Order and Disorder in African Cities: Governance, Politics, and Urban Land Development Processes

Perceptions of urban chaos and fear of disorder have two historical roots. The first is the experience of unregulated urbanization in other parts of the world, especially 19th-century Europe. The unsatisfactory urban growth patterns, mixed land use, and unsanitary living conditions which resulted led to concerted efforts to develop municipal infrastructure technology and to plan and regulate urban development. Colonial administrators' dismay at the apparent squalor and disorder of indigenous urban settlements stemmed, therefore, from their ideas about the form that orderly urban development should take, as well as their lack of understanding of the social and political organization of indigenous societies. Second, the fragility of the colonial military, political, and administrative apparatus fueled fears that they might not be able to maintain control over large and politically volatile urban populations. Such fears have not receded with the demise of colonialism: postindependence governments have been unable to keep pace with the demands of rapid urban growth, and the need to maintain social and political control over urban populations has strongly influenced their political and economic strategies.

Both colonial and postindependence governments have, therefore, attempted to impose "order" on towns and cities, in particular with respect to politico-administrative systems for urban management and the development of urban land. This paper will, first, explore the form which these attempts to achieve orderly urban development have taken, and second, assess the extent to which they have succeeded. We will see that their influence on urban development has been limited, with the result that most political activity and land development do not comply with the rules of the formal political and land administration systems. The common view that cities are, as a result, chaotic will, however, be challenged. It will be asserted that, far from behaving in an anarchic fashion, actors in urban politics and land development base their behavior on widely understood and accepted, if informal, rules for social interaction. As a result, the appropriateness of the concepts of order underlying attempts to institutionalize particular forms of political and physical order is questionable. Finally, some pointers for the future will be identified.

The urban experience on which this paper draws is largely that of the medium-sized cities of African states which have not been embroiled in civil war. These are cities in which relatively few people are, literally, homeless; some services and utilities continue to function, however inadequate their coverage and unreliable the provision; and social and political relationships between city governments and their citizens and between residents themselves are not, generally speaking, anarchic, although dysfunctional political relationships, social malaise, and poor urban governance are widespread.² Most of the examples are drawn neither from collapsed states nor from urban settlements at the extremes of the size continuum. This paper therefore complements the in-depth analyses of the cities on which this volume focuses, although the generalizations it contains should, of course, be treated with caution.

In trying to achieve political control over growing urban populations, regulate urban growth, and improve health and efficiency, decision makers have imposed their own ideas of what constitutes order and systems by which that order is to be achieved. Here, colonial and postcolonial attempts to impose political and physical order will be outlined.

The political systems introduced by colonial administrators reflected both the political philosophies and systems of their home countries and the means adopted to secure control over the resources required by colonial trade and enterprise. Both of the principal colonial powers — Britain and France — had liberal democratic political systems in formation. However, their legal systems, degree of centralization, and view of the political relationship of their colonies to the mother state differed. As a result, the political system instituted in the urban settlements established alongside indigenous towns or in areas of agriculture or mining enterprise reflected both the characteristics of and differences between their home countries, and the differing approaches to colonial rule adopted in different parts of Africa. Where the new towns were established by mining companies (as in the Copperbelt of Zambia), democratic structures were absent or weak. Elsewhere in anglophone countries, local councils were established based on ward-level elections and the enfranchisement (as in the UK at the time) of property owners, which conveniently restricted representation and decision-making power to male settlers, businesspeople, and colonial administrators.

Gradually, as independence approached, the representation of previously excluded groups improved: initially tokenistic, by the time of independence the scene was set for representative local councils with a degree of autonomy (if often apparent rather than real), although often the preexisting indigenous settlements (especially in countries under indirect rule) were not fully part of this political system, and in the settler colonies, the parts of urban settlements reserved for African occupation were separately administered. Unlike anglophone colonies, francophone colonies' status and the centralized political and administrative structure of the French state resulted in a top-down system of urban administration which persisted long after independence.

Capitalist development in Europe and North America was accompanied by the individualization and formalization of land and property rights, with legal systems designed to protect the interests of owners. Following rapid uncontrolled urbanization during the industrial revolution and the early years of the 20th century, systems for planning and regulating urban development were gradually introduced. These were underpinned by conceptualizations of the nature of property rights and the legitimate role of the state, and were influenced by the principles on which the legal and political systems were based in the countries concerned, their differing historical experiences and geographical characteristics, and the constellations of interests dominant in particular historical periods. As noted above, the primacy of individual rights to the enjoyment of property in perpetuity and to dispose of it freely was the basis of a liberal capitalist economic system. Moreover, it was believed, on the basis of economic theory, that given accessibility constraints, the most efficient urban land use patterns would emerge if owners were free to sell their property to the highest bidder, leading to a concentration of uses capable of paying the highest rents and using land most intensively at the most accessible nodes of the urban area, usually the city center. Only where one rightsholder's enjoyment of his or her property interferes with the ability of others to do the same is state intervention, in theory, legitimate. Although in practice state intervention may be extended based on additional goals that are also considered legitimate, in principle, the rights of property holders are primary and the state can only curtail these rights with their consent.

One of the purposes of state intervention is therefore to protect and guarantee the rights of property owners and occupiers by registering those rights and enforcing contracts for their transfer. Land registration is also needed for property to be used as collateral for raising investment funds. To this minimum, historical experience has added a series of other state roles which are, to a greater or lesser extent, accepted as legitimate.

First, action to provide (or ensure the provision of) physical infrastructure and regulate standards of development in order to protect public health was accepted, not least because many of the diseases related to deficient utilities or poor waste management affect all income groups (e.g., air pollution, cholera), even if they have the greatest impact on the poor. Municipal engineering technologies and standards, therefore, were developed initially to deal with rapid urbanization associated with industrialization in the North. When the colonial powers began to develop urban enclaves to accommodate their businesses, officials, and other nationals, as well as their local employees, they applied the same engineering technologies and standards, although often adjusted in line with perceptions of their new environment. Thus colonial appropriation of large tracts of land for urban development was underlaid by a failure to recognize existing land rights and a perception that land was abundant,³ leading to the adoption of lower densities than in Europe.

Second, the European experience of unregulated urban expansion resulted in cities where the juxtaposition of residential areas with industries spewing uncontrolled emissions of waste and pollution, as well as a shortage of sites for public facilities, adversely affected health and quality of life. It also resulted in the conversion of large swathes of agricultural land for urban use, a danger brought home to high-density countries such as the Netherlands and the UK by food shortages during and after World War II. Such experiences led to the institution of effective systems of planning and development control. It was accepted that the state had the right and capacity to curtail the interests of property owners in the wider social interest. The common experience of the adverse consequences of unregulated urban development and belief in an efficacious and well-meaning state generated such wide public support that commercial developers and investors were forced to accept the limitations on their freedom to operate. Although the restrictions also affected individuals, increasing home ownership led more and more urban households to recognize that the restrictions were outweighed by the degree of protection thereby afforded their living environments and property values.

In the colonial era, therefore, the political control necessary to ensure the viability of colonial enterprises, the financial self-sufficiency of (especially British) colonial administrations, and the safety of colonial and settler populations were inextricably linked to claims on land and property as well as the European tenure and land administration systems considered necessary to promote enterprise, safeguard the interests of business, and protect the health and living standards of European urban residents. The use of land policy and planning as an instrument of social control was demonstrated most starkly in the settler economies of countries such as South Africa and Zimbabwe. The refusal to extend political and property ownership rights to other population groups, especially indigenous rural-urban migrants, however, was strongly contested throughout the colonial period, and especially from the mid-1940s onward.

In the years leading up to independence, efforts had already been made, especially in the British colonies, to prepare the ground for representative democratic political systems at the urban level. The structures of broader-based democracy were, therefore, already in place, with administrative structures staffed largely by expatriate professionals capable of managing the (generally small) cities which were the capitals, ports, mining towns, and agricultural service centers of the newly independent states. In theory, the extension of the franchise and abolition of separate administrative arrangements for the European and African sectors of cities created conditions for continued orderly development of the inherited urban fabric. In practice, the imperatives of state formation, the pursuit of national economic development goals, the desire to fulfill political promises made to the peasantry, conflicts between factions (often along ethnic lines), and a desire to hold onto power invariably led to centralization of political and administrative power and the imposition of authoritarian political systems.

In most countries, a military ruler or executive president emerged who was able to stand above and manipulate factions or bypass parties and place the bureaucracy under his own control (e.g., Zambia, Tanzania, Senegal, Cameroon, Kenya). Representative institutions such as parties, parliament, local government, and trade unions were downgraded or abolished. Although many of the regimes instituted mechanisms for allowing the expression of political voice, generally through the machinery of the dominant party ("participatory democracy"), the scope for freedom of expression was rapidly eroded and the machinery used predominantly for information giving, mobilization, and eventually repression. Single-party or military rule at the national level was reproduced at local government levels, although elections were, in some cases, retained. In some of the more populist single-party systems, political stability and developmental progress were achieved until, from the mid-1970s onward, economic crisis, growing authoritarianism, and decreasing accountability undermined the regime. Elsewhere, periods of military rule alternated with shaky civilian regimes, opening the way for extreme clientelist spoils politics (e.g., Nigeria, Sierra Leone, Liberia) and sometimes complete breakdown. In such systems, the winner (the dominant faction) takes all, looting an economy dominated by the black market, general economic crisis, pervasive corruption, and the use of communalism to mobilize political support, resulting in widespread instability and the erosion of authority.⁵

Everywhere, in the 1980s, economic crisis followed by structural adjustment depleted the resource base not only for achieving economic growth but also for the patronage resources on which both civilian and military regimes depended to manufacture consent. Eventually, by the later 1980s, domestic movements for political change, reinforced by external pressures, aid conditionality, and the changing geopolitical order, resulted in a widespread process of (re)democratization. The core areas of political reform in the 1990s have been those associated with liberal democracy: political plurality, fair elections, and decentralization.⁶ Thus processes of political change at the national level have been reproduced at local government levels: where elected single-party local councils had survived, local elections were increasingly contested on party political lines, and where they had not, local elections were reintroduced. The processes leading to redemocratization differed, and the constitutional outcomes and subsequent experience vary as well. However, the vast majority of Africa's democracies today are fragile and unconsolidated. The key problems are identified by some as primarily external but by others as essentially internal. Rita Abrahamsen, for example, argues that democratization is unsustainable in societies dependent on externally controlled resources which are also subject to conditionality based on neoliberal economic theory.⁷ The pressure for democratization, she suggests, was first and foremost a demand for socioeconomic change and, although political rights may be valued, the new regimes were expected to restore mass prosperity. Unable to do this because of their lack of control over policy and resources, they are "exclusionary democracies: they allow for political parties and elections but cannot respond to the demands of the majority or incorporate the masses in any meaningful way"⁸ and are therefore unlikely to become consolidated. Yusuf Bangura, in contrast, attributes failures of democratic consolidation to an inability to maintain elite cohesion (resulting in factional struggles to gain political power and control over state spoils), failure to hold the bureaucracy and the armed forces to account, constraints on the expression of political voice, and lack of respect for some of the key assumptions of a liberal democratic model (e.g., fair elections, peaceful alternation of power, separation of the bureaucracy from the ruling political party).⁹

Both in the independence settlements of most ex-colonies and in the late 1980s, therefore, the dominant ideology of political order has been liberal representative democracy, in which

The concern of liberal democratic theory is therefore not with the rightful place of active citizens in the life of a political community, but with the legitimate pursuit by individuals of their interests and with government as a means of enhancing those interests.¹¹ The state can only govern by consent, which can be withdrawn if individuals believe their rights (to liberty, prosperity, or freedom of expression) have been violated.

Prosperity is to be achieved through the operations of a capitalist economy based on the matching principles of economic liberalism. In theory, therefore, the state should not pursue large-scale economic and social goals, since property is privately owned, economic functions operate through the market, and government interference is deemed counterproductive. But in practice the boundary between the public and private spheres is contested and views on the legitimate role of government vary. Thus the compacts between colonial powers and early nationalist leaders during the transition to independence instituted liberal democratic political systems, but coupled these with a strongly interventionist role for the state in the economy. In comparison, the move toward liberal democracy in the 1990s, again resulting from a combination of internal and external pressures, was associated with a neoliberal approach to economic policy.

Based on the assumptions of economic and political liberalism, much of the longstanding debate about democracy has, therefore, taken the form of attempting to identify the most suitable mechanisms for selecting rulers who will ensure that the rights of individuals are protected and can be held accountable in this respect. In its consolidated late 20th-century form, representative democracy is organized around a general cluster of rules and institutions, which include those identified by Bhikhu Parekh above and also political competition based on the principle of alternation of power and organized through political parties which embody different political platforms and ensure pluralism; associational autonomy in order, inter alia, to enable government to manage conflicting interests, to disperse power between them (organized as interest groups or political parties) without any becoming dominant, and to ensure accountability; and institutional separation of powers between the executive, the legislature, and the judiciary, in order to ensure the rule of law and the subordination of officials of the state to the law and accountability of the executive to the legislature.

By adopting these mechanisms of democracy, it was expected, relations between state and people would be satisfactorily regulated, competing interests would be balanced and reconciled, development objectives of economic growth and basic needs satisfaction would be achieved, and elected representatives and the bureaucracies which implement their decisions would be held to account. Moreover, citizens would be willing to pay taxes and comply with regulatory systems, since the transparency of resource allocation and decision making would give them confidence in the ability of decision makers to respond to their needs and demands, and avenues of redress would be available through either voice in the political process or resort to the legal system.

With respect to the administration of private property and physical order, little has changed. The conceptions of physical order, and the planning and regulatory systems designed to ensure that the urban built environment fitted these conceptions, have persisted. Exported to Africa by the colonial powers, they have been perpetuated by legislation and practice, and reinforced by the continued export of Northern ideas and technology through models of professional training, the use of international consultants and contractors, and the preconceptions and practices of international development agencies and their staff. These ideas and practices relate to tenure and administration of land and property, planning for urban development and infrastructure provision, and systems for development regulation. The belief persists that individual title, underpinned by administrative and legal systems to protect and reinforce property rights, is the most appropriate land tenure system. It is accompanied by a belief that engineering efficiency is best achieved through installing infrastructure in straight lines; a supply-driven approach to urban water provision which downplays the real costs and discourages demand management; and the notion that universal road access and water-borne sewerage are desirable. The latter led to excessive width standards for roads, and human waste-disposal methods which are costly in terms of finance and water use. With respect to regulation, notions of physical order based on land use segregation reinforce attempts to universalize zoning and land use control and formalize informal activities. In addition, there is also a belief that detailed control over the construction of individual houses is essential in order to protect urban dwellers against builders of substandard housing, rapacious landlords, fire, infectious disease, and other health hazards.

As is clearly evident in African cities, attempts to base urban political organization on liberal democracy and land administration on individual property title and land use control have been problematic. An important question, therefore, is whether the problems arise from the assumptions on which the models are based — assumptions about what characteristics political and physical order should have and how it should be achieved — or from shortcomings in the way the political and land administration systems have been designed and operationalized.

In Africa, despite progress in restoring civil and political rights and democracy in the 1990s, political instability is endemic, and some multiparty democratic reforms have been reversed. There are clearly a number of problems, in practice and principle, in applying the liberal democratic model to Africa:

These problems, and the last in particular, lead to the question of whether, if competitive liberal democracy is so hard to institutionalize in Africa, the conception of politics on which it rests and the assumptions of the model itself are appropriate to the African context.

Although democratic theory holds that democratic local governance will increase political participation, resulting in the more adequate representation and empowerment of varied political interests, in turn leading to more wide-spread policy benefits such as poverty reduction,²³ liberal representative democracy is no less problematic, if less studied, at the urban level. At this level, the first key issue is the division of resources and functions between central and urban levels of government, the political and operational framework at each level, and the relations between them. A second set of issues relates to the design and operation of the electoral system, and a third to questions of accountability.

Except to some extent in federal systems, where local autonomy is enshrined in a constitution, the degree of local autonomy depends primarily on central government motives, which are supposedly to share power and achieve economic efficiency.²⁴ In theory, devolution is intended to enhance local responsiveness and accountability and to provide a check on the power of higher levels of government. In reality, a national regime's motive for power sharing is generally to enhance political stability and its own hold on power, by providing opportunities for the expression of preferences at the local level and the representation of local groups and minorities in decision making, and by reducing the extent to which central government can be blamed for poor performance. In practice, local representative political organization has more often been seen as a challenge to both political stability and the ruling party, by providing sectional interests or opposition parties with a political platform. Even in democratic periods, poor performance has often been used as an excuse to suspend elected city councils when a political impasse is reached, e.g., in Dar es Salaam in 1996. In the 1990s, as elsewhere in the world, it was common for the majority party in city councils to be an opposition party, leading to political tensions with the national government and failure to provide a satisfactory financial resource base. Consequently, autonomous decision making is limited and, in part as a result, it is difficult to attract good-quality candidates to stand for political office at the municipal level.

The economic efficiency argument for devolution rests on an assumption that local spending decisions will result in more cost-effective provision of services, especially when the financial resources are locally generated and there is local accountability. In practice, central government's motives may have more to do with reducing demands on national revenue. Moreover, the greater efficiency (or pro-poor) allocation of resources by local government remains a hypothesis: the extent to which it is realized will depend on the local political system and the capacity of the administrative arm of local government to deliver.²⁵ The limited administrative and financial capacity of African municipalities in a situation of rapid urban growth and the disruptive effect of successive reform attempts has resulted in severe infrastructure and service deficiencies, particularly for poor residents.

The second key aspect of the formal urban political system is the electoral arrangements. Even within the general framework of representative democracy, various systems are possible, featuring different arrangements for the election of representatives, executive control, the role of parties, and arrangements for city-wide and sub-city-level decision making and administration.

Election of representatives can be ward-based or at-large; first-past-the-post, proportional representation, or a mixture of the two. In a ward-based system, councillors are more likely to identify with and represent the interests of their constituents.²⁶ However, this is by no means assured. Their motives for seeking political office may be power and status, to advance a political career or to personally benefit from access to public-sector resources. The latter may not involve grand corruption, but derive from privileged access to information (on property development, infrastructure routes, contracts), allowances/salaries, or patronage resources such as municipal jobs, public housing, or trade licenses. Urban wards are likely to contain a mixture of income groups and residential and economic interests, and even when they are predominantly poor, councillors themselves may not share the socioeconomic characteristics of voters, and levels of voter turnout may be low. In addition, ward councillors are likely to advance particularistic arguments, reducing the attention given to strategic issues and city-wide redistribution needs. Ward-based elections tend to be associated with a first-past-the-post system, in which many representatives may be elected with only minority support, while other minorities are unrepresented in the political system. While ward-based systems predominate in anglophone countries, party list systems are more common elsewhere. Closed party list systems reduce the choice available to voters.²⁷ Difficulties arising from the single list electoral system led some of the ten municipalities in Abidjan to introduce neighborhood management councils or committees to ensure more participatory management of local services.²⁸ Elections at-large, based on a proportional representation system, may reduce the extent to which councillors identify with the interests of their constituents but also make it more likely that strategic objectives can be addressed and minorities and small parties adequately represented. In terms of the design of the electoral system, therefore, some of the disadvantages of majoritanian competitive democracy can be overcome by combining first-past-the-post and proportional representation elements, as in the design of the new electoral system for Johannesburg.²⁹

Executive control. There is relatively little variation in the structure and roles of legislatures at city level, but considerable differences in the way executive bodies are structured and their functions performed, with implications for representativeness, accountability, and effectiveness. These differences relate to whether the executive is single (a mayor) or plural (a committee or cabinet), whether it is elected or appointed, whether it has political or managerial powers and how clearly these are separated, and the system for appointing senior staff and rules on terms of office.³⁰ For example, following the political settlement in Mozambique in 1994, a municipalities law was passed which provides for an elected administrator and local assembly, from which a municipal council is appointed by the administrator. Subsequent research detected concerns about whether elected administrators could successfully combine executive and legislative roles, especially given the inherited experience of a coercive and militarized state.³¹ Central appointment of the executive reduces autonomy and accountability to the local legislature. This is seen clearly in Ghanaian cities, especially Kumasi,³² although it may result in benefits if the appointee is able to use his or her position to secure a larger flow of central funds than might otherwise be available. An executive (and councillors) with managerial powers tends to become closely engaged in day-to-day operations, opening the door to political favoritism, inefficiency, and corruption. If senior staff are centrally appointed, as in Kenya,³³accountability to local councillors is reduced. If locally appointed, they may be unable to uphold bureaucratic norms against dubious political practices and less competent because of the lack of a good career structure, although they may also be more committed to local development objectives. Rules that restrict elected representatives to a single term may exacerbate the tendency to favor politically visible short-term projects over long-term solutions to problems.

The role of parties. Democratic theory suggests that individual interests can be aggregated and represented most adequately through a party system. Political parties, it is argued, sharpen accountability by presenting voters with policy choices, which on election representatives have a remit and responsibility to implement. However, David Pasteur concludes that in practice there is no obvious correlation between multiparty competition and the responsiveness and integrity of urban management in developing countries.³⁴ Partisan interests may well be put before either overall city development objectives or the needs of supporters of other parties. Moreover, it is rare for political parties in Africa to have a coherent policy platform or to aggregate and articulate local interests, in part because they lack funds and also because they are not themselves democratic institutions. Pluralism may instead encourage destructive competition between factions, levels of government, and neighboring jurisdictions, without giving voters policy choices. The strategies parties pursue to widen their political base include cooption or camouflaging party political activity as grassroots (community-level) organization. Moreover, parties do not command loyalty, because they are seen merely as a means to attaining power. Politicians often change parties to improve their electoral chances or gain access to the resources of the ruling party. Also parties and politicians have to raise funds for campaigning and patronage, so they are obligated to both their sponsors and supporters, resulting in clientelistic and particularistic practices. All of these features are seen in, for example, Kenya's political system.

City-wide and local decision making and implementation. City governments with comprehensive boundaries and functions and a strong financial base are probably best placed to manage urban growth.³⁵ However, on the one hand, metropolitan governments are often administratively fragmented and, on the other, city-level government is remote from residents, adversely affecting access and accountability, unless there is a ward-based electoral system. Possible solutions to the former problem (amalgamation of adjacent municipalities, extension of the core municipality, or directly elected metropolitan government) may face political resistance. Alternative solutions include voluntary cooperation between municipalities, the establishment of metropolitan bodies for specific functions, or an indirectly elected metropolitan body. A possible solution to the latter problem is the establishment of sub-city political and administrative structures. In Abidjan, for example, the metropolitan area (current population 3.5 million) was divided into ten municipalities in 1980. Each elects five councillors and a mayor who, together with a metropolitan mayor, are also members of the supra-metropolitan body, Ville & Abidjan. The eleven mayors form an executive committee. This federal system has not, however, solved all the problems of lack of clarity in the allocation and overlapping of responsibilities, reluctance to devolve authority, lack of coordination between local and central government, difficulties in coordinating municipalities with very different levels of wealth, lack of municipal capacity, and weakness of metropolitan leadership.³⁶ A similar system was established in Greater Dakar in 1996, with an additional layer at the city level. In Pikine (population one million), AbdouMaliq Simone identifies severe fiscal shortfalls and conflicting interests within and between the sixteen wards as hindrances to a serious attempt to develop ward action plans and a city strategic development plan. The lack of legitimacy which results from arbitrary administrative boundaries and failure to deliver services gives rise, he suggests, to a sense of disconnection between residents and the city administration.³⁷

The ability of elected sub-municipal governments to represent and respond to the interests of residents depends on their relationships with municipal councillors as well as the resources available to them. For example, in Kumasi, sub-metro assemblies, town councils, and unit committees have never become effective, largely because the resources made available to them by the chief executive and municipal assembly have been insufficient.³⁸ In both authoritarian and some ostensibly democratic systems, sub-municipal structures may have administrative responsibilities but in practice be merely tools for political control from above. Kenya's chiefs and sub-chiefs, who are appointed by the administrative/ security apparatus of the central government, are a case in point.³⁹ In contrast, local representative organizations (and NG0s) may be outside the hierarchical structure of government or party but nevertheless recognized for consultation purposes (e.g., Bulawayo Residents' Associations⁴⁰).

A third important aspect of urban politics is accountability.⁴¹ In theory, a democratic system, through periodic elections and other arrangements, ensures both external and internal accountability. Elected municipal government is, in theory, open to the public gaze: it is relatively accessible compared to national government, local politicians are relatively active and numerous, there are likely to be local media, and residents have political rights. This does not mean that those with the greatest influence are the poor, and even if they are, the need to ensure electoral support does not ensure integrity. Mechanisms to enhance external accountability may include executive, judicial, and legislative controls, central government regulation, the exercise of political voice, and independent scrutiny bodies.

The first of these mechanisms implies that the bureaucratic and political executive are separate. However, in practice, politicians and officials exchange mutual favors, and the ability of the legislature to ensure accountability is limited because of its political weakness. In addition, in British-model systems, because the elected council has executive functions, it is not an independent accountability mechanism, especially if it is dominated by a single party. Where there is a strong opposition party, it can encourage accountability, but these rarely exist in Africa's fragmented and unstable party systems.⁴² According to Dele Olowu, the effectiveness of judicial remedy is only medium-low in Africa, while independent scrutiny bodies are rare.⁴³ Central government has control and regulatory powers but is ineffective in holding local government to account when local politicians and the ruling party or employees in the field offices of central government have forged an alliance for mutual benefit.

In theory, the electoral system allows voters to register their general approval or disapproval of an incumbent's performance and to select new representatives. However, electoral politics may only legitimize the power of the local elite. Middle-class residents tend to have interests in common with the bureaucracy and either manage to secure services for themselves or insulate themselves by selfprovision or the purchase of private services (from gated communities and private health care to individual septic tanks and generators). They do not, therefore, try to hold local government to account for its failure to tackle environmental health problems. In addition, elections are occasional, only address the broadest issues, and the system is affected by the ratio of citizens to councillors, the timing of municipal vis-a-vis national elections, and the voter registration process. Under-representation is common in the largest cities (either because the redrawing of boundaries lags behind population growth or because the ruling party is reluctant to provide opposition supporters with greater representation) and where voter registration is voluntary. Furthermore, timing municipal to coincide with national elections reduces them to a sideshow and may rob them of the most able candidates.

Democratic pragmatists suggest that limitations on the accountability provided by the electoral system can be offset by the use of public meetings or hearings, opinion surveys on the extent of citizen satisfaction with municipal services, as well as a watchdog role for the media and civil society organizations. Although greater freedom of expression accompanied democratization and has increased the role of the media in improving accountability in the 1990s, investigative journalism is still hindered by intimidation, repression, and lack of resources. There is much ambivalence about the political salience of NGOs, grassroots and other membership organizations. Seen by liberal democracy theorists as a means of increasing accountability, others regard them as, in effect, a parallel administration: a para-political structure linking the official political apparatus to residents.⁴⁴ The functions and motives of NGOs and the political space available to them vary, influencing their political behavior. Some undertake developmental activities or provide services, typically to residents underserved by public agencies. Others act as intermediaries between residents and public-sector agencies or create structures to coordinate unorganized interests and increase their political influence at the city or community level. Sometimes NGOs develop positive relationships with residents, politicians, and local government, as in the Nairobi Informal Settlements Coordinating Committee. However, governments often regard them as a political or security threat, or view them merely as useful instruments to advance government aims or counter public-sector deficiencies.⁴⁵ The former view leads to NGOs being ignored, regulated, or oppressed; the latter to corruption or establishment of governmentoriented NGOs. NGOs may support communities or they may encourage dependent relationships. They are accountable only to their boards, or perhaps to donors. This accountability gap results in failure to monitor their own performance and even to fraud, as seen in some particularly dysfunctional "brief-case" NGOs.

Grassroots membership organizations are most widespread in residential communities of the poor, although where they have been formed in high-income areas they can be important in both the provision of services (especially security) and in the exercise of political voice (e.g., the Sandton Residents Association whose rates boycott has created a financial crisis for Johannesburg Metro Council⁵⁰). Most of the social groups important to the lives of residents are informal associations, only some of which are engaged in political activity. Many are ethnically based, including cultural and burial groups and hometown associations. In Tanzania, hometown associations, previously banned as a potential source of ethnic divisiveness, have been tolerated again since the 1980s.⁵¹ In Lagos, landlords manipulated their ties with both patrons and clients so astutely that their leaders came to be recognized as traditional chiefs.⁵² Landlord and tenant associations for matters (such as service provision) where the interests of these two groups do not conflict are important in some neighborhoods.⁵³ In the 1980s, new voluntary associations emerged, partly to cope with economic and governmental crisis and partly out of a concern with the wider interests of society. In Tanzania, the new groups included local defense teams (sungusungu) which formed from about 1987 onward to deal with the absence of state-provided security, and were formalized in Dar es Salaam in 1990, leading to an immediate drop in crime rates. Women's associations are among the fastest growing new associations. Most organizations emphasize self-reliance and the development of solutions to everyday problems.⁵⁴

However, civil society organizations are generally only weakly developed, often emerge to compensate for state failure, and do not have a clear role in enhancing the accountability of the formal political system in most African cities.

Both earlier and recent attempts to base urban political systems on liberal representative democracy have, therefore, encountered problems. As at the national level, urban political systems are typically fragile and unconsolidated. However, the democratic systems under consideration are not alone in this: the authoritarian political regimes of the 1970s and 1980s also experienced problems. Despite some early achievements, lack of accountability and economic crisis ultimately undermined those in power at both national and city levels. Extensive patronage systems favored the better off over the poor and some ethnic groups over others. As local government's efficacy and legitimacy declined, residents without access to patron-client networks increasingly adopted the political strategies of exit (as seen in the growth of mutual support organizations noted above), noncompliance, or "agitational practices.”⁵⁵ In Dar es Salaam, for example, widespread noncompliance with government regulations on construction and informal-sector activity have led to political responses, notably partial retreats by the government from many of its past heavy-handed interventions.⁵⁶ Mamadou Diouf describes young people in Dakar as having abandoned both government-constituted youth organizations and party politics for confrontations with the state, parents, and educators, targeting symbols of the state in election-related riots, or attempting to reclaim authority over urban districts by establishing militias, engaging in punitive expeditions against addicts, drunks, and thieves, challenging service providers, and organizing neighborhood cleanup operations.⁵⁷ Whether the prevalence of patronage, exit, and noncompliance despite the formal trappings of democracy constitutes chaos will be further discussed in the next section. First, however, the experience of formal land administration will be briefly analyzed.

As noted above, the form of tenure considered suitable for urban areas is individual, based on the issue of formal title, and intermediate rights of occupancy are considered inferior or incomplete. To administer a system of formal title, a full cadastre and efficient registration system are needed as a minimum, resting on the one hand on recognition of individual ownership rights and on the other on technical requirements for mapping, plot survey, registration, and conveyancing. Shortages of qualified professionals and finance, coupled with rapid urban growth, mean that mapping has lagged further and further behind urban expansion, and the cumbersome procedures and centralized administration for registration and transfer have been unable to respond to increased demand. Moreover, on the outskirts of many urban areas, the claims of indigenous residents to land were regulated under customary tenure systems. Even if these systems are recognized in law, the relationship between them and formal land administration in the urban context has often been problematic. Generally, either the contradictions have not been resolved or formal land law supposedly takes precedence, a position which those with indigenous claims are not prepared to accept.

In addition, urban planning systems based on the preparation of comprehensive land use plans implemented by means of public investment in infrastructure and development, accompanied by detailed development and building control over private subdivision and construction, have proved unable to cope with the demands of rapid urban growth. Limited professional and financial resources have restricted investment in new infrastructure, and insufficient resources for and attention to maintenance have resulted in the rapid deterioration of those roads and utilities that were installed. During the heyday of state-led development, a desire to control urban development and, on the part of some leaders, to restrict speculation and profiteering or modernize land administration, led to land nationalization. However, the state machinery, predictably, lacked the capacity to make the land in its ownership available for development sufficiently rapidly to keep pace with demand, or to approve transactions between private leaseholders. Instead, public land subdivision, allocation, and regulation provided extensive opportunities for rent seeking and patronage, produced bottlenecks in the supply of land for urban use, and resulted in extensive evasion. Reversal of this legislation in the 1990s in most countries has produced few improvements.

The result is that the formal land administration system has only ever succeeded in meeting the demands of a few — typically between a third and half of the urban population.⁵⁸ In practice, much land is occupied without formal title, much development takes place in areas which are not designated for urban use in development plans (or is mixed use in areas zoned for sole residential or other use), and the majority of residents live in areas which are inadequately served or unserved by road access and publicly provided water, sanitation, electricity, and social facilities.

Attempts to impose order on political practice and urban development, to regulate the activities of private enterprises and citizens, and to outlaw systems and practices regarded as unsuitable and undesirable have, therefore, largely failed. Patron-clientelistic politics, exit, and noncompliance with regulatory requirements are prevalent. The question is whether the current situation in African cities is correctly characterized as chaotic, or whether systems of social rule are operating which are different from those of the formal systems but which are either more effective or command greater legitimacy.

As noted above, liberal democratic political systems are based on an individualistic exchange view of politics⁵⁹ and political practices which do not fit into formal systems of electoral representation and decision making are often regarded as undesirable and disorderly. However, there is another view of politics which regards political actors as culturally dependent and socially constructed, and political action, therefore, as embedded in social rules and institutions. Such institutions define the behavioral and social bases of the terms of political exchange and provide a framework of roles, identities, and rules (of which self-interested calculation is only one). In practice, some argue that in the African context, many of the formal institutions of governance, including (but not confined to) a liberal democratic political system, are not rooted in local culture: the shared meanings, representations, and values underlying the social system, and expressed in social relationships and practices.

The disconnect, Mamadou Dia suggests, is seen in political institutions that lack moral and political legitimacy. Indigenous institutions, on the contrary, he argues, are anchored in local culture and values, and so are legitimate and enforceable.⁶⁰ Dia suggests that traditional government in Africa was hierarchical but broadly representative and that most societies were governed by consensus: rulers had authority but shared power. They were often selected by a specific body, whose choice then had to be approved; failure of rulers to discharge their functions led to retribution, usually deposition. The abuse of power was checked by rotating office, imposing fixed terms, or creating positions for individuals who would not be removable by the leader. The checks and balances operating in the system were expressed in moral concepts and axioms. Decision making was generally consensual rather than authoritarian and, although the acquisition of wealth by leaders was considered legitimate, they were expected to share this wealth by guaranteeing the welfare of their followers and by assisting those who were in need. The central values were, therefore, sharing (because of the need for security), deference to rank, the sanctity of group commitment, a regard for compromise and consensus (win-win solutions), and economic, social, and personal relations based on trust. Dia's is an essentially positive view although he admits that traditional institutions can harbor undesirable practices, such as discrimination on the basis of gender and age, and may find it difficult to change to accommodate present-day demands (including, I would add, those associated with urbanization). However, others have a less positive view, pointing to the links between chiefly structures and patronage systems of politics and the practice of calling on "culture" to justify political practices in which ethnic rivalry is deliberately mobilized as a means of retaining or gaining power.

The result of colonial ideologies which encouraged Africans to view government as a source of benefits is that now, Francis Enemuo suggests, there are two public spheres: an amoral civic public sphere from which benefits are expected but which is unimportant in the definition of duties, and a moral primordial public sphere, defined in terms of an ethnic group, within which relationships are seen in terms of duties.⁶¹ The result is a failure to pay taxes and a disinclination to acknowledge the duty of citizens to create enduring and legitimate political institutions. Both Dia's and Enemuo's views are oversimplified: not only were precolonial authority and state structures varied and often contested, ethnicity was also (re)constructed by the colonial powers. Nevertheless, in many countries, traditional organizational and political systems do provide systems of social rule which continue to have moral and political authority. Some, sharing Dia's positive view, consider that they could be a resource for social stability in a mixed government system without abandoning democratic reforms,⁶² while others see them as a hindrance, encouraging resource allocation on an ethnic basis and allowing the state to twist traditional institutions to its own purpose.⁶³ Moreover, consideration of alternative sources of social rules for associational and political life should not be confined to the supposedly "traditional" bases for social organization.

Potential roles for traditional authorities and other forms of social organization in urban governance can be considered at both city and sub-city levels. Akin Mabogunje, for example, advocates that traditional associations such as neighborhoods or urban quarters and their chiefs be integrated into urban management.⁶⁴ The 1994 municipalities law in Mozambique provides for a limited role for the traditional authorities, which were not recognized by the previous Frelimo regime but had been used by Renamo as a basis for a rudimentary administrative system. However, views on this role are mixed, because of conflicts between chiefs and their failure to always represent the wider community.⁶⁵

All political experience introduces participants to rules for engagement. Even if systems change, therefore, the legacy of earlier systems both influences subsequent political practice and provides a potential basis for political and social relationships. This is illustrated in the African urban context by the experience of the one-party "democracies" of the 1970s (Zambia, Mozambique, Tanzania), where grassroots organization was associated with the attempts of the ruling party to secure support and exert control. That is not to say that residents did not benefit from the hierarchical party structures, which were used for delivery of infrastructure improvements in many of the informal settlement upgrading projects of the time and which often included separate organizations for "youth" and women. Nor does it mean that there was no scope for participation, although they tended to be fairly autocratic, top-down organizations, and groups that refused to join political organizations (such as the Jehovah's Witnesses) were often penalized. With the reintroduction of multiparty democracy, the discredited local organization of the former ruling party could no longer claim it represented communities as a whole. The extent to which elements of the system persist, constituting a store of social and political capital, has not been systematically researched. Ann Schlyter shows that in George, an informal settlement in Lusaka, eighteen months after the transition to multiparty democracy, the idea of the ruling party as the sole community organization had persisted. MMD leaders in the area regarded themselves as community leaders, even though no functional local organizational hierarchy similar to that previously maintained by UNIP had been established and they had lost the role played by UNIP leaders in dispute resolution. In addition, women's space for political engagement had narrowed.⁶⁶ Elsewhere, however, there is some evidence that the structures and leaders left over from the one-party era still form an important part (officially or otherwise) of the administrative system (e.g., in Mozambique and Tanzania) and still have legitimacy in the eyes of the population.⁶⁷

The political rules of patron-clientelism are today well understood in Africa, even if accepted by beneficiaries and resented by those sidelined in equal measure. Moreover, there is a legacy of institutions, strong in some countries (e.g., Ghana) and weak in others (e.g., Kenya), inherited from traditional authority and kinship systems. In addition, successive and specific urban political and social organizational experience has given rise to new bases for associational and political activity. In this context, transition and complexity would be more appropriate terms than chaos to describe the evolving social and political relationships and practices of African cities.

Inherited formal land administration systems and their successors are, in theory, governed by formal rules (those embedded in statutes and regulations). "Traditional" tenure systems, on the other hand, are regulated by "custom" (a series of embedded but fluid social norms and practices which constitute generally unwritten rules).

Donald Krueckeberg, in his critique of three so-called creation stories or myths about the nature of land ownership in Africa, notes that the portrayal of African ownership as communal served the interests of both the colonial powers (who used it to reinforce/establish authority structures beholden to them, i.e., chiefs) and Africans (to protect their rights against further expropriation or taxation rather than as a description of actual rights). In addition, although such tenure might express African cultural values, it also suppresses the rights of some, especially women. In practice, much land is individually acquired and held, and the process of adjusting the rules governing customary tenure to the process of buying and selling land began early in the 20th century, despite the hostility or indifference of the colonial government.⁶⁹

In urban areas, land transactions and conflicts are structured by hybrid institutions which are neither strictly formal nor customary and will be referred to here as informal. In contrast to the failures of the formal land administration system discussed above, processes of informal land development have been successful in delivering large quantities of land for urban residential development (often between 50 and 70 percent).⁷⁰ This success can be attributed to the social legitimacy of these processes, but the institutions that regulate transactions in and the use of land (including trust) come under pressure during the process of urban development. The systems of formal, "customary," and informal rules exist in parallel, although views about the relationship between them vary. Some see them as conflictual. Alternatively, in an interactive or synergistic view, elements of each institutional form are used by actors as appropriate to minimize transaction costs. Elements of each are combined in the notion of societal noncompliance, which involves synergy, conflict, domination, and manipulation in the relations between actors.⁷¹ "[It] is considered a form of 'protest,' albeit with its own 'rules of the game,' which may be used to contradict, pre-empt or control those of the state, but which also observes and upholds other state rules."⁷² It exploits contradictions and areas of ambiguity. Of interest here are the characteristics of social, economic, and political relations between the actors involved; the relative functions of formal, informal, and customary rules in rendering acceptable the institutions governing transactions; and the circumstances in which these institutions break down. The discussion will be illustrated by research findings from Tanzania, Lesotho, and South Africa.

Dar es Salaam. In Dar es Salaam, the formal land administration system has had great difficulty coping with the demand for urban land and services, and today over 70 percent of residents are accommodated in informal settlements.⁷³ Before independence, land transactions in the rural areas beyond the urban boundary were based on verbal agreements between buyers and sellers, often witnessed by adjacent owners. Where land is still held by indigenous residents under customary or pseudo-customary law, intensification of the demand for subdivision is weakening the strength of customary practices in land (especially where economic pressure or family crisis forces sale). Thus, while land sales are widespread, because there is some social stigma attached to selling family land, it is done discretely (meaning quietly without publicity, not one by one).

Beginning in the 1960s, the informal market devised its own quasi-legal procedures to authenticate ownership, transfer, and sale.⁷⁴ Land transactions were, and still are, authenticated by local politicians (elected Ten Cell and sub-ward or Mtaa leaders), despite such transactions being contrary to government policy, which prohibited transactions in unimproved land. Individual rights, which are ceded in perpetuity, are certified and the use of written sale agreements has gradually become common practice. Transactions are given greater legitimacy by the use of local witnesses, who can be relatives, close associates of the transacting parties, or neighbors. In the 1980s, with expansion of the city and further development of land markets, more robust instruments for enhancing individual rights in property were developed, including registration. In certain areas, local political leaders have, with the agreement of local residents, established a set of mandatory procedures that have to be followed by all parties. In some, administrative fees are levied by sub-ward offices and honoraria paid to witnesses. However, leaders in other areas doubt their ability and right to enforce the norms necessary to safeguard interests in land. In Rangi Tatu, one pen-urban study area, 40 percent of owners had informal sale agreements authenticating their purchases, less than 10 percent had acquired land through the formal process, and nearly 50 percent did not have evidence to support their ownership but were increasingly trying to regularize it through the informal system, in which they placed considerable confidence. In addition to rights and transfers being authenticated at the local level, disputes are often arbitrated and resolved where possible at the sub-ward level, by a committee of elected local elders, and are only referred to the courts when the elders are unable to resolve the dispute.⁷⁵ Leaders also sometimes take the initiative to arrange housing plots in an orderly way, following existing features (e.g., tracks), allowing for access and circulation, and designating sites for public uses (school, market, cemetery). Their concepts of order, which seem to be acceptable to residents, are based on facilitating the routine activities of residents, providing security, and supporting social and cultural values. Efforts to provide services generally come last — they tend to be sporadic, sponsored by local leaders or private individuals, but often with community committees, e.g., construction of a primary school or improvements to water supply. Higher levels of the local administration (e.g., the ward) only occasionally get involved in providing social infrastructure. The shortcomings of this approach are evident: it results in uncoordinated and patchy provision, and becomes increasingly difficult as market pressures on land increase, especially where local organizations have poor links to the local authority.

As urbanization proceeds, the social recognition of land rights by leaders is vulnerable to the erosion of trust, pressure on land, and the increased power of the market. Research found that, while hybrid informal/formal institutions successfully regulate transactions in newly developing or partly consolidated residential settlements, they are unable to cope with the pressures in a consolidated inner city informal settlement.⁷⁶ In Keko Mwanga, access tracks were being encroached upon, there was no land for social facilities, and problems of trespass and multiple sales of the same plot had increased. Sub-ward leaders reported that, since the introduction of multiparty politics, they had been unable to restrain building activities, but this may also be because sub-ward leaders in this area are associated with the ruling party. In this area, Ten Cell leaders were no longer able to resolve disputes between neighbors and the aggrieved parties resorted to the formal legal system. Despite litigants' lack of formal title, in settling such disputes, state courts take into account only the merits of individual cases, without challenging or questioning the legal bases of transactions or the rights of disputing agents to transact in land.⁷⁷ Therefore

Maseru. A significant proportion of the recent population increase in Maseru has occurred in peri-urban settlements outside the formal urban boundaries, where agricultural land is privately subdivided for sale under the authority of the traditional leaders, the chiefs. The result is unplanned urban sprawl, with initially large (1,000 square meters plus), irregularly shaped plots which are inefficient to service. The Land Act of 1979 attempted to regularize the situation by nationalizing land and distinguishing between land leased from the state and rural arable land to which use rights were to be allocated in perpetuity by "democratically elected" land allocation committees. The Act also includes provision for the designation of Selected Development Areas to allow the acquisition of land within urban boundaries for new areas of urban use or upgrading of unplanned residential areas. The effect of the SDA declaration is to cancel existing rights and interests in land without compensation, pending the issue of leaseholds. The Act failed to streamline land delivery, reduce the loss of agricultural land to urban development, or promote orderly" urban growth, because of various forms of subversion by the state itself, local chiefs, and owners of use rights to peri-urban cropland (field owners). The problem is defined officially not as inappropriate law but as the lawlessness of chiefs and their subjects.

Because of the difficulties faced, especially by middle- and low-income people, in accessing land through the formal system, for the majority the only feasible option has been to purchase directly from field owners. Whereas by custom, chiefs allocate land freely to their immediate subjects and to a small number of incomers from other areas, increasingly they are dealing with requests from unknown people. Under these conditions, a "moral economy" of custom, where access to free land was guaranteed by allegiance to a chief, has been supplanted to some degree by an "amoral economy" of impersonal market exchange. Popular opinion is that individual field owners are encouraged by their chiefs to subdivide and sell their agricultural landholdings or face state appropriation without compensation. In turn, chiefs issue certificates of allocation to plot buyers (so-called Form Cs, backdated to before June 1980, when the 1979 Land Act came into force). While in the past, chiefs would have been expected to undertake this task free, today they charge a fee for issuing such certificates. Such land subdivision aims to avoid direct confrontation with state law enforcement machinery by occurring at night, over the weekend, or on holidays, copying official land parceling practice, and involving public officials in the process (buying and selling land, advising the chiefs and replenishing their supply of Form Cs, or advising prospective land buyers in return for cash "for lunch").

The role of chiefs and field owners in subdividing and selling land for urban development has been recognized de facto if not de jure, in part because of specific instances of noncompliance. Following the acquisition of land without compensation at Khutetsoana, on the outskirts of Maseru, for a World Bank— funded sites-and-services scheme, field owners in the adjacent area of Ha Mabote, who under the 1979 Act had only licenses issued by the state for agricultural land within urban boundaries, rushed to sell their land in case it was also expropriated without compensation.⁷⁹ The government moved in 1984 to declare Ha Mabote an SDA before it could be completely subdivided and developed. However, not only did subdivision continue apace, with the backing of the chiefs, but the latter also succeeded in preventing the demolition of "illegal" structures. With the connivance of officials, provided routes reserved for roads were respected, chiefs urged continued subdivision, while negotiating their acceptance of the project in return for recognition of existing use rights. Eventually (in 1986) they succeeded in getting the provision for licenses repealed and compensation payable for all land expropriated by the state. By exploiting their political position and inconsistencies in the land law and administration, in this instance chiefs were able to derive benefits for themselves, their subjects (the relatively low income field owners), state employees, and the purchasers of subdivided agricultural land, while enabling government to produce an urban extension broadly in accordance with the intended layout.⁸⁰

Maseru therefore illustrates not only the operation of social rules but also how noncompliance as a political relationship between state and society has served to achieve the political objectives of reinforcing the efficacy and position of chiefs, to the greater satisfaction of many citizens than formal state decision making and land administration. The Dar es Salaam and Maseru examples illustrate the close links between urban politics and the rules governing land administration, and these are demonstrated even more graphically in our final case study of an area on the outskirts of Durban, which illustrates how the trajectory of an area of informal settlement and the well-being of its residents are determined by its changing relationships with the wider political system.⁸¹

Durban. The story of Besters Camp in Durban illustrates the interrelationship between informal rules for land development and electoral politics and also the way in which rules change over time.⁸² The early settlement in this area on the far northern outskirts of the city consisted of illegal subdivision ("shackfarming") by legal Indian landowners. In the mid-1970s, Inkatha was reconstituted and began to assert political control by establishing committees in the area. In the second half of the 1980s, as the struggle between Inkatha and the UDF/COSATU heightened, there were widespread disturbances. Violence against the Indian owners caused them to flee, permitting Inkatha to assert control (and seize vacated property), with the state turning a blind eye to what was going on. For example, in the unsettled area of Ekutholeni, a series of invasions were organized by self-proclaimed Inkatha "councillors." Settlers were required to pay an Inkatha membership fee, a token payment for the site, and a fee to the warlord's functionaries who demarcated the plot. The warlord in this area established four wards, each with a chairman appointed by him. This formed the territorial base from which he controlled the politics of the area, using force, blackmail, and protection rackets to control land ownership and subdivision, businesses, water supply, taxis, and the social activities of residents. The warlords and their men acted as both police and judiciary in the absence of an effective state presence. In the adjacent areas of illegal subdivision, the warlords organized infill development, with non-Zulu residents sometimes being forced out. Because Inkatha was able to maintain control during these turbulent times, a symbiotic relationship between the movement and the national and Kwa Zulu governments developed, although not without conflict and tension, in return for which basic services (e.g., tanker water) were provided.

By the 1990s, Inkatha control was undermined by its reliance on force and inability to provide services. Residents switched their allegiance in the hope that the UDF/ANC would deliver services, with the result that some warlords were driven out and others changed their allegiance, even adopting leadership roles in the new regime. This period was marked by greater autonomy from the state, but also by a period of organizational turmoil marked by power struggles between criminal gangs, "youth," and women's groups. Excluded from mainstream civic politics and emerging from the control of warlords who had suppressed other forms of social organization, new forms of order developed, but these depended on the arbitrary exercise of power by youth and so-called people's courts. The emerging ANC and civic leadership therefore had to both assert control and try to deliver services.

In the 1990s, the area was the site of the first major upgrading project in South Africa, initiated by a private-sector NGO with city and IDT funding. This was made possible by the new legal framework, and adopted an innovative organizational approach in which there was a high degree of resident participation. A Community Development Trust (with resident representation) was established as the organization in charge, responsible for installing infrastructure and issuing freehold titles. However, some unregulated buying and selling of surveyed sites has continued, so it is expected that differences between registered and de facto ownership will emerge in time. With the CDT, elected councillors, and Ward Development Forums similar to those elsewhere in the city, the area is now integrated into the local political structures, but there are pressures because the settlement leaders are still dependent on political patronage, and power struggles over access to external resources have increased. Informal organization, especially the women-dominated Health and Welfare Committee, continues to, be strong, reaching down to the household level. "Strong men," often ex-Inkatha leaders who have kept control over the key economic enterprises in the area, have benefited from the increased cash circulation which resulted from upgrading, especially the labor-based construction contracts which they won, and on which they were able to employ their followers.

External political backing, with the support of these powerful informal networks, was important to ensure that the CDT and WDFs were able to, secure upgrading, from which the informal networks benefited in turn. The settlement is now legally incorporated into the land administration system and urben management structures, but the informal internal sociopolitical fabric remains,, in parallel with and often in support of the legally constituted structures, Political legitimation has not brought notable changes to the internal social fabric, leading to ongoing challenges to integrate formal/ informal and de facto / de jure aspects of politics and land administration.

Recognition of the ability of informal urban land development processes to deliver large quantities of land relatively cheaply to meet urban demand does not deny their disadvantages: informal settlements are characterized by environmental degradation, public health hazards, and the use of unsuitable land for urban development. They give rise to water pollution, encroachment on public land, and additional costs of retrofitting services, especially when densities reach a high level. They cannot necessarily guarantee security of occupancy, protect against fraud, or resolve all disputes. Nevertheless, they result in residential development which, while far from orderly in conventional terms, is not chaotic.

In reaction to fears of uncontrolled urban growth and the political mobilization of urban mobs, as well as a desire to secure urban environments conducive to efficient enterprise and public health and responsive and accountable political systems, colonial and postcolonial regimes have tried to institute systems of political representation and land administration capable of dealing with the pressures of urban development. However, the consolidation of political systems based on conventional theories of liberal democracy has proved problematic in Africa at both the national and urban levels. Furthermore, formal land administration systems designed to protect individual property rights and regulate urban development have benefited fewer than half of all urban residents and have failed to cope with rapid demographic growth. This paper has argued that we need, therefore, to question the concepts of order and the formal rules on which these systems are based. Their colonial origins, theoretical assumptions, and specific organizational design together make them, to a considerable degree, inappropriate and unrealistic in contemporary Africa.

In practice, much political practice consists of patronage, exit, or noncompliance, and much land development is informal, in the sense that it does not comply in one or more respects with formal state law and regulations. These practices are in some respects disorderly, but they are far from chaotic. While the social rules of authoritarian, clientelist spoils politics are widely understood, they cannot be said to have a high degree of legitimacy or to be based on high levels of trust. In other respects, however, political and land development practices are governed by systems of social rules that are widely accepted. These are not drawn from an idealized and unchanging inheritance of traditional authority, social organization, and kinship. Although all of these institutions may be important, to a greater or lesser degree, in different societies, they have evolved in response to a long series of colonial and postcolonial changes, have not always successfully adapted to the demands of contemporary societies, and can be oppressive and exclusive. In addition to these sources of institutions, the urban experience of recent decades, especially since the postindependence state-led development model has been discredited, has spawned a variety of new associational forms. While most are ostensibly nonpolitical, many have gained legitimacy because of their success in supporting their members during periods of government incapacity, and some have challenged the political order or made political gains through practices of widespread noncompliance.

This is not to say that formal political and administrative systems should be abandoned: informal patronage practices have wasteful and inequitable outcomes, exit and noncompliance are essentially negative, and unplanned land development gives rise to insecurity and poor living environments. Moreover, even such informal institutions are not divorced from political and administrative systems based on formal rules. During periods of both democratic and authoritarian rule, informal practices have borrowed from formal rules systems and in some instances modified or transformed them in turn. Many aspects of informal and customary political and land development practices have wide social recognition — it is this familiarity, understanding, and legitimacy, this rootedness, that helps to explain their prevalence and their relative success. However, they also have dysfunctional and conservative aspects, while the formal rules on which liberal democratic political systems, private property rights, and planning and regulatory systems are based are becoming increasingly embedded and accepted. In addition, experiments designed to overcome some of the shortcomings of formal representative politics and land administration systems, such as democratic decentralization,⁸³ participatory and deliberative democracy,⁸⁴ regularization of informal tenure, and reforms of planning and building codes, are under way.

Future research needs to develop a better understanding of the characteristics of informal political practices, property rights, and land development processes, of the relationships between formal and informal systems, and of the outcomes of promising experiments. Such an analysis will enable the strengths and limitations of contemporary political practices, informal tenure and land development processes, and new approaches to be identified. It will provide a framework both for the development of distinctive and locally appropriate rules frameworks and politico-administrative systems and for the transfer and adaptation of solutions from other societies within Africa and beyond. These might include forms of democratic engagement which complement representative democracy and mechanisms of accountability which can crosscut ethnic divisions and give voice to the poor, replacing particularistic patronage politics and majoritarian winner-takes-all elite politics with a more truly social democratic system.⁸⁵ They might also include forms of tenure and development regulation which are manageable in a context of demographic growth and limited resources, provide affordable land for development in appropriate locations, and ensure an acceptable level of security for urban residents and enterprises.⁸⁶

That cities demonstrate multiple conflicts and contradictions (between wealth and poverty, tradition and modernity/postcolonialism, men and women, old and young, power and powerlessness, opportunity and marginalization) is not denied. That the attempts of formal urban political and land administration systems to manage these conflicts and contradictions have had serious shortcomings is evident. Clearly, aspects of urban political practices and land development processes are disorderly. However, the continued (if inadequate) functioning of electoral processes, city administrations, formal property rights systems and regulatory procedures, and the parallel operation of multiple sets of social rules that govern so-called informal political practices and land development processes bely the notion that cities are chaotic and contain promise that more appropriate conceptualizations of political and physical order can lead to improved urban governance arrangements in future.

Carole Rakodi. Order and Disorder in African Cities: Governance, Politics, and Urban Land Development Processes, in: Okwui Enwezor, Carlos Basualdo, Ute Meta Bauer, Susanne Ghez, Sarat Maharaj, Mark Nash, Octavio Zaya (ed.): Créolite and Creolization. Documenta11_Platform3. Hatje Cantz Verlag, Ostfildern-Ruit. 2003, pp. 45-80.

Download "Order and Disorder in African Cities"