Assessing State Failure
Assessing State Failure
Assessing State Failure
407
ARTICLES
DAVID CARMENT
Coming Anarchy offered a more bleak assessment of state failure in Africa. In this neo-Malthusian perspective the worldespecially the Southis beset by increasing crises generated by fast-growing populations, demographic changes and weakening state capacity to regulate conflict.3 That state failures serve as the breeding ground for many extremist groups is indisputable. Today most wars are fought either within nation-states or between states and non-state actors. Few wars pit one nation-state against another. One legacy of the Cold War is that many governments are more readily prepared to wage a conventional war against non-state actors. However, non-state actors differ fundamentally from state actors; the former are moving targets. They depend on highly decentralised structures that are semi-autonomous and can act and survive on their own. The list of internal challenges is extensive and growing and no region of the world is unaffected. In brief, the perceived pressure to anticipate and respond to state failure has increased in the past several years. The international communitys track record in this regard is not good. First, it has failed to prevent the slow collapse of states in Central and West Africa, despite a clear understanding of when and where such events would occur and the availability of forecasts predicting and explaining their causes and manifestations (as in the Congo, Guinea, Liberia and Sierra Leone). Second, it has failed to anticipate the moral hazards that are generated by efforts to address refugee flows, ethnic cleansing, and clan warfare (as in Rwanda and Somalia).4 Third, it has failed to understand the way biased outside involvement can actually accelerate conflict between combatants (as in Kosovo, Somalia, and Bosnia). Fourth, it has failed to produce credible responses to warring factions, thereby generating even greater violence (as in Rwanda and Bosnia). When examining these failures, the question to ask is: why? What do these failures suggest about when and under what conditions the international community should intervene to prevent tensions from escalating out of control, and how to manage violent situations when they do? It is well known that anticipating state failure is as much a matter of being able to generate an effective response as it is of getting the analysis right. But is the international community furnished with a solid analytical base from which to generate high-quality response strategies? In addressing these questions I make a threefold argument. First, most explanations of why states fail, including those that rely on comparative case study, historical trends, leading indicators, events-based data, field monitoring and expert opinion, are, in isolation, inadequate analytical tools for either risk assessment or early warning. This inadequacy exists for a variety of reasons. Many analyses point to fundamentally different causes of state failure; others rely on the monitoring of background factors and enabling conditions that are associated with the risk of conflict but do not themselves provide accurate information on the probability of specific events leading to failure. Still others do not distinguish between causality and correlation, while others are engrossed in issue-specific problems that are symptomatic of state weakness and human insecurityeg illicit gun flows, child soldiers, black market activity and AIDSproblems that are by themselves significant and important but not necessarily associated with 408
or causes of failure. Second, these disparate and often contending analytical approaches constitute a formidable and potentially useful tool-kit for risk assessment and early warning. However, there is a large and very real analytical gap between academics and practitioners on how to develop and use early warning techniques and methodologies. This is partly because, to be policy-relevant, analytical tools must also be useful operationally, organisationally and strategically. Further, the accumulation and integration of research findings is vital if theoretical insights are to generate important policy-relevant implications, especially at a time when academic early warning research is being criticised for its failure to provide policy-relevant diagnosis. As a result, many regional organisations as well as the United Nations are now developing an in-house capacity for conducting their own risk assessments and are developing independent procedures for conducting early warning, monitoring and response. One of the obvious dangers in creating independent analytical tools of this sort is that these lenses can and do point to fundamentally different problems of concern. If there is to be an improvement in the quality of response, future funding efforts should emphasise the integration of analytical findings and methodologies of various research programmes.5 In this regard, models and frameworks that relate directly to decision-making processes should have the highest priority. The causes of state failure: contending or complementary analyses? In this section I evaluate some of the contending and complementary claims on the causes of state failure in order to identify how these might fit into a coherent framework of analysis.6 Many explanations of state failure raise two interrelated issues. The first is the inherent problem of defining state failure in terms of consequences. I argue that state failure is a non-linear process of relative decay. Placed along a developmental continuum states can be characterised as strong, weak, failed and collapsed. Some states may never achieve the status of strong, moving instead from weak at independence to failed and in extreme cases to collapsed. Others may linger on as weak states for years and even decades. Others remain strong. In this vein, Somalia, Congo, Liberia, Sierra Leone and Bosnia are examples of state failure and collapse.7 In each of these cases the central government ceased to function, and was unable to provide for the well-being of its population or protect it from internal and external threats.8 States weaken and fail when they are unable to provide basic functions for their citizens.9 The economy weakens. Education and health care are non-existent. Physical infrastructure breaks down. Crime and violence escalate out of control. These conditions generate opposition groups which often turn to armed uprising. More often than not, the weapons of choice are small arms, light weapons and explosives because they are cheap, plentiful, durable, easily transported and simple to use.10 These conflicts create huge population shifts and refugee crises, long-term food shortages, failing economies, and the death of large numbers of civilians from disease, starvation and direct conflict. The second assumption is that there is a logical connection between state 409
DAVID CARMENT
performance and state failure and that there is a need to separate out the absolute development of a states capabilities within the international system, and the relative development of a states capabilities within the international system. Whereas the former process will always be unidirectional (though reversible) as a state develops (or regresses) over time, the latter process is going to be curvilinear because a states performance is being measured against other states in the international system which will or will not be developing at a more rapid pace. Thus, the proper referents for understanding state failure are not only a states own past, present and future performance in absolute terms but its performance relative to other states at any given point. The rate of change (which is understood by examining a states relative performance as opposed to absolute performance), whether progressive or regressive, tells us whether a state is moving either towards collapse or improvement. In other words, characteristics and indicators are useful for defining state failure only if there are appropriate reference cases with which to compare. And since these reference points are themselves evolving over time it is important to understand that failure is a relative term and has meaning only with respect to state performance at specific points. This distinction between absolute and relative performance not only helps us separate out causes from consequences, it also provides us with some clues about where we should look for explanations. Most scholars who seek to explain state failure are confronted by three distinct sets of empirical puzzles. Each puzzle is drawn from the perspectives of systemic transformation, statesociety relations and violent interactions and events. The first perspective associates state failure with macro-level changes in the international system.11 The second emphasises intermediate statesociety relations and the third emphasises micro-level strategic interactions between groups. More generally: Macro- or long-term processes are associated with system-structure transformations and the associated problems of the emergence of weak states; Intermediate mechanisms are associated with institutional viability and state weakness; Micro- or short term selection processes and mechanisms account for preferences for violence over pacific forms of strategic interactions and the subsequent escalation and/or duration of ethnic hatreds, violence, repression, and war at specific points in time. Much of the literature addresses state failure from the perspective of the first two puzzles, while comparatively less time has been spent addressing microquestions about the timing, escalation and the duration of political interactions leading to violence. This empirical gap is, of course, understandablelong- and medium-term perspectives furnish a useful overarching historical framework for studying system change and comparing state performance over relatively long periods of time, while explanations of specific choices, events or behaviours tend to focus on environmental stimuli in the context of standard social scientific models. 410
Macro-level perspectives: system and structure The development of political capacity, legitimacy and authority, all essential features of state building, is not a linear process.12 This is especially relevant to explaining state failure, since changing environmental conditions can reverse (in very short periods of time, eg months and years) these essential features. For example, changes in system structure can reverse state building in at least two non-mutually exclusive ways: through the creation of highly dependent weak states (compounded by the subsequent withdrawal of powerful patron-states) on the one hand; and through processes of economic development, and the strengthening of international norms of self-determination on the other. With respect to the former process, there have been at least four great waves of state building, each following the collapse of empires:13 South America in the 19th century (the Spanish Empire); Europe after the first world war (Russia, Austria-Hungary, Ottoman Turkey); Asia and Africa after the second world war (Belgium, Holland, France, Britain and Portugal) and Central Asia and Eastern Europe in the late 1980s (the Soviet Union).14 Most of these systemic transitions were associated with the abrupt creation of new states in hostile environments involving conflicts over territory and identity.15 For example, both the post-second world war phase of decolonisation and the break-up of the Soviet union introduced many more new and weak ex-client states into the international system.16 However, unlike the elites of the West European and Latin American nations, the leaders of these new African, Asian and East European states were faced with three compounding problems which enhanced their perception of insecurity. The borders they had to defend were arbitrary; their societies were usually diverse in composition; and few leaders had experience in building inclusive civic and democratic cultures. In essence, the security threats of these states were and are as much internal as external.17 HansHenrik Holm argues that the latest wave of weak states in the 1990s is a consequence of the way the international system has developed over the past 10 years. Like Ignatieff, Holm believes that the Cold War ensured that most weak states at least survived but with its end most of these states have been left to sink or swim.18 Similarly, Robert Rosh and Mohammed Ayoob argue that state failure is largely a function of the withdrawal of outside support to weak states.19 To the extent that regional conflicts as well as the maintenance of state integrity were both key features of the international system during the Cold War, there may be some validity to these claims.20 The net result, as Ayoob suggests, is the absence of effective statehood in much of the Third World, or what some scholars have termed quasi-states.21 Historical perspectives suggest that the political configuration of ethnic groups and the degree of constraint they exercise over the state and its decision makers is determined in part by colonial experience and outside involvement.22 In colonised states, demographics have largely determined changes in political power. This is particularly true when political power is not coterminous with economic power. A single ethnic group may be encouraged by the colonial power(s) to dominate economic and political processes at the national level and as a result be confronted by challenges from other groups.23 411
DAVID CARMENT
Consistent with this view, William Zartman proposes that state collapse in Africa has occurred in two wavesthe first came towards the end of the second decade of independence and the second, a decade later and into the 1990s.24 Zartman notes that state collapse is usually marked by the loss of control not only of political space, but of economic space as well. The two work in opposite directions, with neighbouring states encroaching on the collapsing states sovereignty by meddling in its politics, thereby making the political space wider than the states boundary.25 As this happens, the economic space retracts, with parts of the economic space captured by the neighbouring states and the informal economy dominating the rest.26 Charles Alao provides empirical evidence for this argument through an examination of the causes of state weakness in Africa during the postindependence and post-cold war periods. He finds that Africas weak states and subsequent failures were a result of the way African states were formed: colonialism brought people of different ethnic, political and religious affiliations together to form a state and forge a common sense of citizenship.27 In addition, most African economies were incorporated into the European capitalist framework, which made most of these economies structurally too weak to cope with the challenges of nation building. Jeffrey Herbst suggests that the paradox of decolonisation in Africa stems from the formal colonisation of Africa and the replacement of the continents diverse political systems with an artificial state system which was carried forward in post-independent Africa.28 The faade of sovereignty was to be overturned only a few years after independence by pseudo-Marxist regimes, one party-states and patronclient fiefdoms.29 Chadwick Alger notes that most Western powers failed to pay attention to developing viable institutions of governance in Africa which could support the independence of new states. As a consequence, African authoritarianism emerged from a series of interrelated phenomena that arose out of the colonial legacy (most states were conceived in violence, there was little transformation in the economy and the local ethnic elites commitment to the Western imposed structures was low).30 When there was a convergence of interests between Africas newly emergent classes it was to end colonisation but these goals became fragmented by ethnic loyalties. The consequence was the formation of patronclient relations and the development of personal rule resulting in a state based on personal authority and coercion. The political system became structured not by institutions but by politicised ethnically based patronclient relationships. In a few short years following independence, the African state became dominated by divisive ethnic rivalries. A second and related set of macro-level perspectives traces state failure to processes associated with economic development and the development of international norms of self-determination.31 Both affect the likelihood of state failure only indirectly. More importantly, empirical research finds support for the general claim that economic conditions influence a variety of political and social events, including violence and government instability. With respect to the former, economic equity issues are at the forefront of many analyses of state failure. For example, Gurr and Duvall state that greater social justice within nations in the distribution of economic goods and political autonomy is the most potent path to 412
social peace.32 Gurr cites further evidence of the link between minority rebellion and economic differentials, while Gurr and Duvall and Kpsowa and Jenkins, among others, draw out the link between external economic dependence and a heightened vulnerability to various forms of civil disorder.33 The international economic context began to turn against existing state structures in Africa in the early 1980s with the rise of structural adjustment, then the end of the Cold War precipitated a much more hostile international context for state formation. Stedman suggests the triumph of free market ideasundermined the external sources of support for Africas patrimonial regimes and left some with no legs to stand on. In sum, he adds, economic conditionality cut at the heart of the patrimonial state.34 Like the international economic factors specified above, international norms also contribute to conflictual state structures and, indirectly, to the propensity to state failure. The rise of multi-ethnic authoritarian states usually results in a minority groups perceived sense of exclusion and failure in the social, economic and political domains. As a result, minority groups recognise that internationalisation of their demands can both simultaneously encourage internal mobilisation and weaken the salience and effectiveness of the state by creating international forums for sub-state grievances.35 This legitimisation process is supported by the existence of supranational organisations and international norms which provide a forum and focal point for sub-national claims.36 Specifically, international organisations promote sub-state mobilisation by providing human rights recognition and support which legitimise self-determination claims.37 Christopher Allen has suggested that the patterns of violence and warfare in Africa, as well as the characteristics of the new violence, are attributable to this kind of internationalisation.38 In particular, ethnic conflict and the desire for independence arise out of the systematic suppression by the modern state of minority political and economic interests.39 Similarly Edward Azar and John Burton, among others, have long argued that the move to violence begins with the denial of separate identities, the absence of security for minorities and the clear absence of effective participation for these minorities.40 Intermediate perspectives: statesociety relations Whereas systemic perspectives emphasise the weakening of states as a result of minority group mobilisation in the face of global economic and normative transformations, intermediate-level perspectives emphasise the weakening and in some cases the collapse of the state in the face of internal pressures. The assumption here is that the emergence of state disorder results from the failure of prevailing societal values to legitimise existing divisions of labour and political order.41 States in decay are in transitional stages in which existing ideologies fail to legitimise the positions of various actors in a hierarchical social structure. Under such conditions most scholars predict the result will be the breakdown of the social and political order.42 Mark Lichbachs examination of non-co-operation is useful in bringing out the level of difficulty involved in reversing the process of decay. Each type of solutionmarket, contract, community and hierarchylogically precedes any 413
DAVID CARMENT
one of the others: contract requires market, because beneficial mutual exchange permits parties to arrive at the terms of a contract; market requires community, because common values create the trust needed to conduct market transactions; community requires hierarchy, because common values must be authoritatively enforced and passed on to future generations; and hierarchy requires contract, because in the very long run only mutually agreed upon coercion will be accepted.43 In the context of state failure, the destruction of national identities stands in the way of acquiring shared values that could provide a basis for inter-group cooperation. Identity-based politics also detract from the publics ability to appreciate the value of market exchange; the utility of depriving a rival group of benefits may be perceived as greater than the disutility of foregoing gains from trade. The obstacles to contractual or hierarchical solutions to the problem of mutual non-co-operation are analogous: both the trust required for a contract and the legitimacy needed for stable hierarchy will be elusive when ascriptive, exclusive identification holds sway in a society. Thus as Douglas Dearth suggests, a state is said to have failed if it does not fulfil the most basic obligations of statehood. The leadership does not have the means and credibility to compel internal order or to deter or repel external aggression. In addition, the leadership does not, or cannot, provide sufficiently for the people to attract minimal sufficient domestic support. Consequently, as Pauline Baker and John Ausink argue, a fully collapsed state is one that has lost legitimacy, has few functioning institutions, offers little or no public service to its constituents and is unable to contain fragmentation.44 State collapse begins when the central state starts to deteriorate, leading to the fractionalisation of society, with loyalties shifting from the state to traditional communities that seem to offer better protection. For Dearth this process is a three-step progression. First, institutions fail to provide adequate services to the population. Second, improperly channelled ethnic, social and ideological competition erode the effectiveness of these weak institutions even more. Finally, the cumulative effects of poverty, overpopulation, rural flight and rapid urbanisation, as well as environmental degradation, overwhelm the weak state to the point of collapse.45 Decay has both internal and external implications. As Vernon Hewitt suggests: high levels of domestic instability limit a states ability to act authoritatively within the international community, limit its ability to act on domestic society with any legitimacy, and to deliver socio-economic packages aimed at bringing about widespread industrialization.46 Nirvikar Singh suggests that a states legitimacy is closely tied to the kinds of ethnic policies its pursues. Narrow policies favouring one group are less sound than broad distributive ones.47 In the absence of strong, secular organised parties and strong institutional structures, ideology and culture become the focus for understanding state decay and failure. Ultimately, it is the states actions that are directly responsible for the erosion of the political system in the first place.48 According to Paul Brass and Atul Kohli, among others, the state does not merely respond to crises produced by uneven ethnic mobilisation and social change, but is itself the dominating force providing differential advantages to regions and 414
ethnic groups.49 Peter Wallensteen sees the convergence of the internal and external dynamics as the ultimate basis for evaluating state performance. There are instances of decay where the state is under-consolidateda situation where the state is not effective in the performance of its duties; and cases where the state it is over-extendedwhere it becomes a threat to its inhabitants.50 As states begin to rely more extensively on coercive forms of managing internal (mostly ethnic) tensions, power tends to become more concentrated in the hands of a few and potentially homogeneous ethnic groups.51 This disjuncture creates recurring problems of governability for those in power. The resulting breakdown begins at the state centre as hierarchical patterns of authority give way to regional, decentralised, ethnic and informal forms of political and economic organisation. The net result is conflict between a single ethnic group dominating political institutions and the counterbalancing efforts by minority groups to wrest control from the centre. Ultimately, as Gros argues, states fail when public authorities are either unable or unwilling to carry out their end of what Hobbes long ago called the social contract, but which now includes more than maintaining the peace among societys many factions and interests.52 Empirical support for this argument is provided by Ren Lemarchand, who reflects on the crisis in the Great Lakes region of AfricaRwanda, Burundi, and Zaireand the patterns of state decay affecting these countries.53 He suggests that state decay here occurred within vertically structured social arenas. Exclusionary policies were a major source of erosion of state legitimacy, as evident in the way politics were played out between Hutus and Tutsis in Rwanda and Burundi.54 Similarly, Indonesias institutional system, political structure and popular media reinforce the identity of state-centric nationalism. The state is not so much a subordinate to the dominant ethnic group but works in partnership with it. This partnership is reinforced when the state is challenged by minority groups, itself a response generated by assimilative pressures, policies on in-migration, economic competition and more direct political threats of secession. The net result is a lethal policy feedback process in which the central governments policies in the form of entitlements for the majority ethnic groups induce minority groups to organise for political action. This challenge in turn generates greater resistance to change from the state centre. Micro-level perspectives: dynamic interactions Macro- and intermediate perspectives are extremely useful for understanding the root causes, enabling conditions and background factors associated with state failure and collapse.55 They identify structural and societal factors associated with weakness and can account for changes in political, social and economic demands over time. They may, under some circumstances, be able to explain why groups end up fighting over resources and territory. But they cannot explain organised violence; that particular subset of human social interaction which involves a high level of inter-group hostility. Nor can they account for variations in the scope, severity and timing of organised violence more generally. Individuals and groups may be persuaded by elites to hate and fear members of other groups and they 415
DAVID CARMENT
may be driven by mass pressures to rebel, but the probability of war, violence, ethnic cleansing and genocide depends on the opportunities and constraints that present themselves to the warring factions and their leaders at specific points in time. Thus, micro-level perspectives are premised on two interrelated processes. The first is the interaction between armed factions and groups and the second is between armed factions and groups and outside forces in a position, through actions and statements, to alter the direction and intensity of violence. With respect to the former, most assessments of intra-state violence underscore the prominent role played by elites in the mobilisation process. Failed states are viewed as a problem of emerging anarchy where organised groups that lack many of the attributes of statehood must pay attention to the primary problem of their own security. 56 In a state of emerging anarchy, or whenever the internal balance of power shifts, questions of control become pre-eminent. This strategic environment can cause hostile groups to fear extinction and yield to mob violence. Accordingly, political opponents may emulate traditional state behaviour by seeking relative power gains against other groups. The lack of an arbiterinternal or externalinduces problems of credible commitment between groups that do not trust one another and are liable to misrepresent information for relative gains.57 According to Beverly Crawford and Ronnie Lipschutz, broken social contracts and weakened oppressive institutions open political space for political entrepreneurs to mobilise support. If the political gains made available to these entrepreneurs are achieved through the re-allocation of resources or the disproportionate economic deprivation of one group in favour of another, the net result will be the escalation of conflict towards inter-group violence. Similarly, Charles Tilley suggests that successful use of coercion by a state in order to suppress local ethnically based challenges enhances the assessment of its future utility. Hence, coercion against minority ethnic groups is also a normative factor, since elites who use violence become habituated to violence.58 Violence becomes a part of elite political culture that is assimilated into the national identity. Violence becomes a useful political tool.59 Under conditions of decay, if the state centre loses its autonomy by favouring one group over another, the disadvantaged group is likely to believe that whatever social contract there was is broken and cannot be fixed without some sort of third party to provide minimal security guarantees. A battle for independence is likely to follow if a third party guaranteeing agreements between groups cannot be found.60 In the absence of a third party guarantor, negotiation will be extremely difficult because groups possess fundamental incentives to defect. The mistrust that develops increases the desirability for disadvantaged groups to pursue a proactive conflictual stance and to mobilise against the state.61 The second kind of micro-level interactions are those between belligerents and outside forces that are in a position to influence the dynamics of the conflict from onset to termination. Relying on only macro- and intermediate-level explanations of state failure is as unwise as evaluating only interactions between factions. This argument becomes clearer when one considers the role and impact of outside parties in affecting the course of specific violent behaviours. In this view, outside 416
forces greatly influence the selection of violence at key junctures in a conflict as a strategy for securing group survival. Here the concern is not only partisan support for factions through processes of diffusion and escalation but the less well understood impact that third party interveners have on conflict dynamics. For example, recent research on third party involvement shows that humanitarian assistance can exacerbate tensions between groups because of the incipient moral hazard problem. Others suggest that a lack of resolve and credibility within security organisations creates additional incentives for escalation and prolonged conflict. Structural (macro- and intermediate-level conditions) imperatives may have accounted for the mutual hatred underlying fighting in Bosnia, Kosovo and Rwanda, but these wars were waged with specific objectives in mind. Decisions by Serb leaders to escalate the fighting in Bosnia and by Hutu leaders to initiate a genocide in Rwanda depended on the prospects of winning (and losing) specific pre-planned battles and confrontations. With respect to Bosnia, whenever Western leaders mounted a prolonged and stable threat of retaliation backed by ultimata, deadlines and a clear commitment to punish, credibility was high and coercive diplomacy worked. Weak threats, on the other hand, promoted violence. The genocide in Rwanda was a direct consequence of strategic decisions by political and military officials within the UN Security Council not to mount an effective and pre-emptive peace enforcement mission as late as March 1994.62 Generating evidence: from general theories to model development Thus far, I have examined general theories on state failure and/or processes associated with state weakness, collapse, breakdown and violence. Unfortunately, theoretical insights alone are insufficient to generate effective and specific responses to state failure. This is because most theories by themselves lack specificity and rarely consider the operational milieu in which effective responses have to be generated. Theoretical insights are useful as a starting point for more in-depth analysis and then only if decision makers can be persuaded that the information is useful to finding an appropriate fit between strategy, the problem at hand and the resources available. These problems mean that analysts must establish a time frame appropriate to the issue at hand. In this sense, anticipating state failure is like peeling an onion in which each analytical layer noted above reveals progressively longer time lines: long-term fundamental dynamics relating to macro-level preconditions and consequences; mid-term intermediate behavioural patterns; and immediate microlevel events such as political crises and ethnic cleansing. For example, warning must come years in advance to respond strategically to structural problems (development, institution building, establishing infrastructure) but only a year or two when escalation is imminent and when the tasks are to engage in preventive diplomacy, dialogue and mediation.63 Model development is important because it specifies the relationship between these levels of analysis rather than treating them as independent and isolated causal factors. Drawing on the above discussion a model should focus on two types of variables: configurational variables, which defines state failure processes in terms of the interaction between units of analysis (eg state and society, warring 417
DAVID CARMENT
factions etc); and composite variables which characterise state failures by summarising the background conditions, attributes and performance indicators of the state over time. Thus, identifying state failure is a three-step process involving the use of both composite and configurational variables. These three steps are: 1) identifying the relevant configurational and composite variables; 2) postulating thresholds in order to identify significant transformations and shifts from states of equilibrium; and 3) determining the independence of variables in order to isolate the causal significance of each variable. Articulating such constructs and concepts is useful in the generation of propositions or hypotheses about state change. These propositions can in turn be tested empirically to determine whether or not they have factual support. To the extent that use of conceptualisations enables the theorist to describe and explain, in simple language, the complex processes of state failure, the use of concepts is justified. However, the ultimate focus of the theorist should not be these constructs and models. These should be understood only as heuristic devices to explain complex phenomena to policy makers. Models used to generate evidence for the explanation and prediction of state failure correspond to the configurational and composite approaches described above. These kinds of models include forecasting as well as risk assessment models.64 As Dipak Gupta shows, in addition to distinct levels of analyses these approaches can also be distinguished by their methodologies as depicted in Figure 1.65
Forecasting Models
Time series
Neural network
Expert-based
Delphi technique
ARMA
MARMA
Hidden-Markov method
Sequential model
Bayesian model
Figure 1 Source: Dipak Gupta, An early warning about forecasts: oracle to academics, in Susanne Schmeidl & Howard Adelman (eds), Synergy in Early Warning Conference Proceedings, Toronto, 1518 March 1997, pp 375396.
418
The following approaches are identified according to the methodology employed and the level of analysis. 1a) Macro-level evaluation of structural indicators (econometrically or through pattern recognition techniques) (eg parts of the State Failure Project; PIOOM; CIFP; HEWS; ICB; FIRST; Rummels Democide database; Uppsalas Conflict database).66 1b) Macro-level time series of leading indicators67 (eg IOM; Refworld; the Food and Agriculture Organisations (FAO) GIEWS; Reliefweb; the UN systemwide Earthwatch; HazardNet for disasters; the global early warning system for displaced personsGEWS). 2a) Intermediate-level conjunctural models that track changes in pre-specified events and interactions between groups (eg conflict/co-operation, genocide, non-violent protest) using machine-coded data, pattern recognition and neural networks (eg GEDS; PANDA; KEDS). 2b) Intermediate-level structured (Delphi) and subjective models, which utilise a team of experts who identify key actors and estimate their future position on a given issue (regime stability, turmoil likelihood, investment restrictions and trade restrictions) with regard to their power to influence the outcome, the importance (salience) they attach to the issue, and the certainty or firmness of the actors orientation (eg Decision Insights; Political Risk Services).68 The scores which emerge from this assessment are used to provide a formal estimate of probability. 3a) Micro-level sequential models which develop risk assessments based on tracking specific behaviours, using accelerators (eg parts of State Failure; CEWS); 3b) Micro-level response models which evaluate outside response to conflict and develop feasibility assessments based therein (eg Helen Feins Life Integrity Violations Approach; IDRCs International Development Research Centre, PCIA). 3c) Micro-level field reporting by NGO networks (eg FEWER; FAST; ICG, CIPDD) using structured and/or unstructured reporting techniques.69 The array of choices in terms of units of analysis, deductive and inductive methodologies, qualitative and quantitative theoretical assumptions and timeframes renders politically relevant and integrated analysis of state failure difficult but not impossible. On the one hand, where conflicts are well understood in both form and content and the causes are proximate and escalation likely, the main problem will be to identify the relevant configurational variables through an evaluation of micro-level interactions (3a,b,c). On the other hand, where the situation is latent, a state of equilibrium is achieved or behaviour is only remotely suggestive of political or economic collapse, careful monitoring of composite indicators and trends at the macro and intermediate level will be essential (1a,b, c; 2a,b,c).70 An example of an approach integrating composite and configurational variables is Barbara Harffs sequential model for early warning of genocide and politicide. The approach resembles a qualitative time-series approach, but 419
DAVID CARMENT
incorporates the role of accelerators. She identifies 10 background conditions, four intervening conditions and eight accelerators. What is interesting is that she does not assume that state development is linear. Where processual models, without accelerators and triggers, identify stages of a conflict, these static models cannot provide adequate risk assessments that will allow for planning of responses to impending situations. This is where the dynamic role of accelerators and triggers comes in, and ideally those that are essential and necessary. An alternative methodology is Moore and Gurrs employment of data from the Minorities at Risk project to compare three empirical approaches to long-term risk assessments. Their work generates risk profiles; lists of high-risk factors, or leading indicators based on general theoretical knowledge such as group incentives, capacity and opportunity. They then apply a theoretical regression model in which an argument is expressed as a multiple equation model, and a statistical techniquethree-stage least squaresis applied to the data to estimate the parameters of a predictive equation. Finally, they employ an empirical regression model, an inductive approach similar to the State Failure project, in which statistical software determines specific indicators for assessing probabilities.71 It should be noted that each model produces slightly different results, although with a proportion of overlap. The obvious conclusion is that, barring any weaknesses in the internal validity and reliability of these methods, it is difficult to select, on the basis of findings and rigour alone, one methodology over the other. Each purports to explain and predict different facets of state failure and its causes. There remains the need to reconcile empirically valid but potentially contending claims on the causes of state failure, on the one hand, and the desire for accumulation, integration and policy relevance on the other. How can multiple approaches and the accumulation of findings be simultaneously encouraged and developed? One approach would be to integrate research at the level of findings. The focus would be on those causal factors that appear in multiple assessment lists. This would entail a brief description of the method employed in policyrelevant terms and then the establishment of a watch list. While it is true that one does not need a complex model in order to put states on a watch list, it also true that models and theoretically generated insights can direct the analyst towards causal factors that are potentially unique to a given situation (in other words they provide details about what specifically is to be warned about). They can also be counterintuitive (they direct the analysts attention to something that might otherwise be overlooked or ignored). A simplified version of the watch list approach is that espoused by the International Crisis Group (ICG). ICG provides fairly detailed and regular assessments of current hot-spots based on expert analysis. But the reports do not provide the kind of specificity required for accurate prediction of events nor do they necessarily provide an overview of the relative risks based on regional or incountry performance (for example, Indonesia faces several internal insurgencies, all or none of which might contribute to the countrys demise). In general a watch list is useful as a complement to the more detailed analysis required to generate policy options. That is, analyses can be diagnostic (if x then y) or they can be diagnostic and prescriptive (if x then y and we should do z). 420
Therefore, a second approach would be to integrate methodologically dissimilar risk assessment procedures, frameworks and models through a consortium of analysts, policy advocates and practitioners into a dynamic exchange of information. Such an approach could provide a more comprehensive and accurate picture than would any single methodology. It would also be better placed to provide the kind of specificity required to link solutions with causes. Conclusions Debates on state failure have mainly focused on definitional issues, the strengths and weaknesses of contending methodologies and evaluation procedures as well as the causes, manifestations and processes of state failure. Much less attention has been paid to the question of how to link theoretical insights to policy options. To date states and international organisations have done little towards the creation of working and useful conflict prevention regimes at the regional and global level. While there is no lack of rhetoric on the necessity of prevention, serious attempts to give organisations the tools to put preventive systems into place are modest at best. In an effort to moved the debate forward, I have argued that understanding and responding to state failure requires a multifaceted, multilayered and multi-actor methodology. This approach entails three levels of analysis, relative performance measures and an appreciation of dynamic processes of conflict. The ramifications of such an approach are straightforward. With respect to policy, preferences for solutions to state failure will depend on the explanations we accept for their decay and potential collapse. If one emphasises root structural causes (economic, social and political composite indicators) the list of solutions might include longterm, developmentally orientated structural prevention. If one emphasises medium- and micro-level political configurations and interactions, the range of solutions might include everything from partition, power sharing, democratisation or constitutional entrenchment of ethnic or minority rights, to more specific operational responses such as sanctions, peace enforcement and the long-term institution building.72 A key question I have tried to answer is how to render academic analyses accessible to the practitioner. I have suggested that it is important that those who focus on developing methodologies of risk assessment and early warning for the purposes of analysing and predicting state failure are clear about how an institution can best use the array of political instruments available to it to provide an effective response. Ultimately, anticipating state failure is a process-based approach requiring sound analysis as well as an explicit connection to policy options for preventive measures. A process-based approach means that the method and format of applied early warning is shaped directly by the operational focus of the process itself, in this case preventative action as opposed to preparedness. All of these elements point to the relevance of basic policy analysis and planning methods to close the warningaction gap. Such methods incorporate the structuring of problems, the application of appropriate analytical tools to solve these problems, and the communication of analysis and recommendations in a format useful 421
DAVID CARMENT
to decision makers. In short, policy planning is a type of decision-support procedure. Such an approach requires that organisations have a better sense of their own institutional needs and capabilitiesfar more than they do now. Such activities especially those premised on long-term structural transformationsshould have a built-in evaluative process or impact assessment capability that will, in effect, ensure self-monitoring and provide policy guidance on what to do. At the very minimum activity in an economically and politically fragile society should not further destabilise that society.
Notes
The author would like to thank Robert Rotberg and Rasheed Draman for very helpful comments on earlier drafts of this paper and the World Peace Foundation at Harvard University, the Social Sciences and Humanities Research Council of Canada and the Security and Defence Forum of Canada for their support in this research. An earlier draft of this paper was presented at a Conference on Failed States, Harvard University, 2001. Portions of this paper appear in D Carment, Preventing state failure, in Robert Rotberg (ed) The Failure of Nation-States: Causes, Consequences, and Reconstruction, Princeton, NJ: Princeton University Press, forthcoming. This research builds on the Country Indicators for Foreign Policy Risk Assesment initiative (located at www.carleton.ca/cifp). 1 Unprecedented power, colliding ambitions, The Economist, 26 September 2002. 2 Michael Ignatieff, Blood and Belonging, Toronto: Viking Press, 1993, p 8. 3 Robert D Kaplan, The coming anarchy, Atlantic Monthly, February, 1994. 4 Robert H Dorff, Democratization and failed states: the challenge of ungovernability, Parameters, Summer, 1996, pp 1731. 5 The consequences of not acting early are obvious. Effective peace building, essentially picking up the pieces after a violent conflict has torn a country or region apart, is a costly and time consuming process. 6 Charles Alao identifies a number of interrelated factors. These include weak state structures and their inability to cope with post-cold war transition; deteriorating economic conditions; and the rise in ethnic conflict. See Charles Alao, The problem of the failed state in Africa, in Muthiah Alagappa & Takashi Inoguchi, (eds), International Security Management and the United Nations (Tokyo: 1999, pp 83102. See also Gerald B Helman & Steven R Ratner, Saving failed states, Foreign Policy, Brookings, 89, 1992/93, pp 320. 7 From a video transcript, Small arms and failed states, 24 October 1999, at http://www.cdi.org/ adm/1307/transcript.html. 8 To understand what a failed state is, it is important to understand a successful state. At its core, a successful state provides for the basic security of its population, protecting it from both internal and external threats. It also has the capacity to provide for the health and welfare of its population. See http://www.cdi.org/adm/1307/transcript.html. 9 State Failure Task Force Report, 30 November 1995. Prepared by Daniel C Esty, Jack A Goldstone, Ted Robert Gurr, Barbara Harff, Marc Levy, Geoffrey D Dabelko, Pamela T Surko, and Alan N Unger. According to the Task Force a failed state is one that is utterly incapable of sustaining itself as a member of the international community (p 1). Narrowly defined, however, state failures consist of instances in which central state authority collapses for several years (ibid.). However, since fewer than 20 such episodes have occurred during the past 40 years, statistical analysis is difficult. Therefore the Task Force broadened the concept of state failure to include a wider range of civil conflicts, political crises and massive violations of human rights that are typically associated with state breakdown. In line with such a broad definition, the Task Force isolated four kinds of state failure: (1) revolutionary wars, (2) ethnic wars, (3) mass killings, and (4) adverse or disruptive regime change. 10 See http://www.cdi.org/adm/1307/transcript.html. See also Pauline H Baker & John A Ausink, State collapse and ethnic violence: toward a predictive model, Parameters, Spring, 1996, pp 1931. 11 See Ted Robert Gurr, The internationalization of protracted communal conflicts since 1945: which groups, where and how?, in Manus Midlarsky (ed), The Internationalization of Communal Strife, London: 1992, pp 424; and Michael Brecher & Jonathon Wilkenfeld, The ethnic dimension of international crises, in David Carment & Patrick James (eds), Wars in the Midst of Peace: The International Politics of Ethnic Conflict, Pittsburgh, PA: Routledge, 1997, pp 164193.
422
13 14
15
16
17
18
19
20
21
22 23
24
25
26
27 28
Ted Robert Gurr (ed), Handbook of Political Conflict: Theory and Research, New York: 1980; and Samuel Huntington, Political Order in Changing Society, New Haven, CT: Yale University Press, 1968. Naomi Chazan (ed), Irredentism and International Politics, Boulder, CO: Lynne Rienner, 1991. Wallensteen argues that social stratification most of the time leads to the break down of states. Using data on how civil wars end since 1945, Wallensteen argues that state break-up, when well managed, does not have to be as disruptive as often thought. Peter Wallensteen, State failure, ethnocracy and democracy: new conceptions of governance, paper presented at the Failed States Conference, Purdue University, West Lafayette, 2527 February 1998. There are several distinct causal mechanisms: (1) the internalisation of external conflict in which weaker states invite outside involvement. For example, external involvement exacerbates internal conflict to the extent that the poorer the nation, the more invidious the inequalities, and the more dependent the state is, the more susceptible it is to violence internally. (2) The externalisation of internal conflict involving: a) a shift in the dyadic balance of power, or b) external intervention. See Dina Zinnes, Why war? Evidence on the outbreak of international conflict, in Gurr, Handbook of Political Conflict, pp 331360; Jack S Levy, The diversionary theory of war: a critique, in Manus I Midlarsky (ed), Handbook of War Studies, Boston, MA: Unwin Hyman, 1989, pp 259288; Patrick James, Conflict and cohesion: a review of the literature and recommendations for future research, Cooperation and Conflict, 22, 1987, pp 2133. Ted Robert Gurr, Resolving ethnopolitical conflicts: exit, autonomy or access, in Ted Robert Gurr et.al, Minorities at Risk: A Global View of Ethnopolitical Conflicts, Washington, DC: USIP Press, 1993; Peter Calvert, The Foreign Policies of New States, New York: St. Martins Press, 1986; and Edward E Azar & Chung-in Moon, National Security in the Third World: The Management of Internal and External Threats, Aldershot: Edward Elgar, 1988. See Azar & Chung-in Moon, National Security in the Third World. The absence of interstate ethnic strife in South America supports this conclusion. Although homogeneous, these states also have lower levels of internal cleavage and very low levels of transnational affinities. According to Holm, the international system is created on the basis of the norms from the dominant states concerning the idea of the state, legitimacy, and the legal framework for the state. The weak states are unable to live up to these norms. The weak states represent both a system failure and a system responsibility. Hans-Henrik Holm, The responsibility that will not go away: weak states in the international system, paper presented at the Failed States Conference, Purdue University, West Lafayette, 2527 February 1998. Robert Rosh, Ethnic cleavages as a component of global military expenditures, Journal of Peace Research, 33, 1989, pp 2130; Mohammed Ayoob, State-making, state-breaking and state failure: explaining the roots of Third World insecurity, in Luc van de Goor, Kumar Rupesinghe & Paul Sciarone (eds), Between Development and Destruction, London: Palgrave, 1996, pp 6786. Mohammed Ayoob, State-making, state-breaking and state failure examines the Western model of state making (in the 17th and 18th centuries) and tries to draw a parallel with what is currently taking place in the Third World. Ayoob also examines the twin concepts of ethno-nationalism and selfdetermination, which relate to state failure. According to Ayoob, state failure predominates when institutions collapse, when existing institutions are not fulfilling peoples basic needs and when satisfactory alternative structures are not readily available (p 80). Ibid. Ayoobs main argument is to link the emergence of state failure to superpower competition in the Third World. Donald Horowitz, Ethnic Groups in Conflict, Berkeley, CA: University of Berkeley Press, 1985). Rene Lemarchand, Patterns of state collapse and reconstruction in Central Africa: reflections on the crisis in the Great Lakes, Afrika Spectrum, 32, 1997, pp 173193. I William Zartman, Introduction: posing the problem of state collapse, in Zartman (ed), Collapsed States: The Disintegration and Restoration of Legitimate Authority, Boulder, CO: Lynne Rienner, 1995, pp 111. Ibid. According to Zartman, such scenarios lack clear turning points and warning signals. Nonetheless, failure is a slippery slope that has some notable characteristics towards the end and these sometimes serve as warning signals of imminent collapse He identifies five such characteristics: the devolution of power to the peripheries because the centre fights among itself; the withering away of power from the central government; government malfunctions by avoiding necessary but difficult choices; the incumbents practise only defensive politics; and the ultimate dangerthe lost of control by the centre over its own agents, pp 111. Ibid. Zartman suggests that state collapse arises because of the poor performance of the states functionsrepresentation, interest articulation, output efficiency, etc. Charles Alao, The problem of the failed state in Africa, pp 83102. Jeffrey Herbst, Responding to state failure in Africa, International Security , 21, 1996/97,
423
DAVID CARMENT pp 120144. Chadwick F Alger, Failed states and the failure of states: self-determination, states, nations and global governance, paper presented at the Failed States Conference, Purdue University, 2527 February 1998. The absence of confrontations and conflicts between classes may have prevented the growth of liberalism, with its ideological and legal emphasis on individual rights and liberties. The European models of democracy dissolved quickly as a foreign arrangement. John Burton, The history of international conflict resolution, in Edward E Azar & John W Burton (eds), International Conflict Resolution, Boulder, CO: Lynne Rienner, 1986, pp 2840; Edward E Azar, The Management of Protracted Social Conflict, Aldershot: Dartmouth Press, 1990; Ted Robert Gurr, Peoples against states: ethnopolitical conflict and the changing world system, International Studies Quarterly, 38, 1994, pp 347377; and Anthony D Smith, The ethnic sources of nationalism, Survival, 35, 1993, pp 4864. Ted Robert Gurr & Raymond Duvall, Civil conflict in the 1960s: a reciprocal theoretical system with parameter estimates, Comparative Political Studies, 6, 1973, p 160. For a full analysis see Dane Rowlands & Troy Joseph, The International Monetary Fund, civil strife and conflict prevention, in David Carment & Albrecht Schnabel, Conflict Prevention: Grand Illusion or Path to Peace?, Tokyo: UNV Press, 2003; Ted Robert Gurr, Why minorities rebela global analysis of communal mobilization and conflict since 1945, International Political Science Review, 14, 1993, pp 161201; Gurr & Duvall, Civil conflict in the 1960s; and Augustine Kpsowa & J Craig Jenkins, The structural sources of military coups in postcolonial Africa, 19571984, American Journal of Sociology, 99, 1993, pp 126163. See Nicolas Van de Walle, The economic correlates of state collapse, paper presented at a Conference on State Failure, Harvard University, 2001; and Stephen Stedman, Conflict and conciliation in sub-Saharan Africa, in Michael E Brown (ed), The International Dimensions of Internal Conflict, Cambridge: MIT Press, 1996, p 243. Anthony D Smith, Conflict and collective identity: class, ethnic and nation, in Azar & Burton (eds), International Conflict Resolution, pp 6384. See also Michael Nicholson, Failing states, failing systems, paper presented at the Failed States Conference, Purdue University, West Lafayette, 2528 February 1998. Holm, The responsibility that will not go away, examines what he calls the new international system and its attendant negative consequences for the survival of weak states. He looks at the recent construction of a bifocal system with wealthy nations at the top and poor nations at the bottom and, consequently, the emergence of two zones: that of peace and that of war. While grappling with the instability that has been created by this bifocal system, the international system was further affected by the emergence of globalisation, which has overshadowed the bifocal system, creating further instability. Chris Allen, Warfare, endemic violence & state collapse in Africa, Review of African Political Economy, 81, 1999, pp 367384. Allen examines the patterns of violence and warfare in Africa as well as the characteristics of what he calls the new violence, including barbarism and the economics of war. Ibid. Allen uses the term spoils politics to describe what is at play in Africa. Spoils politics occurs when the primary goal of those competing for political office or power is self-enrichment (p 377). Prolonged spoils lead to terminal spoils and ultimately to state collapse. At the terminal stages of spoils politics, where state failure sets in, the main political features include the decline or disappearance of state functions and offices; abusive use of remaining institutions, notably the army and police; the contraction, fragmentation or disappearance of central authority; and a relationship between the state and society that consists largely of mutual avoidance or violence and resource extraction. Allied to the political features are a number of economic features: general contractions of the economy leading to its decline or the emergence of a second economy. The social consequences of all these, according to Allen, is endemic violence. Stephen Ryan, Explaining ethnic conflict: the neglected international dimension, Review of International Studies, 14, 1988, 161177. Azar & Burton, International Conflict Resolution, Boulder, CO: Lynne Rienner, 1986. See Atul Kohli, Democracy and Discontent: Indias Growing Crisis of Governability, Cambridge: 1990. Huntington, Political Order. The gap between tradition and modernity marks a transitional stage characterised by anomie and the decay and corrosive impact of economic development on established beliefs and behaviours. Mark Lichbach, The Rebels Dilemma , Ann Arbor, MI: University of Michigan Press, 1995, pp 279280. Baker & Ausink, State collapse and ethnic violence.
29
30
31
32
33
34
35
36
37
38
39
40 41
42
43
44
424
46
47
48
49
50 51
52
53 54
55
56 57 58
59
Douglas H Dearth, Failed states: an international conundrum, Defense Intelligence Journal, 5, 1996, pp 119130. Failed states are a matter of concern for two reasons. The first is formalisticbecause the state system has for the past 350 years characterised the organisation of the international system, when states fail, especially in large numbers, the entire system is thought to be in danger. The second reason is humanitarian state failure results in widespread human suffering. Vernon Hewitt, The International New Politics of South Asia, Manchester: University of Manchester Press, 1997, p 198. For example, Kohli shows that, since 1967, Indias political system has undergone a decline in order and authority. This decline has arisen as a result of two factors: the emergence of political parties based on ethnic identity and the diminishing capacity of secular elites at the state centre to influence the political behaviour of those below them. Kohli, Democracy and Discontent; and Nirvikar Singh, Cultural conflict in India: Punjab and Kashmir, in Ronnie Lipschutz & Beverly Crawford (eds), The Myth of Ethnic Conflict, Berkeley, CA: University of Berkeley Press, 1999, 352360. See Paul Brass, The Politics of India Since Independence, New York: Cambridge University Press, 1990. Wallensteen, State failure, ethnocracy and democracy. In examining the roots of state failure, Wallensteen argues that it has to do with state building. He examines the issue of ethnocracy and the dissolution of empires and argues that the ethnisation of society can be a cause of its falling apart. Ethnocracyrule by one ethnic groupinstils in all social groups an understanding that ethnic security is associated with state power. Ibid. See Rajat Ganguly (ed), Ethnic Conflict and Secessionism in Asia: Causes, Dynamics, Solutions, New York: Sage, 2003. Jean-Germain Gros, Towards a taxonomy of failed states in the New World Order: decaying Somalia, Liberia, Rwanda and Haiti, Third World Quarterly, 17, 1996, p 456. Ultimately, as Gros suggests, the factors in state failure also include poor economic performance; lack of social synergy; authoritarianism; militarism; and environmental degradation caused by rampant population growth. According to Gros, since the degree of stateness varies considerably across most cases of state failure, rather than putting all cases together as instances of failure, it makes sense to situate them along a continuum. Based on this argument, Gros presents a taxonomy of five types of failed state. These are the anarchic statewhere there is no central government; the phantom or mirage state where a semblance of authority exists; the anaemic statewhere the energy of the state has been sapped by counter-insurgency; the captured statewhere there is a strong central authority but one that is captured by members of insecure elites to frustrate rival elites; and states that failed in vitro (aborted states)before the process of state formation was consolidated. Lemarchand, Patterns of state collapse and reconstruction in Central Africa. According to Lemarchand, long before Rwanda and Burundi hit collapse they were faced with demographic pressures which they could not contain, among other things, land shortage, a key ingredient of ethnic strife in that region. Another important factor, of relevance to all the countries in the region, is the shrinking of the political bases of state authority. Ethnic elites often intentionally promote feelings of cultural and economic inequality for political gain in the hope of establishing a strong, unified base for action. The strategy often involves tapping economic grievances and re-framing them in the discourse of ethno-nationalism. Where culturally divergent groups inhabit a common society, there is a structural imperative in which one group becomes subordinate to another. This is especially common when ethnic cleavages are reinforced by differences in class and status or when labour markets are divided along ethnic lines. The inequalities often result in exploitation, discrimination and blocked mobility for members of subordinate groups, which create the grievances, feelings of relative deprivation and, by implication, the motives for ethnic mobilisation and violence. Barry Posen, The security dilemma and ethnic conflict, Survival, 35, 1993, pp 2747. Lipschutz & Crawford, The Myth of Ethnic Conflict, p 36. When inter-group violence ensues, states take control through the provision of policing and similar functions. The degree of enforcement available to states is variable. At one end of the spectrum are police states in which all forms of political conflict are discouraged. Frequent success in the use of state-organised violence (for example to achieve national consolidation and suppress internal challenges) leads to the development of police states These include military regimes and one-party states. In states that have little or no experience of managing ethnic tensions, and where constraints are low, hegemonial exchange and its more coercive variant, the control model, are the usual alternatives. Control models differ from hegemonialexchange models to the extent that there is a superordinate ethnic group in power. The elites of these groups have developed the techniques of coercion, depoliticisation and co-option in order to maintain power. Control becomes institutionalised and usually arises when the state is faced with imminent
425
DAVID CARMENT collapse. See David A Lake & Donald Rothchild (eds), The International Spread of Ethnic Conflict: Fear Diffusion and Escalation, Princeton, NJ: Princeton University Press, 1998. Russell Hardin, One for All: The Logic of Group Conflict, Princeton, NJ: Princeton University Press, 1995, p 143. Research that combines insights from primordialist, political-economy and instrumental perspectives generally focuses on economic and political disparities between the state centre and minority. Conflict between dominant groups and minorities usually involves issues of national identity; the expansion and centralisation of nationalist political authority, which creates a competitive arena for state controlled resources; and the recognition of ethnicity as a basis for resource competition and political access. Establishing a credible reputation for responding to internal violence is important, and, arguably, military measures have a greater chance of lowering pay-offs from violence than less vigorous forms of intervention such as sanctions. Mary O McCarthy, Potential humanitarian crises: the warning process and roles for intelligence, in Susanne Schmeidl & Howard Adelman (eds), Synergy in Early Warning Conference Proceedings, Toronto, 1518 March 1997, pp 1516. Forecasting is about the likelihood an event will happen. By itself it has no strategic connotation or purpose. Forecasting can be either passive (about events over which we have no control) or active (about events over which we have some control). For example, weather forecasts are a form of passive forecasting: they do not tell us, when there is a 50% chance of rain, whether it will rain half the day, or whether it will it rain every half hour or whether it will it cover half the region. To be policy-relevant, forecasting needs to take on additional qualities. It must be diagnostic, by which emphasis is placed on describing how and why things work as they do, and it must be prescriptive, offering explicit recommendations to policy makers faced with certain kinds of problems. Dipak Gupta, An early warning about forecasts: oracle to academics in Schmeidl & Adelman, Synergy in Early Warning Conference Proceedings, pp 375396. I would like to thank Luc van de Goor of the Clingendael Institute for insights on the conceptual organisation of the these initiatives. Listed acronyms, organisations and urls are current as of June 2001 and include: PIOOMthe Dutch acronym for Interdisciplinary Research on Root Causes of Human Rights Violations (http://www.fsw.leidenuniv.nl/www/w3_liswo/pioom.htm); CIFPCountry Indicators for Foreign Policy (http://www.carleton.ca/cifp); HEWSHumanitarian Early Warning System (http://www.reliefweb.int/ocha_ol/programs/pad/hews.html); ICB International Crisis Behaviour Data Project (http://web.missouri.edu/~polsjjh/ICB/); FIRSTFacts on International Relations and Security Trends (http://www.sipri.se/projects/database/index.html); Democide Research Homepagehttp://www2.hawaii.edu/~rummel/; Uppsalas Conflict Databasehttp://www.pcr.uu.se/ data.htm; IOMInternational Organization for Migration (http://www.iom.int/); Refworldhttp:// www.unhcr.ch/refworld/country/cdr/cdrsom2.htm; GIEWSGlobal Information and Early Warning System (http://www.fao.org/giews/english/giewse.htm); The UN-based EarthWatchhttp:// www.unep.ch/earthw.html; HazardNethttp://hoshi.cic.sfu.ca/~hazard/; GEWS Global Early Warning System (http://fugimodel.t.soka.ac.jp/FUGI/chapter6/chapter6.html); GEDSGlobal Event Data System (http://geds.umd.edu/geds/) (this system has been replaced by a more comprehensive integrated database, The Integrated Network for Societal Conflict Research (INSCR), under the direction of Monty Marshall at CIDCM, Maryland University; PANDAProtocol for the Assessment of Nonviolent Direct Action (http://www-vdc.fas.harvard.edu/cfia/pnscs/panda.htm); KEDSKansas Events Data System (http://www.ukans.edu/~keds/); Political Risk Serviceshttp://www.polrisk. com/; CEWSConflict Early Warning Systems (http://www.usc.edu/dept/LAS/ir/cis/cews/); PCIA Peace and Conflict Impact Assessment (http://www.idrc.ca/peace/); USAIDs Impact Assessment Participatory Country Program Strategic Planning, and Performance Monitoring, Field Manual, USAID/BHR; FASTEarly Recognition of Tension and Fact Finding (http://www.swisspeace.ch/); ICG International Crisis Group (http://www.intl-crisis-group.org/); CIPDDCaucasian Institute for Peace, Democracy and Development (http://www.armazi.demon.co.uk/cipdd.html). A leading indicator approach would use previously identified relationships or sequences of events to identify the precursors of instability or conflict. The basis for this approach holds that there are sequencing regularities that allow the forecaster to discover what variable to focus on in order to project a trend. The major disadvantage of a leading indicator approach is that, while it can often predict the direction of change, it gives no indication of its magnitude For an excellent example, see the work of Sean OBrien Anticipating the good , the bad and the ugly: an early warning approach to conflict instability analysis, 19752015, paper presented at the annual APSA meeting, San Francisco, 2001. For an example of predictions using the Decision Insights model, see Patrick James & Michael Lusztig, Assessing the reliability of predictions on the future of Quebec, in David Carment, John F Stack, Jr & Frank Harvey (eds), The International Politics of Quebec Secession: State making and
60
61
62
63
64
65
66
67
68
426
ASSESSING STATE FAILURE State Breaking in North America, Westport, CT: Praeger, 2001. In order to establish an integrated framework for analysing the emergence of violent conflict and conflict management, it is necessary to understand how each given type of crisis typically develops and which possible avoidance efforts can be effective. In general terms, the factors that contribute to conflict escalation are categorised as structural factors (root causes), accelerators (precipitators/ facilitators), or triggers (catalysing events). Structural Factors: background conditions that form the preconditions of crisis situations, such as systematic political exclusion, inherent economic inequities, lack of adequate and responsive institutions, the presence of ethnic minorities, resource exhaustion, and over-dependence on international trade. Accelerators: feedback events that rapidly increase the level of significance of the most volatile of the general conditions, but may also signify systembreakdown or basic changes in political causality. Triggers: sudden events that act as catalysts igniting a crisis or conflict, such as the assas-sin-ation of a leader, election fraud, a political scandal. Portions of this part of the paper are based on the risk assessment and conflict indicators approach developed by CIFP (www.carleton.ca/cifp). Will Moore & Ted R Gurr, Assessing risks of ethnopolitical rebellion in the year 2000: three empirical approaches, in Schmeidl & Adelman, Synergy in Early Warning Conference Proceedings, pp 4570. Definitional and operationalisation issues are always problematic when dealing with events that are controversial, rare and not well understood.. John Cockell, Early warning and the United Nations system, in Carment & Schnabel, Conflict Prevention.
69
70
71
72
427