Political Informality: Deals, Trust Networks, and the Negotiation of Value in the Urban Realm

ABSTRACT This article develops a conceptual framework for analysing political informality, before going on to explore it through a discussion of urban politics in Kampala, Uganda. It builds on recent theoretical developments concerning the relationship between ‘deals’ and ‘rules’, the role of informal trust networks in politics, and different conceptualisations of patron-client relations, to consider varying forms of informal political negotiation and bargaining. Informal politics is ubiquitous, but varies significantly in terms of its interface with formal institutions and the degree to which it is shaped by shared norms and expectations. The article draws out some of these variations through a typology setting out four partially-overlapping categories that aim to capture the diverse and dynamic nature of informal political interaction: pro-formal, anti-formal, para-formal, and a-formal. It then turns to the application of these in an urban context through an analytical discussion of marketplace politics in Kampala. The article argues that conceptualising political informality in this way can facilitate granular and comparative analyses of urban political processes often just described as ‘messy’, ‘chaotic’, or ‘fluid’.


Introduction
This is a capital city, we are fighting with so many interests! A minister says this, or the President says this. So this causes [government] most of the time to break down […]. At the end of the day also we have our interests, because we all are fishing in the same lake. 1 The above quote from the former Mayor of Kampala is strangely counterintuitive. While it should come as no surprise that a capital city involves many conflicting interestsand no great revelation to anyone who knows Kampala that political elites view the city as a 'lake' full of depleting resources to be grabbedbetween these two observations comes something more peculiar. 'A minister' and 'the President' are both part of the government. So why, in the brief narrative provided by the Mayor, should them saying things cause the government to 'break down'? The reason is that these hypothetical utterances from key political figures are not part of the official political process, not within the politicians' official mandates, and often consciously intended to subvert formal institutions for political dialogue. They are forms of political informality. There are many other varieties of informal politics, too; yet this kind, which involves the weakening of government itself through the sporadic invocation of informal top-down political linkages, may have had more impact on Kampala than anything associated with the organisation of political parties, patterns of voting behaviour, or the other staples of mainstream political science.
In academic debates on informality, the political element has often only been implicit within a literature concerned primarily with economic and spatial dimensions. Informal politics has long been an object of interest in political science, but for the most part has been debated and

Specifying the informal political
Informality as a generic concept has been defined in innumerable ways over the past five decades. Here I follow Lindell in classing as informal those activities that in some respects 'lie beyond or circumvent state regulation' (Lindell, 2010, p. 5). This accepts that few actors fully evade any form of regulation or taxation by the state, although the essence of informality is that their integration into formal state regulatory systems is at most partial. Acknowledging early criticisms of Hart's concept as overly rigid and dualistic (Bromley, 1978;Roitman, 1990), this approach also alludes to a continuum of formality-informality rather than a clear dichotomy.
How might such a definition apply to politics? We are here talking about the organisation of political interests and bargaining within (and between) societies, including through elections, political organisations, demonstrations, local community mobilisation, and practices of government (national or local) in general. This distinguishes the focus of this article from informal activities and practices that could be considered political in a broader sense, exemplified by Bayat's idea of the 'quiet encroachment of the ordinary' as a political act (Bayat, 1997, p. 57). Like economic activity or physical development in space, political activity is subject to contextually-specific rules and regulations concerning issues such as: the establishment, activity, and funding of political parties; who can participate in elections or stand for political office; political lobbying and the permitted activities of organisations such as trade unions and community organisations; and conflicts of interest for people involved in public life. Politics thus has its own regulatory framework in a given society that provides official legitimacy to certain processes and organisations.
However, political relations are also social relations, and thus inevitably have informal as well as formal elements. As Mills observed in a foundational text of urban Political Science, members of the 'power elite' come to see that their 'several interests could be realised more easily if they worked together, in informal as well as in more formal ways' (Mills, 1956, p. 20). This unregulated dimension of politics is not, of course, just about elites. Outside the regulated spheres delineated above, political negotiations take place among and between different societal strata and social groups. Protest by subaltern groups to contest political structures, and coercion to secure political support, play out in all sorts of extra-legal and extra regulatory ways. The market power of particular actors and groups can strongly affect how formal rules play out in practice. Interest aggregation and lobbying takes place outside formal structures of pressure group activity; parties are characterised by unofficial factions; and many of the crucial moments in the formation of political coalitions take place through informal (and opaque) negotiations that are impermeable to formal rules. This often also involves the transfer of financial resources in ways that evade formal rules or blatantly contradict them. Sometimes the inherent ambiguity in official rules and regulations provides scope for informal negotiation and bargaining.
Much of the political negotiation that shapes public life and economic opportunity therefore takes place in unregulated or informally-regulated spheres. This much is obvious. What is less obvious is how this varies across and within societies and over time, why, and with what significance. As with economic and spatial informality, it can be argued that in less-developed countries political informality plays a greater role relative to formal politics than in more developed ones. This is certainly true if we define development itself as incremental improvements in organisational and institutional capacities (Brett, 2009). Indeed, much contemporary theory assumes development to be fundamentally linked to the nature and effectiveness of formal institutions (Acemoglu & Robinson, 2012;Rodrik, 2008). Pritchett, Werker, and Sen (2018, p. 24) summarise this perspective well with their claim that 'the key feature that distinguishes "developing" countries is the gap between the official, formal, legal, de jure laws and regulations and what actually happens.' In relation to the political sphere, such ideas date back to Huntington (1968) at least.
Observing a correlation between official rules that 'work' and development is not, however, the same as saying that development is caused by having 'good' formal institutions. This kind of determinism, perhaps most strongly associated with Acemoglu and his colleagues, sees formal institutions as the crucial factor in development success, with informal rules and practices playing at best a secondary and often anti-developmental role. Such perspectives are increasingly questioned through literature that places informal institutions at the centre of development narratives (Lund, 2006;Tsai, 2007;Williamson, 2009;Khan, 2010;Stacey & Lund, 2016). Pritchett et al. (2018) build on this by distinguishing between 'rules' and 'deals' when examining relations between governments and firms. While every society is replete with formal rules of various kinds, many developing countries are characterised by weak capability to enforce them. Such countries, they argue, are dominated instead by 'deals': interactions rooted in informal, personalised relations between people or organisations that are not based on the impersonal application of rules. Deals can vary in the degree to which they are open to people beyond a narrow elite, and the degree to which they are orderedthat is predictable and reliable. In politics, as in state-business relations more specifically, the way deals are struck can be seen as varying along axes of order and openness. These deals operate alongside formal rules, but in developing countries the latter are often subservient. Only when deals themselves become more predictable, open, and impersonalin other words, become more institution-likecan a society based on rules emerge.
The relationship between the formal rules (even if they are largely ignored) and the character of 'deals' comprises the 'political bargaining environment' in a given context (Goodfellow, 2013a), because deals usually involve some negotiation over how to use or abuse the formal rules. Thus, even though we should resist thinking of political informality as a generic residual serving primarily to explain the functionality of formal institutions (as in, for example, , the latter cannot be removed from the equation. The point is to shift the emphasis to reflect where much of the 'action' actually happens in most parts of the world, and to further conceptualise these processes and practices, going beyond a focus on their 'twilight' nature (Lund, 2006) to analyse their varying forms, functions, and meanings.
Writing primarily about spatial informality, Alsayyad and Roy note that 'If formality operates through the fixing of value […] then informality operates through the constant negotiability of value' (Roy & AlSayyad, 2004, p. 5). I suggest that formal rules and regulations attempt to 'fix' systems for accumulating and measuring political value, too: in other words, being a member of a party, voting for a particular candidate, or joining a trade union are means by which individuals help political capital accrue to a particular cause. They establish formal media of exchange between political support and corresponding benefits. Yet alongside these processes, informal bargaining that intersects with these formal processes is central to how political value is negotiated: for example, protests can enable certain groups to enhance their perceived value to a political organisation, ethnic populism can determine which organisations people affiliate with, and clientelistic practices can determine how people actually vote.

Political informality, clientelism, and trust networks
Clientelism may be the most obvious manifestation of political informality. Stokes, Dunning, Nazareno, and Brusco (2013) define clientelism as a form of 'non-programmatic political mobilisation' in which benefits are conferred by people in positions of political power (patrons) conditional on recipients returning these favours with votes or some other form of political support. Clientelism is thus based on a relationship of mutual (if asymmetric) exchange (Kitschelt, 2000, p. 849), often rooted in 'traditionally acquired norm orientations' (Lauth, 2000, p. 27). Although formal opportunities for patronage exist in many countries, much clientelism is informal and based on mutual favours.
The main challenge to institutional theories that posit development itself as a shift from less to more formally institutionalised societiesand concomitantly from more to less clientelistic systemsis not so much that they overstate the difference between developed and developing countries but that they misrepresent the process by which impersonal rules come to dominate. This is exemplified by differing interpretations of the evolution of democracy as a set of formal political institutions. It is 'taken as axiomatic that democracies rely more on formal institutions, while autocracies operate primarily through informal institutions' (Radnitz, 2011, p. 353). This association of democracy with formality is fallacious, however. Charles Tilly's later work, which builds on and critically engages with Putnam's (1993) work on democratisation, highlights the importance of informal institutions and networks for creating effective democracies historically (Tilly, 2004(Tilly, , 2007. This poses an intellectual challenge to currently influential theories that posit the shift to democracy as being driven by the creation of impersonal, open systems for representation and participation (Acemoglu & Robinson, 2012;North, Wallis, & Weingast, 2009). For Tilly, it is the partial integration of pre-existing 'trust networks' into public political life that facilitates democratisation. Trust networks are defined as 'ramified interpersonal connections within which people set valued, consequential, long-term resources and enterprises at risk to the malfeasance of others' (Tilly, 2004, p. 5). These might take such forms as religious sects and solidarities, credit networks, or patron-client chains. Thus clientelism itself represents a form of trust network. Trust networks are informal institutions in the sense that they embody social relations established over long periods, although they may develop their own rules and regulations and become more formal over time.
Tilly conceptualises democratisation as the intertwining of these informal networks of mutual support and cooperation with public authority, such that the former effectively concede some of their 'stock' of trust to governments. Trust between citizens and the state is not created by impersonal formal institutions of democracy, but is a precondition for them. These ideas contrast with a widespread perception in political science that informal institutions (including clientelism) necessarily 'collide' with democratic principles (Lauth, 2000;O'Donnell, 1996). In fact, formal politics in developed countries has not displaced informal, trust-based networks but has partially incorporated them and built upon them. The formation of states in the context of colonialism was generally very different; rooted in external domination, these states generally did not incorporate pre-existing trust networkson the contrary, the former tended to erode the latter, which again speaks to the gulf between formal and informal norms in cities of the South and the political volatility that often characterises these contexts.
These considerations raise important questions about whether shrinking the role of informal institutions to promote impersonal, rule-bound politics in developing countries is feasible or desirable. Lund argues that state institutions are never definitively formed, but 'a constant process of formation takes place' (Lund, 2006, p. 697). In some respects paralleling Tilly, his point is that 'twilight institutions', which may both mimic and give meaning and legitimacy to state institutions, are part of the process of state formation (Lund, 2006, p. 697). This is partly because political informality, being based largely on trust rather than impersonal rules, can help to legitimate formal politics. The opposite can also be true of course; political informality can de-legitimise the state. Unlike those such as Putnam (1993) who celebrate civic associational activity more uncritically, Tilly is well aware that some forms of trust network can be inimical to both democracy and well-functioning states (Tilly, 2007). This underscores the fact that there is enormous diversity in types of informal politics, and a need to understand the varying ways in which these interact with state institutions.
Tilly's work also alerts us to the relationship between these processes and cities, which were his other great preoccupation. His final, unfinished book focused on the interaction between trust networks (which embody the accumulation and coordination of commitment), cities (sites for the accumulation and coordination of capital), and states (representing the accumulation and coordination of coercion). The book argues that different configurations of capital, coercion, and commitment fundamentally shape people's quality of life (Tilly, 2010, p. 274). Although he was writing about the gradual evolution of human societies, his historical focus on the interface of cities, states, and trust has a contemporary parallel: the interaction of financial transfers, threats of coercion, and social networks in the bargaining over political value that takes place in contemporary urban settings. Before exploring how informal politics can play out in an urban development context, I build on the above discussion to explore and typologise different types of political informality.

Towards a typology of political informality
All politics is messy, and political strategies and tactics come in and out of focus in the dynamic reality of political action. This places limitations on any effort to put political processes into boxes, but the simplification that comes with heuristic typologies can provide tools to help penetrate and analyse this messiness. As noted previously, any concept of political informality has to be situated in relation to formal politics, and by extension to formal political institutions. Official rules, regulations, and processes, while not the focus here, are part of the 'dynamic process of move, countermove, adjustment, and negotiation' that shapes politics as a whole (Tarrow, 2012, p. 3). In trying to understand the informal dimensions of these moves and countermoves, I here distinguish between four types of political informality: pro-formal, anti-formal, para-formal, and a-formal. Each of these is now considered in turn.

Pro-formal
Some political practices that occur outside official procedures are nevertheless supportive of the formal political institutions in place. In their typology of informal institutions, Helmke and Levitsky (2004) include 'complementary' ones, which increase the likelihood that formal rules will be enforced (for better or worse). Building on this, in their discussion of unwritten rules in established democracies, Azari and Smith (2012) propose a category of 'completing' informal institutions, which are generally accepted norms that fill in gaps and ambiguities in existing formal political institutions. They also propose a category of 'coordinating' informal institutions: informal practices that aim to resolve conflicts and overlaps among existing formal rules that require reconciliation. Both of these can be considered as aspects of pro-formal politics, in that they exist to help formal systems work better. These can emerge as a consequence of elite practices, or in more bottom-up ways. Research on informal settlements in Ghana, for example, highlights how informal popular norms and practices help to legitimise formal political institutions as well as filling gaps in state provision of certain public goods (Paller, 2014;Stacey & Lund, 2016).
Sophisticated systems of unwritten political rules can support formal institutions in less democratic contexts, too. In her study of local governance in China, Tsai (2007) shows how even where state structures for accountability are weak, solidarity groups that overlap socially with government personnel can generate unofficial rules with which local officials feel obliged to comply. Informal norms of accountability thus fill gaps in formal institutional capacity (Tsai, 2007, p. 371). Some forms of clientelism can also be pro-formal. Analyses of the rise of Asian 'developmental states', for example, have highlighted the importance of informal strategic political alliances and cronyism for the effective functioning of those states (Evans, 1995;Gerlach, 1992). Indeed, the 'embedded' in Evans' (1995) conception of 'embedded autonomy' in states like South Korea refers primarily to informal political relations between government and business, which under specific conditions helped to make formal institutional systems function better through their coordinative utility.
If we take Tilly's ideas seriously, all modern democratic states emerged through informal politics that was at least partially pro-formal, rooted in volitional processes of integrating trust networks into coercive public authorities. Even today, analysts point to bottom-up processes of state formation emerging through traditional practices in places like Somaliland where increasingly formal political orders have emerged out of informal institutions (Boege, Brown, Clements, & Nolan, 2009). However, Somaliland is unusual in the context of contemporary developing countries due to its relative lack of internationally-imposed political-institutional systems. Most developing countries today already have in place an internationallysanctioned, sometimes donor-supported set of formal political institutions. As noted above, this kind of external imposition does not correspond to Tillyean negotiated processes of integrating trust networks. Such contexts are often replete with informal practices that actively undermine or disrupt this formal framework: in other words, anti-formal politics.

Anti-formal
Political activities that deliberately challenge or weaken formal institutions can be broadly subdivided into two: those that are top-down in nature, involving elite practices to build political support by creating a state of exception that exempts people from formal constraints; and those that are bottom-Political informality 283 up and involve popular mobilisation against extant rules. Politicians managing their linkages with citizens under conditions of competitive clientelism often engage in highly anti-formal behaviour. Political science has traditionally focused on 'programmatic linkages' between politicians and citizens, following its more general preoccupation with formal institutions, leaving other forms of linkage under-theorised (Kitschelt, 2000). While programmatic linkages are procedural in nature (Kitschelt, 2000, p. 853), informal political practices that seek to build political linkages in other ways can be highly anti-procedural. The kinds of top-down mobilisation strategies that have come to be termed 'neo-populist' are of particular relevance here.
Populism itself is not necessarily informal; historically, much populist politics has taken place through highly organised entities including labour unions as well as political parties. The idea of 'neo-populism', however, invokes linkages between charismatic leaders and largely unorganised subaltern groups (Resnick, 2013). The building of such linkages often takes place in ways that deliberately and consciously undermine formal institutions such as parties, as well as using a rejection of formal laws and systems to construct popular support. Neo-populism as defined here thus implies a marked anti-organisational and anti-institutional orientation (Weyland, 2001, p. 14). While long associated with Latin America, changes in Africa in recent decades, including the opening of political competition, have contributed to increasing recourse to 'neo-populist' strategies (Boone, 2009;Carbone, 2005;Resnick, 2013).
Political parties and formal institutions are often targets of rhetorical attack in this kind of populism (Carbone, 2005;Weyland, 2001), with the very sphere of the formal being vilified. This is particularly relevant to urban areas, where planning and regulatory systems are a convenient scapegoat for political figures seeking to position themselves as popular heroes. More particularistic clientelistic relations can also be highly anti-formal, especially where they are predicated on patrons pledging to 'protect' particular groups from rules and regulations. Building political linkages through the creation of a 'state of exception' is a hallmark of some cities in the developing world today (Roy, 2005). The degree to which groups are able to secure 'exceptions' from what they perceive as punitive formal systems reflects their political capital, and the deals they can consequently strike with politicians (Goodfellow & Titeca, 2012). Through their power as voters, or mobilisers of popular discontent, they can informally negotiate their political value in order to try and secure 'protection' from formal institutions.
These popular processes of negotiation reflect how bottom-up as well as top-down political strategies can coalesce around the contestation of formal rules and structures. More broadly, bottomup anti-formal politics can manifest in a number of ways, the most obvious being through protests or riots that emerge precisely because formal channels for popular participation are deemed unsatisfactory, repressive, or corrupt. Social movements, which necessarily emerge to contest formal structures, often use deliberately illegal tactics to make their claims 'at the gates' of the formal system (Tarrow, 2012). These activities are themselves often rule-bound, and in some contexts even violent protest can be seen as becoming so normal as to be informally institutionalised (Goodfellow, 2013b). Eventually anti-formal activism can even lead to new formal rules. Note, however, the contrast between these practices and those such as petitioning by a registered special interest association for a change in the law, which would not count as anti-formal since it works with official process and seeks to formally change structures rather than subverting or seeking unofficial 'exemptions' to them.

Para-formal
Unlike the above two categories of political informality, some political practices coexist with formal institutions without either actively supporting or intentionally undermining them. Some of these form distinct systems in their own right, which we can call para-formal. Certain types of clientelism and corruption fall within this category, including routinised forms of mutual exchange among elites which operate alongside formal rules for political engagement, often contradicting but not seeking to change the latter. For example prebendalism, a concept particularly associated with the literature on Nigeria and some other African countries (Joseph, 1987; Van de Walle, 2007), involves public resource streams being strategically allocated to key elites to help political leaders to secure political support. The official purpose of any particular public office, and the rules which govern it, remain in placebut they are a secondary concern (Joseph, 1987, p. 8). Prebendalism amounts to systematised corruption which reflects a moral code different from that embodied in the formal rules; it is expected that state offices will be acquired and used for the benefit of the office holders, their constituents, and kin groups (Diamond, 2013, p. viii). While corruption is predicated on the exertion of raw power, it also amounts to 'social activity which is regulated de facto and in accordance with complex rules, and tightly controlled by a series of tacit codes and practical norms' (Blundo, Olivier de Sardan, Bako Arifari, & Alou, 2006, p. 5).
Again, para-formal activity is not something confined to the sphere of elites. Consider one of the most influential concepts in recent subaltern studies -Chatterjee's idea of 'political society'. This is based on a distinction between two domains of politics: one which is linked to ideas of citizenship and rights-based participation (civil society) and another that relates to the connections between subject populations and governments that seek to provide for and control them (political society) (Chatterjee, 2004, pp. 39-40). The former is a domain of legal institutions while the latter is the domain of claims that are not rooted in formal citizenship rights, emerging from a (postcolonial) expectation that the state is obliged to govern and look after the poor, even though these obligations may not be codified. Thus political society is not devoid of rules or norms, but constitutes a sphere of informal institutions rooted in moral claims based on particular notions of community and reciprocal obligation. These two institutional domains co-exist, and politics as a whole involves 'a constantly shifting compromise between the normative values of modernity and the moral assertion of popular demands (Chatterjee, 2004, p. 41).
These ideas evolved in India, but are far from irrelevant to other postcolonial settings. For example, Ekeh's earlier argument that post-colonial African societies are characterised by 'two publics' has some parallels with Chatterjee's approach. For Ekeh (1975), in postcolonial African societies a 'primordial public' exists alongside a 'civic' public, with the formerbut not the latteroperating on the same moral principles as the private realm. These two publics thus involve different moral foundations, but the same political actors simultaneously operate in both (Ekeh, 1975, p. 93). The conceptual affinity with 'neopatrimonialism', which subsequently came to dominate African studies, is evident. Neopatrimonialism can be defined as 'a mixture of two co-existing, partly interwoven, types of domination: namely, patrimonial and legal-rational bureaucratic domination' (Erdmann & Engel, 2007, p. 105). Although these involve two distinct 'logics', they permeate and shape one another and cannot be considered separately (Erdmann & Engel, 2007, p. 105).
Notwithstanding their nuances and conceptual differences, these theoretical perspectives share a common concern with how informal norms and practices shaped by the post-colonial context exist in interactive parallel with the formal. Many of them amount to forms of patron-client relations, but the framework developed here enables a distinction between clientelisms that actively support or subvert formal rules and those that merely coexist and intersect with them, which we can call para-formal. Paraformal politics is akin to Pritchett et al.'s 'ordered deals', where there is a normative structuring framework that differs from the official rules. Not all kinds of political informality are necessarily this normatively structured, however; hence the need for a category of a-formal.

A-formal
The degree to which informal political bargaining and negotiation is bound by shared expectations is often difficult to determine; there is no obvious moment when bargaining transitions from a space of uncertainty and unpredictability into one of order and stability. At the extreme end, however, there are informal political processes that correspond to 'disordered' deals, embodying a politics distinct from all those above in that it neither exists to support existing formal institutions nor to systematically subvert them, and is relatively non-rule-bound in itself.
This kind of politics characterises the spaces where rules are unclear or non-existent, or where different norm systems clash to the point that no rules have effective force. In the sphere of spatial planning, zones beyond the writ of rules have been characterised by Yiftachel (2009) as 'grey spaces'

Political informality 285
where people exist between the 'lightness' of legality and security, and the 'darkness' of eviction and destruction. Building on Roy's (2005) characterisation of informality as essentially created by the state, Yiftachel suggests that the state consciously creates these situations of pervasive insecurity, which politicians and officials can capitalise on through promises of 'whitening' or threats of 'darkening'. The negotiation that takes place in these 'grey spaces' largely occurs outside of trust networks, and it is the relative weakness of norms of trust and reciprocity that creates opportunities for political manipulation and exploitation.
This idea of negotiation in grey spaces has broad relevance beyond the sphere of planning. Chattaraj uses the concept of 'jugaad' governance, which in Indian popular discourse means 'cobbled together' and implies 'making do, crudely improvising or even innovating under constraints'. 'Jugaad governance' therefore refers to interactions that are 'negotiated, flexible, improvised and often extralegal' (Chattaraj, 2016, p. 2). Unlike clientelism (which is normatively structured), the a-formal is a sphere of tactics, uncertainty, and unpredictability where institutions are withheld or break down, rather than a space of alternative institutions. Here political value is in a constant state of negotiation. Some influential scholarship on cities in Africa and Asia suggests that much of urban life plays out in this way. Most notably, the work of Simone (2004Simone ( , 2005 has extensively explored the provisional and contingent nature of practices and relationships in these contexts. For Simone and Pieterse, the way in which most urban-dwellers try to improve their lot is through improvisations and makeshift relationships that are 'rarely institutionalised in to a fixed set of practices, locales or organisational forms' (Simone & Pieterse, 2018, p. xi). Such a-formality may be fleeting, characterising situations where new norm systems have yet to fall into alignment, but in some sectors of society it might be a more pervasive and enduring feature of everyday life.
The purpose of outlining these categories is not to suggest a rigid framework for putting particular practices in boxes, especially given significant overlaps. The point is to offer a typological language for analysing the diverse array of informal political practices prevalent to varying degrees in contemporary societies, and how they interact with formal structures and systems of governance and political organisation. Figure 1 maps these categories on a two-way axis, where the vertical axis refers to the degree to which informal political practice is norm-oriented/rule-bound, and the horizontal axis is the degree to which the practice is supportive or subversive of formal institutions. 6 This formulation emphasises how the categories are not neatly bound, but can serve as tools to situate forms of political informality against axes that capture salient aspects such as their predictability, regularity, and interface with formal systems. While neither para-formal nor a-formal practices are necessarily subversive of formal institutions, their positioning in Figure 1 indicates that they are relatively likely to be subversive in effect. Similarly, proand anti-formal activity are not necessarily highly rule-bound, but are likely to be, particularly in cases where they are relatively successful in supporting or contesting existing institutions. The diagram aims to capture fluidity, not fixity: the reasonably norm-oriented sphere of political society, for example, which above is classed as 'para-formal', could easily slip down the axis into more unpredictable and contingent sphere of 'jugaad' negotiations.
This framework does not in itself explain why different kinds of informal political processes and practices emerge, and under what conditions politics is likely to become (for example) especially anti-formal or a-formal in nature. Literature on formal institutions, including Levitsky and Murillo's (2009) work on stability and enforcement as two dimensions of formal institutional strength, offers some clues here. One can speculate from their framework that where formal institutions are stable but weakly enforced, para-formal systems are especially likely to emerge. However, where formal institutions are unstable (that is they change frequently), this may stimulate anti-formal politics because groups who do not benefit from the rules in place see that it is possible to change them. Concomitantly, those who do benefit from existing (but unstable) rules are likely to engage in proformal practices to prevent those rules from being subverted. More generally, the degree to which formal institutions are concordant with existing social dynamics (for example the extent to which they have integrated trust networks) is likely to influence the intensity of informal politics.
This kind of speculation can only take us so far, however, and there is a danger in always taking formal institutions as the epistemological starting point for understanding politics, even if we accept they are always part of the political bargaining environment. The question of why political informality takes certain forms is primarily an empirical one. Answers will depend on political and institutional conditions in specific contexts and how they have changed over time. Different kinds of political informality may also coexist in the same space, and it takes empirical analysis to tease out and explain this. While there is not adequate space here to examine an empirical case in great detail, the remainder of this article explores the potential of the above framework for understanding political informality in a complex urban environment: specifically that of marketplaces in Kampala, Uganda.

Political informality and urban analysis: the case of Kampala
As noted previously, in colonised societies it was often only in urban areas that 'modern' rules, and therefore formality itself, were intended to take root; hence it is in urban areas that the idea of informality acquires its greatest salience. One of the first conceptual frameworks in political science explicitly concerned with informal politics was also focused on urban areas: namely the idea of 'machine politics', popularised in mid-twentieth century analyses of political organisation in the United States in the late nineteenth century (Scott, 1969). Urban political machines involved layered systems of patronage, where material rewards were traded for political support in ways that violated the norms of public office. However, the urban political machine as a concept is now widely seen as being inseparable from the US historical context, and of little relevance to developing countries today, where the competing pull of ethnic, religious, and regional appeals often provides more effective and less costly ways of forming linkages with potential supporters (Nelson, 1979, p. 201).
Marketplace politics in Kampala in recent decades is a particularly apt arena for studying political informality in practice, as it reflects many of the broader dynamics of urban politics in the city during a period of significant change. The market trading sector since the 1990s has been one in which the Political informality 287 roles of public, private, and civil society actors have constantly been in flux, as a wide range of interests sought to capitalise on economic growth, and the National Resistance Movement (NRM) government rolled out a wave of new decentralisation and privatisation policies before finally imposing an entirely new governance structure on the city. Over the same period, in response to ongoing pressure, the NRM formally introduced multi-party democracy while always attempting to find new ways to retain power, and ethnic and regional politics also re-intensified. All this created an extremely fertile and competitive political environment in the city. As Lindell and Ampaire note, the politics of Kampala's market traders in this period was 'fluid, untamed and pragmatic, but also contradictory and fractured' (Lindell and Ampaire,206,p. 257). The framework presented in this paper offers a way of analysing some of the deeper patterns of political process and practice that these contradictions and fluid interactions represent.
Kampala's marketplaces have a long history of anti-government mobilisation dating back to the colonial era (Monteith, 2017). In the post-colonial period, strong social networks were maintained and regularised practices of engagement with the state evolvedbut these were radically challenged in the context of changing government policy from the 1990s. In Nakasero market in the city centre, for example, Kampala City Council's (KCC) efforts to redevelop the market in the early 1990s proved deeply unpopular with vendors and a series of strikes, blockades, and violent acts unfolded. Vendors also began to re-organise their market associations, both within and across marketplaces (Monteith, 2016, p. 155; see also Lindell & Ampaire, 2016). This led to some increased formal political engagementfor example vendors successfully lobbying the Inspector General of Government to launch a report into the mismanagement of Kampala's markets. As government policy shifted towards privatisation later in the decade however, much of the political activity that ensued was informal in nature and there was a breakdown in established relations between KCC and the vendors, which had previously consisted of regularised negotiations over issues such as market rates. Social ties among vendors were instrumentalised by business 'tycoons' seeking to build support for their bids to formally purchase leases for entire marketplaces in the face of strong resistance, which created a parallel system of governance predicated on new and divisive spheres of influence. From the vendors' perspective, this was disorientating and linked to the perceived arrival of 'bad' people and concealed allegiances (Monteith, 2016, p. 161).
Meanwhile, the controversy over selling off the market leases at a time of increased political competition in Ugandaparticularly after multi-party politics was re-introduced in 2005provided significant scope for politicians to enter the marketplace with promises of a better deal for market vendors. This approach was personified above all in Nasser Ssebagala, the city's first directly elected mayor from 1998. Ssebagala was typical of populist leaders in that he was an outsider relative to conventional political elites, and a Muslim rather than a member of the Protestant or Catholic elite of the Baganda ethnic group who have traditionally dominated Kampala's politics. Associated for most of his political career with the opposition Democratic Party, he built his populist appeal around being the son of a Nakasero vendor, with an agenda to empower vendors. His catchphrase was 'seya', which has been translated as someone who 'wanders' and 'makes visits' (Monteith, 2016, p. 158), but also as a local adaptation of the word 'share' used by the urban poor to denote camaraderie. 7 'Seya' became his nickname, accompanying his slogan, 'forward with the common man's revolution'. 8 Shortly after being elected, Ssebagala was imprisoned for 11 months in the United States on fraud charges. 9 This only increased his popularity as he was seen as a martyr, as well as a source of wealth that might be shared with his followers; he returned in 2000 to a hero's welcome, ran again for mayor and won decisively in 2006. The markets were crucial sites of support. 10 Despite his promises to empower vendors, once back in power Ssebagala made a dramatic volteface and embraced the national government's privatisation agenda, continuing to negotiate the lease of some of Kampala's markets to politically-connected private companies. 11 As anger at his betrayal grew among vendors, the Ugandan President Museveni saw an opportunity to reclaim urban support for the NRM and made a highly populist effort to usurp the Mayor's role as the 'saviour' of market vendors. Museveni thus cast aside the official policy of privatising markets by announcing support for vendors' taking direct control, which led to a surge of NRM support in the marketplaces (Goodfellow & Titeca, 2012). Ssebagala too made a further U-turn, reverting to his earlier claims of support for vendors' rights, but this came too late as the sale of Nakasero market to a prominent tycoon had already been processed. 12 Not wanting to lose his own newly-acquired political capital in the markets, Museveni met personally with vendors, declared the sale nullified, and gave clearance to the vendors to redevelop Nakasero market themselves. 13 In so doing, he managed to significantly boost his popularity among a key sector of the population in Kampala, an opposition stronghold.
Both Ssebagala's and Museveni's behaviour in this period epitomised anti-formal politics, characterised by competing informal personal interventions that threw official government institutions and policies into disarray in the interests of building political support. Meanwhile, vendors themselves were engaging in a range of different 'modalities of struggle' (Lindell & Ampaire, 2016) that encompassed not only anti-formal behaviour such as unauthorised protests and riots, but a number of distinctly pro-formal forms of mobilisation and formal political activities. For example, in their appeals to the President, vendors explicitly invoked formal policies and laws such as the 1998 Land Act to claim to their rights as 'sitting tenants' in the market. As Monteith explains, they drew upon 'an impersonal language of legal "rights"a tool of their historical exclusion in Kampalain order to prompt a personal intervention from a leader in need of new followers' (Monteith, 2016, p. 63). They became what Holston terms 'legal strategists', strategically using law 'to avoid becoming its victim' (Holston, 1991, p. 722). Through a variety of forms of informal politics, they managed to achieve an official turnaround in government policy in late 2007 (Lindell & Ampaire, 2016, p. 268).
These events did not mark the end of heightened marketplace politics. In Nakasero, a dispute stretched two further years before vendors were officially granted legal ownership of the market, and this was followed by increasingly centralised control by market leaders as they sought to raise the revenue to pay the market lease fees and fund the proposed vendor-led redevelopment. Market governance became increasingly bureaucratic, and requirements such as having proper identification cards were rigorously and often violently enforced. A culture of 'hard' management and discipline came to the fore, some of which was formalised though market laws, but backed up with beatings and intimidation (Monteith, 2016, pp. 169-174). This amounted to pro-formal intervention in the sense that it was used to harden a new set of formalised market rules.
Even as this bureaucratic centralisation was under way, another dynamic was emerging in relation to party politics. Although previously an opposition stronghold, the new market leadership structures were deeply penetrated by NRM supporters since Museveni's increased attention to markets discussed above. This highly partisan interference in marketplace organisation eroded existing social networks and support systems. Trust networks were felt to have disintegrated, and an environment of heightened insecurity and confusion emerged, particularly in the context of the 2011 elections. The environment for making 'deals' became significantly disordered from this point onwards. When KCC was replaced by the unelected Kampala Capital City Authority (KCCA) in 2011, some vendors wrote to KCCA complaining about the market leadership, precipitating an effective government takeover of the market. By this time, market vendors were severely divided and disoriented; so much so that there were no protests, despite KCCA's violence to the long-held dream of vendor ownership and control (Monteith, 2016).
These two decades of intense upheaval, interpreted by some vendors as a shift from marketplace 'culture' to marketplace 'politics' (Monteith, 2016), can also be parsed through a lens of political informality. The value of this is in helping to illuminate how political practices vary in response to changes in rules and associated shifts in expectations and social norms. The markets were always characterised by a degree of para-formal politics, in the form of networks of solidarity and political clientelism underpinned by relatively stable expectations. However, the radical policy prescriptions and abrupt policy reversals characterising the above narrative transformed the political bargaining environment over time. A series of anti-formal political interventions by politicians, who were seeking electoral support in the context of vendors' anger about privatisation, was matched by anti-formal activism by vendors interspersed with pro-formal efforts to root their case in legal frameworks. This combination of pro-and anti-Political informality 289 formal politics was effective for them initially, securing both informal and then formal changes in their favour.
However, the reality of vendor-ruled markets meant that an elite cadre of vendors emerged who then sought to protect their own position and interests through aggressive and sometimes violent pro-formal activity in the markets. The combination of these dynamics with heightened formal party interest in the market further undermined the foundation of the market's original trust networks, eventually generating an everyday marketplace politics that was predominantly a-formal and contingent, with no new protests or effective mobilisation emerging. Interestingly, it was the greater penetration of the market by formal politics in the guise of political party networks that exacerbated this a-formality in the vendor response. Within the market, 'politics' itself came to be seen by many vendors as amoral and unstructuredeven as a sort of 'madness' (Monteith, 2016). Thus by the mid-2010s there was no certainty of the norms through which vendors could negotiate their value (in other words, secure themselves relatively stable and predictable relationships of mutual favours). This echoed the situation in other markets in the city over the same period (Goodfellow, 2013b;Philipps & Kagoro, 2016).
Existing frameworks for understanding urban politics cannot adequately capture the unwieldiness and complexity of these overlapping phases of politics identified within Kampala's market sector. There is no political 'machine' in operation here, and nor is this a straightforward narrative of patronclient ties, or of the evolution of a social movement. Yet nor is this a narrative without any shape or trajectory; it is not all about 'uncertain, unanticipated connections […] trying to operate outside the accustomed discursive sites of sociality' (Simone, 2005, p. 525), although it increasingly became so towards the end. Examining how different kinds of informal politics respond to changes in formal institutions and to each other, generating a bargaining environment that leans towards one or other modality of political informality at a given time, can help us to better understand why people make and break political linkages in particular ways.
This analysis also helps to problematise the commonplace association between informal workers and anti-or a-formal politics. Although market vendors did riot and engage in informal disruptive tactics, it was often also they who initiated the more formal political practices (such as petitioning and registering new legal entities to push for their claims) and pro-formal activities, including popularising their own interpretation of the law and strategic 'switching' between different organisational forms to represent themselves in different fora (Lindell & Ampaire, 2016, p. 279). Elites, on the other hand, used some formal processes such as the courts but were more notable for their illegal and anti-formal behaviour, allegedly forging documentation and engaging in bribery as well as circumventing formal procedures (Lindell & Ampaire, 2016, p. 277).
These reflections on Kampala's marketplace politics also support the earlier speculation about how in places where formal institutions change frequently, pro-and anti-formal politics dominates over relatively stable forms of para-formal clientelist politics. 14 Successive formal institutional changes since the 1990s, partly spurred by international donor agendas, disrupted what had been relatively stable forms of clientelism in the markets. Over time, with the disruption of these patron-client ties came the unravelling of the trust networks embodied in them, leading to increasingly a-formal politics as processes of negotiating political value became uncertain and discordant. Only by exploring how multiple forms of political informality intersect over an extended time can we make sense of a situation that was routinely just described as a 'mess'. 15

Conclusions
This article contributes both to the literature on urban informality and to debates on politics in developing countries. It has presented a new framework for conceptualising political informality, rooted in a wider literature but with particular relevance for granular analysis of how urban politics evolves amid changing policy prescriptions, overlapping institutions, and shifting associational activity. As such, it offers fresh lenses on political dynamism and complexity, indicating how an analysis of shifting political informality over time can illuminate the 'chaos' and fluidity' of urban political life. The kind of informal politics that dominates at a given place and time, and how this relates to formal political structures and processes, matters in some fundamental ways. Para-formal politics is relatively predictable and often based on norms of reciprocity, but undermines the legitimacy and effectiveness of the formal political system so long as the two remain separate. Meanwhile, formal politicsand any pro-formal politics that accompanies itreifies a set of political institutions, and if this is the only kind of politics, institutions will be rigid and persistent even when they are flawed. Anti-formal politics is therefore especially needed in situations of institutionalised repression and exclusion. Intuitively, a healthy political system requires a balance of pro-and anti-formal politics, whereby laws and social norms become mutually reinforcing butcruciallyare also open to effective challenge.
This does not mean that pro-and anti-formal politics will necessarily generate 'good' urban politics. It is also important that formal institutions themselves build on something pre-existing, even if the ultimate aim is to be able to change that 'something' rather than be held hostage by it. In other words, rather than eradicating clientelist relations, which is usually impossible without great violence, inclusive politics involves building formal political institutions onto pre-existing social relations while also creating accompanying mechanisms for these relations to be contested and adapted. The case of Kampala's marketplaces represents a rather different story. Instead of rooting formal institutions for governing markets in the existing socio-cultural networks, a series of new formal rules were imposed from outside, shaped strongly by a privatisation and profit-maximisation agenda. Far from integrating trust networks, policy agendas such as this necessarily deepen the rift between formal institutions and parallel social institutions, which are based on trust networks and non-marketised reciprocal obligation. If we were to speculate on the general implications of the Kampala case, it suggests that when such marketised institutions are combined with partisan politics, a range of extreme pro-and anti-formal political processes can emerge as actors seek to defend or subvert these formal structures depending on their interests. Unlike the healthy balance discussed above, however, this heightening of anti-and pro-formal politics without adequately incorporating pre-existing trust and solidarity networks can eventually produce heightened uncertainty and a-formal politics.
What potential does such a-formality offer? In theory, a weakening of existing norms and collective moral principles holds the potential to be empowering for somefor example those who might be held back by social institutions such as those relating to caste or gender. But ultimately the lack of trust and predictability in a-formal deal-making is hazardous, exposing people to immense risk and manipulation. The organisation of vulnerable groups into formations that can produce new, more equitable trust networkswhich can then assert their relevance in either pro-or anti-formal ways, and ultimately demand incorporation into democratic structuresis much more likely to generate progressive and inclusive formal politics than a-formality. Overcoming hierarchies and exclusion necessarily involves people being able to negotiate their political value, and history demonstrates that this only happens when people mobilise collectively, building on shared principles and expectations, but also informally. Only with some political informality can formal structures for more equitable negotiation be permeated with social meaning. The institutional economists may be right that rules matter for development; but the path to such rules is strewn with deal-making, and with the building of trust networks to achieve the resilience of ordered over disordered deals.