Return to Index

"Failed States III:
Globalization and The Failed State"

Florence, Italy
April 7-10, 2000

Failed States and Non-States in the Modern International Order

by Christopher Clapham
Department of Politics & International Relations,
Lancaster University, LA1 4YL, UK
Christopher.Clapham@lancaster.ac.uk

Introduction

The 'failed state' is one of those unsatisfactory categories that is named after what it isn't, rather than what it is. The phrase carries embedded within it the conception of a global order that, at any rate, ought to be composed of states, and suggests that where this conception is not realised, something has gone wrong. That wrongness in turn creates a sense of threat or incompleteness that needs to be rectified: failure is a challenge that calls for a reversal that can in turn be characterised as 'success', and a successful global order is one that is regulated through the norms of statehood.

This paper does not seek to deny that statehood, where it can be achieved, is for the most part desirable; but it does seek to deny that it is always practicable. It argues that states have historically derived from various specific and by no means universally realised conditions; that the global political system has historically comprised areas under the control of states, areas regulated by other forms of governance, and areas with no stable governance at all; that the idea of the state as a universal form of governance itself rested on assumptions that can no longer be met; and that the analysis of international relations now needs to come to terms with the re-emergence of a once familiar kind of global order, and to work out the ways in which this is likely to operate, beyond the comfort zone represented by effective statehood. The alternative project of attempting to restore universal statehood appears to me to be chimerical.

The Rise of Universal Statehood

Look at any map of the world prior to the twentieth century, or indeed any map of Europe prior to the nineteenth, and striking differences from an atlas of the modern world become apparent. Much of the inhabited world, certainly, is covered by states: territories with reasonably clear boundaries, which may be supposed to come under the ultimate control of a single central authority. Much of it comes under other kinds of jurisdiction, such as the Holy Roman Empire, or various forms of 'native' or 'tribal' authority. There is often a great deal of what may be termed optimistic colouring: large areas over which someone or other claims a jurisdiction which they are quite unable to realise. Large areas, too, are just left blank.

States occur in a variety of forms, and under a variety of circumstances. Charles Tilly has analysed the different kinds of state that emerged in Europe from the start of the second millennium onwards, and characterised the different mixtures of coercion and capital that went into their creation and maintenance. States were by no means as common outside Europe as they were within it, but in most regions of the world discernible states existed. Some of these, like the Sokoto Caliphate or the Khanate of the Golden Horde, covered large and often sparsely inhabited areas, and resembled Tilly's extensive or coercive states. Others, such as the emirate of Harar in what is now eastern Ethiopia, maintained intensive control over small areas, and sustained themselves by extracting surpluses from international trade. Others again, like Ethiopia or Thailand or China, rested essentially on the control and exploitation of fairly dense peasant populations, and most closely resembled the territorial nation-states which eventually attained dominance in Europe. The major difference, in very broad terms, between European states and those in most of the rest of the world was that in Europe, different states were generally contiguous with one another, and were therefore intensely concerned to protect their territoriality through the definition and hardening of their frontiers; elsewhere in the world, states were rarely contiguous, and islands of statehood existed within seas of sparsely inhabited and feebly administered territory. Whereas European states defined fixed frontiers (and established their internal identities) by reference to their neighbours and rivals, non-European states were generally constituted by reference to a core, beyond which the central authority was gradually diluted as it extended into peripheral areas subject to its nominal sovereignty or occasional punitive raids.

The idea that the whole world ought to be divided between states exercising full authority over their territories and populations dates only from the era of European colonialism, and in turn, obviously enough, from the creation of a global economy in which statehood was regarded as necessary to secure and regulate access to resources in hitherto inaccessible or uncontrolled areas of the world. In a few cases, such as Thailand or Ethiopia, existing non-European states were strong or diplomatically agile enough to secure their admission to the new universal state system, and generally took advantage of this status in order to extend their control over previously quasi-independent peripheries (a process that created problems of its own). Elsewhere - in all of the Americas and Oceania, almost all of Africa, and roughly half of Asia - European states just took over. This resulted in the creation of a kind of state radically different from any that had existed before: a state derived from a territory whose extent, boundaries and structure of government, and to a large extent whose economy and sometimes even whose population, were imposed from outside. Sometimes, as in Rwanda or Burma, this involved a superimposition of external authority on previously existing units, though even this brought with it a hardening of the state, in terms of fixed frontiers, citizenship rules, and hierarchical structures of government, that could create extremely damaging tensions between the old form of state and the new. Sometimes, as in the Americas, the importing of a new population, and the destruction or complete suppression of the old, meant that states could effectively be transplanted into a new environment. But very often, it amounted to the creation of 'new states', covering areas that had never previously come under the control of any single state, and that in many cases had never previously been controlled by any form of state at all. The 'problem' of 'failed states' is most basically about whether the 'grafting' of such states (in Jean-François Bayart's well-chosen phrase) onto unpromising rootstock can be made to take, even with the various forms of fertiliser provided by the international system, in the form of universalist ideologies, incorporation into the global economy, and the provision of diplomatic and military support.

 

The Statist Ideal and its Decline

The imposition of statehood as a global norm coincided, unsurprisingly, with the heyday of statehood within the European territories from which this new ideal derived. From its sixteenth and seventeenth century origins, in the Reformation, Hobbes and Westphalia, through to its apogee in the 'totalitarian' states of the first half of the twentieth century, the idea developed that there was nothing that states could not do. This conception of the state as an omnipotent and modern organisational technology, greatly enhanced in the nineteenth century by nationalism and industrialisation, led to the development of state ideologies, state economies, and cradle-to-grave state social services. Though fascism certainly ranks as the ultimate form of state ideology, the state economy (in the form of command planning), and the statist organisational technology (in the form, notably, of the Leninist vanguard party) reached their peak, ironically, under the guise of an ideology which ostensibly relegated states to the status of temporary superstructural expressions of economic forces. In the new developmental orthodoxy, this relationship was reversed, and the creation of a modern economy became a dependent variable, conditioned on the prior existence of an effective state.

This idea of statehood readily fostered the assumptions that states provided the only legitimate and acceptable form of rule, and that all of the world's people, and all of its territory, therefore necessarily had to fall under the control of a designated state. It followed that where states did not exist already, in a form which corresponded to European conceptions of statehood, established states were entitled to incorporate the areas and peoples concerned into their own national territories, destroying in the process many hitherto perfectly viable states (along with numerous alternative forms of political organisation) which lacked the military means to defend themselves against European conquest. As colonialism ebbed, so the ideologies, territories and organisations that it left behind provided the basis from which new leaders of new states could promote their own statist pretensions. Anticolonial nationalism neatly combined opposition to the existing colonial rulers, with adoption of the principles of governance derived from those rulers, which in turn enabled newly independent states to gain an accepted and indeed protected place in the international order. The idea that powerful states provided the essential precondition for effective economic development was particularly welcome. With it, too, came the enthusiastic adoption of those devices through which statehood had been consolidated in its European heartland, ranging from national anthems at one end of the scale to central planning and single party states with totalitarian aspirations at the other.

Hitherto dominant ideologies were adapted in order to enable postcolonial states to gain acceptability, and to encourage their leaders to believe that they too could achieve nation-statehood in its most developed form. One example of this, admirably explored by Robert Jackson, was the alteration in the definition of statehood that was required in order to enable almost any designated territory to 'count' as a state for international purposes, regardless of its inability to meet hitherto established criteria for statehood. Indeed, the resolutely statist biases of IR theory can likewise be regarded as providing ideological underpinning for the international acceptability of a large number of entities that were able to describe themselves as states. A related range of adaptations took place in the conception of 'nationhood', which from the nineteenth century had come to be regarded as an inherent component of statehood. Whereas earlier 'primordialist' ideologies of nationalism had claimed that nations derived from deep underlying cultural and linguistic identities (thus legitimising the emergence of new states in Europe, from the 'unification' of Germany and Italy in the mid-nineteenth century, through to the carve-up of central and eastern Europe in 1919), new 'constructivist' theories argued that 'nations' could be created from almost any collection of available social materials through the purposive use of state power. The Stalinist theory of nationalities, which argued that different cultural collectivities could continue to co-exist, without in any way derogating from the power of the state, also found some adherents in states where unitary nationhood appeared to be beyond the range of plausibility.

The significance of decolonisation continues to be intensely contested, and these debates resonate into the arena of statehood and the international system. On the one hand, what might be described as the conventional reading sees decolonisation as following from the formation of 'nationalist' movements by leaders who used them to oust the colonialists, and to build new and effective states in place of the old colonial empires. On this reading, decolonisation represents the universalisation of statehood, and with it, of nationhood: the spread of the ideology of the nation-state from its European origins throughout the world, and the establishment of statehood as the basis for global order. On the alternative reading, however, decolonisation is seen not as a nationalist triumph but as an imperial withdrawal. Colonial powers withdrew from empire because they were not prepared, and did not need, to pay the costs that it involved. This decision may be seen in turn as deriving from the recognition that they no longer had such need for the raw materials that colonies provided; that they did not need to maintain physical control over territory in order to retain access to these raw materials; and that their economic development depended much more on fostering their relations with one another (through such mechanisms as the European Union), than on continuing to occupy outlying areas of irrelevant territory. Only old-fashioned and underdeveloped imperial powers, such as Portugal and the Soviet Union needed imperialism any longer. Where viable states could be maintained in the formal colonial territories, so much the better; but decolonisation, on this reading, represented not the apotheosis of statehood, but its decline.

The same basic debate extends to the significance of the rapid rise in the numbers of states in the international system, not only following decolonisation but even more markedly with the end of the Cold War. On the one hand, by concentrating on those who form or demand new states, from Scotland through Slovenia to Eritrea and East Timor, one can argue that aspirations for state formation demonstrate the continuing vitality of nation statehood as the ultimate expression of political identity. On the other hand, by concentrating on existing states whose territories are broken up, one can argue that the maintenance of national integrity or territorial statehood does not matter in the way that it once did; and that whereas the era of state supremacy was marked by a steady diminution in the number of states, as powerful states in true realist fashion gobbled up weak ones, so the expansion in state numbers represents a decline in the ideology of statehood itself.

The answers to these questions are inevitably mixed: there's a bit of both. In some cases, the formation of viable and effective post-colonial states validates the nationalist interpretation. In others, the crumbling of post-colonial statehood validates the belief in imperial withdrawal. Some existing states, like Ethiopia, fought bitterly for years against Eritrean secession, and accepted it only once it could no longer be resisted. Others, like the Soviet Union, fell apart because the former state had no basis for its continued existence, and no one could be bothered to fight for it. In the context of this paper, while the existence of the first group is readily acknowledged, it is the second group that is of interest: although some (indeed many) viable post-colonial states continue, decolonisation and the end of the Cold War have done much to shatter the idea of statehood as a universal form of political organisation.

Simultaneously, of course, the state and the ideologies that sustained it have declined in the developed industrial world from which they originally derived. After the Second World War, only the superpowers were able to assure their own security without engaging in sovereignty-sapping alliances, and even these needed such alliances in order to maintain a bloc of like-minded states to sustain their superpower aspirations. The idea that individual states were able to manage their own economies without constant deference to external factors failed to survive the upheavals of the 1970s, and gave way to the spread of global economic liberalism. The centrally planned economy proved, behind its impressive façade, to be a device of extraordinary effectiveness for promoting poverty, alienation, and environmental destruction. In the 1990s, even the social security systems of developed industrial states have turned out to be both economically unsustainable in the light of global competition and penetration, and socially incapable of achieving the welfare goals that were promised in the two or three decades after 1945. The ignominious collapse of Soviet-style communism represented an all-in-one version of the decline in state effectiveness that had taken place over a longer period and in less spectacular fashion among the capitalist states of the West.

A plausible case can be made, indeed, that state collapse has been hastened and intensified by over-ambitious attempts to impose on societies a level of state control that they were ultimately unable to bear. In different ways around the world - in Western capitalist, Soviet-style communist and Third World states - governments had to come to terms with the discovery that the state was not an all-powerful technology through which they could mould their peoples and economies to their own wishes, but was, rather, a structure of management and control that was at the same time awkwardly and sometimes precariously posed between the people whom it sought to manage and the international setting in which it had to exist. This discovery was always likely to bear hardest on those states that were least well adapted both to their own societies and to their international setting in the first place. The remainder of this paper will seek to pursue the implications of this decline in the idea of universal statehood, with particular reference to the states of tropical Africa.

Why do some states collapse?

The indigenous bases for state creation in pre-colonial tropical Africa were for the most part peculiarly feeble. The population of the continent was sparse, and concentrated in a very limited number of locations - parts of the West African forest and savannah, the Ethiopian highlands, the Great Lakes, the southern highlands - which unsurprisingly corresponded to the major zones of statehood. Communications were poor, diseases of both humans and animals were rife, and opportunities to profit from long-distance trade (always a significant element in the development of states) were limited to parts of eastern Africa and the trans-Saharan caravan trade, until European contacts also opened up trading opportunities with the West African coast, largely in the form of slaves. In these circumstances, state formation was necessarily highly discontinuous, with only the Great Lakes region and parts of West Africa providing any significant zones of contiguous statehood. The kinds of state that emerged also varied considerably, from relatively permanent political organisations based on peasant cultivation, through small trading city-states with their immediate hinterlands, and military conquest states of which the most impressive in terms of its capacity for social transformation was undoubtedly the Zulu empire. In some areas nonetheless, most spectacularly in highland Ethiopia, it proved possible to sustain reasonably effective states over substantial periods of time, and in the process to build social structures and values which favoured deference to authority. These structures and values were to prove enormously significant in the post-colonial period. In much of Africa, it was necessary to maintain minimal levels of social cohesion by whatever mechanisms were available to poor and fragmented societies, in which the surpluses necessary to sustain states could be accumulated neither from taxing indigenous production nor from tariffs on long-distance trade. Two mechanisms which acquired considerable significance under these conditions were firstly extended family relationships, and secondly the forms of spiritual authority and control which were readily categorised by Western observers as 'witchcraft'. These localised and particularistic means of maintaining social order were difficult to incorporate into Westernised state structures, despite attempts to do so notably through the ascription of spiritual powers to autocratic leaders.

As a result, over very large areeas of Africa, the idea of a 'state' derives only from the imposition of colonial rule, and is inevitably associated with the decidedly mixed blessings of conquest in the first place, and the subsequent creation of 'modernity' in its manifold forms. The period of quasi-statehood (if such, following Jackson, we may now call it) that followed independence may accordingly be seen as an attempt to build African states on the foundations provided by colonialism, in collaboration with a supportive international system. This attempt was by no means uniformly unsuccessful, and should not be dismissed. There were certainly cases in which it worked relatively well, and may indeed have provided the basis for the long-term development of viable states. A significant element of 'banal nationalism' derived from the experience of living together, from exposure to common colonially-derived experiences such as the creation of a lingua franca, an educational system, and structures of administration. The 'nation-building' pretensions of early leaders were not entirely fraudulent or ineffectual. The point is not that this project did not work anywhere, but that it did not, and could not be expected to, work everywhere.

A familiar weakness of African state formation was nonetheless the very uneven fit between such indigenous bases for statehood as the continent possessed, and the imposed structures of colonial rule. The grafting process to which I have already referred was thoroughly inadequate both spatially and temporarily. Spatially, the continent was divided between colonial territories that broadly corresponded to pre-colonial states (Rwanda, Lesotho), territories in which a 'core' state or group of states was incorporated into a much larger area (Ghana, Uganda), and territories in which pre-colonial state traditions were either subordinated or non-existent. In historical terms, the whole process was so brief (little more than sixty years in many cases) that the level of penetration of external 'governmentalities' (to adopt Bayart's useful term) into the domestic society was necessarily shallow. Whatever the changes in institutional structures that colonialism imposed, the social attitudes which African brought to acting within these structures could only be expected to reflect their own long histories rather than the matching values which (through Christianity and education) the new rulers sought to inculcate.

It must then be recognised that states bring with them very significant costs, in social, economic and political terms; the state-centred and state-supporting literature of international relations has been so heavily concerned to emphasise the benefits of statehood that the other side of the account has gone almost unnoticed. These costs have been paid so long ago in most developed industrial societies that we have forgotten how heavy they often were; and the very success of these societies in meeting such costs has encouraged us to assume that they can everywhere be met with the same success. But even if they can eventually be met throughout the world, there is plenty of evidence to suggest that they cannot always be met yet. Post-colonial societies have moreover, been expected to develop a peculiarly modern form of statehood, modelled on state structures developed over a considerable period of time in now industrialised states, which make much greater demands on their populations than the rough-and-ready forms of personal rule from which states developed in Europe. Early European (and pre-colonial Third World) states rarely had to manage professional armies and bureaucracies; their economic responsibilities were slight; and there was no expectation that they should measure up to Weberian standards of rational-legality or (especially) financial accountability.

The social costs of statehood, and particularly of modern statehood, notably include the sacrifice of identities and social structures which are inimical to the hierarchies of control that states seek to impose. While effective states do not require that all (or by far the greater part) of their populations should share a common sense of identity, they are unquestionably greatly strengthened if this condition is met. Ethnically diverse states have been made to work under certain conditions, as in the Habsburg and Ottoman empires, but these states (even, or perhaps especially, in Europe) proved incapable of coping with the demands made on them by mass political participation. The constitutionalist mechanisms, such as federal arrangements and legally guaranteed rights, through which modern political engineers have attempted to reconcile statehood with ethnic diversity, themselves depend on the shared social values that are needed to uphold the constitutionalist ideal itself. Usually, such mechanisms are tacitly reliant on the hegemonic power of one particular ethnic group or institutional elite to keep the others in order, and preserve the façade of constitutionalism. Once the Communist Party of the Soviet Union collapsed, for instance, the elaborate apparatus of national self-government within the USSR, and indeed the USSR itself, went with it.

But states require more than a broad measure of social homogeneity: they also require people to have the right kind of society, and notably a society in which the reasonably impartial exercise of state authority is both exercised by those in power, and accepted by the governed. Somalia provides a classic example: despite having one of the very few ethnically homogeneous societies in Africa, it was quite incapable of generating forms of social power capable of holding together a society characterised both by intense individualism and by potentially very damaging clan divisions. While some elements of social structure can changes astonishingly quickly, others appear to be remarkable resistant to change - and none more so than the complex of social values concerned with the exercise and recognition of authority. The continuities that Robert Putnam has identified in the societies of northern and southern Italy can be replicated in Africa, and it comes as no surprise to discover that those African societies that have been best able to support effective statehood are those with a history of state formation and maintenance going back well before the era of colonial rule. Somalia and Rwanda, often taken as comparable examples of state breakdown in Africa, are actually complete opposites: whereas state breakdown did take place in Somalia from 1991 onwards, Rwanda at the time of the genocide in 1994 shows both an astonishing level of social discipline, on the part both of the genocidal regime and of its RPF opponents, and a rapid transfer of state power from one regime to another, which despite the traumatic circumstances of the genocide was marked by only the briefest period of intervening anomie. The killing in Rwanda was overwhelmingly carried out by disciplined forces under the control of the state.

This is no isolated comparison, but an example of a much broader phenomenon. Whereas insurgent movements derived from societies without a precolonial tradition of statehood - not just in Somalia, but in northern Uganda, southern Sudan, Liberia, and Sierra Leone - have had the greatest difficulty in establishing effective political structures, those in societies with such a tradition - as in southern Uganda, northern Ethiopia, and Eritrea - have been able to set up effective structures of government, even amidst the chaos resulting from the defeat of their incumbent rivals. In short, some societies have the social values required to sustain states, whereas others do not.

The economic costs of states are far more easily specified. States, and especially modern states - with their armies and bureaucracies, and the expectation that they will provide social services such as health and education - are very expensive organisations. Despite the hopes of nineteenth century imperialists that their African colonies could be made to pay for themselves, out of the profits raised by extracting raw materials for the world market, this rarely proved to be the case; even the very low level of administration maintained by British colonialism could scarcely be funded from internal taxation, while most colonial powers had to meet the costs of colonialism from metropolitan revenues. Independence led to a massive increase in demands on state revenues, both to meet the expectations of the population, and increasingly simply to pay for the state apparatus, and for a leadership whose own standard of living vastly exceeded that of the ordinary population. The weak social basis of African states had to be compensated for, by 'buying' political allegiance through a great variety of clientelist or neo-patrimonial mechanisms, which became increasingly difficult to sustain, and the failure of which further undermined support for the state. Certainly, poor economic policies and corruption exacerbated the problem, but the root causes were structural rather than merely the result of individual cupidity and policy failure; corruption, indeed, is far more plausibly analysed as a systematic problem that arises from the societies in which it occurs, than as the mere incidence of individual wrongdoing. African states also depended on a revenue base that was highly reliant on a small number of primary products, often with wildly fluctuating prices, and their attempts to squeeze out of the productive areas of the economy enough of a surplus to sustain their own activities, only resulted in the depression of the productive base itself. The imposition of economic reforms by international financial institutions, under the title of structural adjustment, had the short-term effect of still further depressing state revenues, while only very partially increasing the underlying level of production.

Structural adjustment programmes rest on the assumption that states are needed, in order to maintain the legal and physical infrastructure that is essential for the effective development of the economy, and successive adaptations of these programmes have placed increasing emphasis on state institutions, at the expense of the unfettered 'free market' ideology with which they originated. Very little attention has been paid, however, to the question of the level of economic production that is needed to maintain a state structure at all; and it seems by no means fanciful to suggest that there may come a point at which the resources required for state formation are simply not there, and that whatever devices are resorted to in order to raise state revenues will prove in one way or another to be counter-productive.

The political costs of statehood are also often high. Though states are commonly characterised as organisations that bring benefits, for many of their people - not least in modern Africa - they have much more evidently been a source of suffering. And while it is likewise readily assumed that such suffering - in the form, for example, of the imposition of government on rebellious peasants - is part of the historic price that has to be paid for the creation of ultimately beneficial state institutions, in modern Africa it has often proved to be politically counterproductive. Sheer bad governance has been at the root of many of Africa's insurgencies, and while a few of these - in Uganda, for example - have resulted in the establishment of more effective regimes than those that they overthrew, in other cases - such as Liberia or Somalia, and very probably Congo - they have destroyed or at least badly undermined the basis for any kind of government at all. Political violence in Africa rarely achieves the state-creating goals of a Henry VIII or a Louis XIV; it is much more likely to be state-destroying.

Even the support of the international system, which has been a key element in maintaining African states, can easily become counter-productive. The ready assumption of the Cold War period that each superpower needed to sustain its African allies, notably by military means, helped to foster a level of militarisation that, so far from sustaining states, ultimately helped to undermine them. Weaponry readily escaped from the control of the governments to which it was supplied - either by capture from government forces by insurgents within the state concerned, or as a result of the support given by states to opposition movements across their frontiers - and was used against states, rather than for them. Ill-disciplined 'national' armies (christened 'sobels' in Sierra Leone - part soldiers, part rebels) behaved in ways scarcely distinguishable from their opponents. Governments, like those of Mengistu in Ethiopia, or Mobutu in Congo/Zaire, were encouraged by the availability of externally supplied armaments to assume that opposition could be dealt with by force, rather than by political compromise. Nor have the changed external priorities of the post-Cold War period always enabled African states to reconstitute themselves on the basis of political compromise, multi-party democracy and respect for human rights: for every case that appears to 'work' (with Mozambique as the most striking example), there have been several others that don't.

Even though the failure of particular states can often be ascribed to specific instances of bad government, at the hands of a Samuel Doe or an Idi Amin, it is plausible to regard these as no more than precipitants that trigger the deeper underlying causes of state collapse. Equally, some states have proved capable of surviving much more intensive levels of bad governance than others: it is a source of wonder, for example, that Ghana has experienced economic collapse, without Ghanaians resorting (save in the north) to any serious political violence. Zimbabwe remains an astonishingly law-abiding society, despite the depredations of an obviously crumbling Mugabe regime. In Uganda, Rwanda and Ethiopia, it has been possible to put states back together again, despite levels of violence greater (or at least, not evidently less) than those that triggered state collapse in Somalia or Liberia. The relationship between the identifiable causes of state collapse and its actual incidence is by no means a linear one. Insofar as it is possible to hypothesise about cases in which state collapse occurs and ones in which it does not, however, by far the plausible guide is provided by the pre-colonial experience of statehood. Regions such as southern Uganda, Rwanda, southern Ghana, and northern Ethiopia, which sustained effective states in the pre-colonial period, are far better able to survive (and even learn from) the experience of bad government than regions without this experience. Social factors, in short, appear to be much more important than economic ones. All the potential wealth of Congo counts for nothing, in the absence of the social formations that are needed to sustain effective governance. And since, as already noted, regions of effective pre-colonial statehood in Africa occupy only a relatively small part of the continent's surface (though accounting for a much higher proportion of its people), opportunities for state failure are widely spread.

Dealing with State Failure

There are now extensive areas of Africa, and indeed of other parts of the globe, in which states can scarcely be said to exist, in anything more than the nominal sense that they are still coloured in on maps. Large parts of Angola and Sudan, indeed, have scarcely ever been under the control of their central governments since these two states became independent in the mid-1970s and mid-1950s respectively. The restoration of effective national government in Congo is looking increasingly like a lost cause. Though some kind of formal political authority has been restored in both Liberia and Sierra Leone, this has been achieved only by handing over that authority to the principal agents of state destruction, Charles Taylor's NPFL and Foday Sankoh's RUF. The principal challenge faced by the international system is not that of restoring states, but that of dealing with zones in which there is no state, or in which states have become so enfeebled as to be quite incapable of carrying out what have conventionally been assumed to be their functions. This is, however, nothing like such a daunting task as the state-centred conception of international relations might lead one to assume. The international system has extensive experience of dealing with non-states, dating back to the period in which these were a perfectly normal element of the global scene, and many of the characteristic devices through which this was done are now making their reappearance.

For a start, of course, statehood itself is by no means that all-or-nothing condition that it is commonly assumed to be in international relations theory. Not only are many of the entities that are conventionally recognised as 'states' completely incapable of performing many of the functions associated with statehood; other entities (such as guerrilla movements) sometimes do perform these functions. International actors are thus not faced by any hard-and-fast line between the familiar and comforting world of state-to-state relations, and the disturbing arena of stateless anarchy. The boundary is a fuzzy one, and allows ample opportunity for quasi-diplomatic relationships with state-like entities, or for the maintenance of the convenient fiction that some group of people represents a state, despite their complete inability to behave like one. Stateness or non-stateness in terms of any accepted set of objective criteria do not have to correspond to any way in which international actors go about their business.

Secondly, non-statehood by no means necessarily results in that war of every man against every man that the Hobbesian predispositions of international relations theory might lead one to expect. The existence of a state is not a necessary precondition for some kind of social order. In parts of Somalia, for example - a society with long experience of non-statehood - forms of social order which predate the external imposition of a Somali state are starting to reassert themselves. Though some parts of the world - such as Germany during the Thirty Years' War, or southern Africa during the mfecane - have had to endure years of horrifying violence, this is not an indefinitely sustainable condition. Sooner or later, people have to find some reasonably peaceful and effective way of living together. Often, this results in the formation or reformation of states, in regions where they had previously collapsed: the Thirty Years' War resulted in the Westphalian compromise, the mfecane in the emergence of a new state order, dominated by the Zulu empire and the other states of pre-colonial southern Africa. Given the artificiality of the state structure bequeathed to Africa by colonialism, it should be no cause for surprise if alternative state systems form over time in regions such as the Congo basin or the Sahel, and such systems may well prove to be more effective and accountable than their externally imposed predecessors.

Third, much of social and economic life is capable of sustaining itself in the absence of states: the ideology of statehood ascribes to states a centrality that they often do not possess, and falsely assumes that state failure must inevitably be associated with the failure of other life-support systems that states have taken over. In parts of Africa, indeed, states are more of a hindrance than a help. Somalia, for example, possesses one of the cheapest and most efficient cellphone systems on the continent, and Somali pastoralists can be found speaking directly from their camels to relatives in Bahrain or Toronto. This free market for telecommunications is certainly vastly more efficient than the cellphone system in states such as Kenya or Zimbabwe, where it has been bedevilled by the efforts of the respective presidents to monopolise it for the financial benefit of themselves, or their relatives and political cronies. The currency, likewise, is much more stable when this is taken out of the hands of the state, and people are free to use whatever store of value they please - preferably US dollars. The idea that a state 'has to' have its own currency, in the absence of an accountable political order that determines how that currency is maintained, is merely a pretence invented by state officials to serve their own interests. One extremely important element in the global political economy, the international trade in prohibited narcotics, benefits immeasurably from the absence of any state system to obstruct it, and provides considerable financial inducements for not having an effective state, which in turn would be liable to regulation and control by other international actors.

In much of Africa, health and education systems have already been 'privatised', as the result of the sheer incapacity of state to pay for them out of its declining revenues, or of the embezzlement of the money by state officials; non-governmental systems, which depend on the willingness of the users of these services to pay for what they receive, are likely to be both more efficient, and more accountable. Even the provision of the most basic of governmental services, justice and personal security, can be and often is privatised: the private security forces which are attracting increasing attention in Africa have come into existence basically because they are capable of providing a service that the government cannot provide itself. A case can certainly be made, in terms of the theory of public goods, that all of these services can best be provided by the state, and that their provision under circumstances where the state is either thoroughly incompetent or completely non-existent is inevitably skewed, unequal, and uncertain. But when the state itself is thoroughly 'privatised' - run for the benefit of its ruler and his close associates - there are no public goods, and the theory is inapplicable. This is the condition to which much of Africa has been reduced.

A fourth important element in dealing with statelessness, and one with long historical antecedents, is extra-territoriality. This is indeed implicit in the use of another country's currency. Structural adjustment programmes effectively amount to the takeover of a state's economic policies by an external agency, in much the way that default on public debt resulted in the assumption by the United Kingdom of control over the management of the Egyptian economy in the nineteenth century. But the most significant element is probably the increasing development of extraterritorial legal jurisdictions. No international corporation investing in Congo, for example, would dream of subjecting its investment to the tender mercies of the Congolese legal system, insofar as there even is one. Such investments are legally protected (even though they obviously require physical protection on the spot) by agreement among the parties that they be subject to the jurisdiction of some mutually acceptable external system of justice, such as that of the United Kingdom or the United States.

Finally, even though the Somali fiasco of 1992-95 has presumably put paid to the idea that the international community as a whole, acting through the United Nations, might take over responsibility for maintaining public order under conditions of state collapse, areas of statelessness effectively form 'free fire zones', in which individual states may intervene to protect their own interests, or in which intrepid individuals and organisations can operate at their own risk. Non-governmental organisations, with Médecins sans Frontières occupying a particularly prominent position, have developed the concept of a right of intervention in situations of humanitarian emergency, in which the overriding demands of 'humanity' take precedence over those of states; but many if not most of these situations arise in conditions of insurgency or state breakdown, in which the state nominally responsible for a particular territory is in any event in no position to enforce its claims. On occasion, with Somalia as a particularly clear example, such NGOs then call on external military forces to accord them the level of protection to which they feel that they are entitled. They are, in effect, the missionaries of the new world order. Even some international corporations, characterised by Reno as 'bottom-feeders', benefit from engagement in high-risk environments which correspondingly offer exceptional levels of profit.

Conclusion

People who think in terms of states, among whom international relations scholars occupy a prominent place, are constantly surprised by the ability of human beings to make do without them. Yet in recent years, states have withdrawn from many of the areas of human life in which they had previously been assumed to be essential, without for the most part leading to the disasters that had widely been predicted. Especially in the realm of economic management, state withdrawal has become an essential precondition for creating more effective and accountable services, and correspondingly enhancing levels of human welfare. As already noted, I certainly do not want to go so far as to claim that states can simply be dispensed with. State failure, not least in Africa, has characteristically been accompanied by tragic human costs. But it is time that students of international relations recognised, both that states bring with them costs as well as benefits, and equally that there are circumstances in which states simply cannot be maintained. In these circumstances, human beings develop alternative forms of governance, which likewise have their own, often very considerable, deficiencies, but which are necessarily adapted to the conditions in which they have to maintain themselves. The international system likewise adapts itself to dealing with such entities, with a flexibility that the state-centred analysis of international relations tends to overlook. There is certainly scope for international action to improve the conditions of life for people living, as best they can, under conditions of state failure; but rather than simply assuming that such action must necessarily involve the reconstruction of states, under conditions in which they have failed before and are likely to fail again, it would be wise to recognise that the brief period in which states covered the whole of the inhabited world has now passed, and that a reversion to patterns of international activity characteristic of previous eras is now called for.