested in such mechanisms (Maung Maung Gyi, 1983;Thawnghmung, 2011). From precolonial to post-junta times, everyday governance has indeed been an essentially "self-governing space" in which state service delivery is simply not expected (Thawnghmung, 2011: 646;McCarthy, 2016).
Instead, state services are often substituted with lu hmu ye, a form of social work tied to (Theravada Buddhist) notions of merit, deservingness, and reciprocity (Hsu, 2019;McCarthy, 2019). The notion of parahita, which refers to the welfare of others, is particularly central to lu hmu ye (Walton, 2016). However, lu hmu ye practices do not only serve immediate welfare purposes, but also have indirect effects. On the one hand, they can be performed as a socio-religious duty, a "moral and meritorious deed" that builds towards one's ultimate liberation from the cycle of reincarnations (Hsu, 2019: 19). On the other, they can constitute an equally moral but more secular way of becoming 'deserving' of material welfare delivery, typically from the state (McCarthy, 2019). As a form of popular claimmaking, the latter mechanism in particular can be read as a practice located beyond both civil and political society.
In fact, whether lu hmu ye constitutes or not a political act is debated: it is apolitical to the extent that it certainly differs from elite politics (nain ngan ye) and does not involve negotiation (including nar lai hmu) or confrontation, but in another sense it is concerned with the distribution of parahita/public welfare (Hsu, 2019: 19-22) or its expansion through indirect state involvement (McCarthy, 2019). Those are increasingly regarded as political activities in Myanmar (Walton, 2016). Further, the Burmese expression lu hmu ye contains a relational notion of exchange and reciprocity that further makes the practice inherently 'lowercase p' political (McCarthy, 2016).
In the wards of Mandalay, lu hmu ye is partly institutionalized in the form of communitybased organizations (CBOs), mutual-help associations, or volunteer groups (Hsu, 2019;McCarthy, 2016). Almost every ward has at least one such organization, whose purposes range from the handling of socio-religious ceremonies such as funerals or donations to monasteries, to various welfare activities. The latter can include ambulance and/or funeral services, the establishment of trust funds for flood victims or encroachers, the provision of community water pumps, and the performance of minor road maintenance activities. Such CBOs are usually funded through donations made by ward residents, and very often headed by local lu gyi, typically administrators or 10/100-household representatives. In Mandalay, some ward offices have become lu hmu ye organizations themselves. These offices have formalized committees of volunteers to handle, for instance, security patrols, blood donation campaigns, or public health promotion. In one ward that had no CBO, the administrator reported that his office was in fact handling all usual lu hmu ye activities with the help of rotating groups of young volunteers. In another ward, a 100-household representative explained how both the elected-appointed and volunteer members of the ward office were handling the dredging of local drains to cope with the lack of municipal service. They mentioned gathering on a weekly basis in a given household cluster to clean drains, calling upon neighbors for help, and eventually calling upon the municipality to carry the sludge away -having become deserving of state support in the process of 'doing [the cleaning] themselves' (McCarthy, 2019).
With regard to the handling of constituency development funds in particular, several administrators mentioned that since state subsidies were generally insufficient to perform the planned infrastructural works, they often invited their constituents to engage in lu hmu ye by contributing either financially or with time and labor. Administrators mentioned that residents were mostly enthusiastic to commit, especially where local infrastructures where lacking the most, although some were reluctant to the idea of doing something directly 'with the state' and would rather have the funds managed entirely locally. A similar sentiment sometimes defined local attitudes towards petty criminality and conflicts within the ward: administrators mentioned calling upon the police only in instances of the most serious criminal cases, and otherwise reported handling situations locally and informally, resorting to other state institutions as little as possible in yet another display of distancing from nain ngan ye (Kyed, 2018;Than Pale, 2018).
In general, because the delivery of parahita through lu hmu ye remains tied to ideas of merit and deservingness, it sustains a "regressive notion of entitlements where the poor bear the costs" (McCarthy, 2019: 328) and where power differences can only flourish. In particular, deservingness can become a powerful tool against the most vulnerable populations (Chambers and Cheesman, 2019), and especially against encroachers who are legally excluded from participating in ward elections and, as a result, from ward governance more generally. This exclusion reinforces their exposure to bureaucratic arbitrariness (Gupta, 2012) and their general insecurity under conditions of nar lai hmu politics (Roberts, 2015).

Conclusion
In this article, we explored everyday ward governance in Mandalay as a gray zone (Thawnghmung, 2011;Truelove, 2019) constituted of complex institutional arrangements, practices, and regimes of legitimacy. We showed how the ward offices of the General Administration Department, the central actors of this gray zone, embody a form of political society linking the state to society in Mandalay. In so doing, they distribute everyday welfare and exercise power over the population in ways that are both informal and governmentalized, disciplinary and biopolitical, legal and non-legal, and democratic and non-democratic. The actors of ward governance further engage in three forms of 'politics', often in intertwining or ambivalent ways. The first two forms, elite politics and popular politics, respectively echo the practices of civil and political society elsewhere in the global South (Chatterjee, 2004). The third form, the politics of merit and self-governance, is culturally rooted in Theravada Buddhism and Myanmar's pre-and post-colonial encounters with authoritarianism.
As a whole, the Mandalay case strengthens and expands the southern urban critique in a twofold manner. First, it affirms its relevance as an epistemic community as it shows that urban politics 'in the global South' are indeed heuristically comparable (Lawhon and Truelove, 2020;Robinson, 2015Robinson, , 2016a beyond democratic contexts, and beyond the 'spaces of neoliberalization' that are southern metropolises (Brenner and Theodore, 2002;Roy and Ong, 2011). Second, however, the case also shows that urban politics in Myanmar are not only rooted in the neoliberal and the colonial encounters. This points to the need of expanding the southern urban critique, geographically and conceptually, to include other conceptions and practices of politics that are historically and culturally situated, notably in religion (Lanz, 2014). We believe this could notably be achieved, as we have attempted here, by better including the theoretical and empirical perspectives of southern and area scholarship (Lawhon and Truelove, 2020), and by conducting research collaboratively in ways that mediate epistemic expropriation (Halvorsen, 2018).