Civil war and related concepts such as state failure have traditionally been studied at the level of nation states, where the nation states are either "at war" or not, and treated as phenomena to be explained by state-level characteristics. A cursory glance at actual civil wars, however, reveals that violence rarely engulfs entire states, but typically occurs within a confined area, leaving other areas within the same state at relative peace. The Albanian KLA insurgency in Yugoslavia, for example, was limited to the province of Kosovo, and did not at the outset imply a direct threat of violence to inhabitants of core areas such as Belgrade. Likewise, a "failing" state does not go from being fully effective over its entire territory to completely ineffective.
State capacity is perhaps better viewed as a matter of degree, where states may be more or less effective in certain geographical areas of their territory. The Northwestern provinces of Pakistan bordering Afghanistan, for example, are largely outside the direct control of the central authorities, even though the government may be quite effective in exercising authority in other areas. Most existing studies of civil war, however, treat civil war as an aggregate outcome at the level of the state, and ignore all variation within states and regions experiencing conflict.
Although existing research has made great progress in understanding conditions under which the onset of civil wars are more or less likely and when conflicts are more likely to be settled or persist, there is often a stark contrast between conclusions of large-N cross-national studies of violence and case studies that narrate individual conflicts. Although case studies may give too much weight to idiosyncrasies of a particular conflict relative to general theory and patterns, many of the non-findings and conundrums in the existing cross-national research on civil war may follow partly from construction, due to their near-exclusive reliance on country-level attributes.
Local Level Versus Aggregate Characteristics
Existing studies have generally neglected how local level characteristics can differ notably from global or aggregate characteristics. For example, many analysts have emphasized that civil wars are not confined to individual states, but show a tendency to involve transnational actors and spread or diffuse once underway. Existing research at the nation state level, however, has largely disregarded the role of the location of conflict and paid little attention to where violence occurs.
Most theories of intervention and diffusion emphasize mechanisms or links that are spatially confined. In the case of Russia, for example, the civil war in Chechnya is unlikely to diffuse or affect distant neighbors such as Norway and Finland. Looking at the conflict region in the Caucasus more specifically, however, suggests that many of the geographical neighbors have been strongly affected by the conflict, and that the behavior of local actors is highly relevant for regional outcomes.
Likewise, case studies and large-N studies have often reached sharply differing conclusions regarding the role of ethnicity and inequality in intrastate conflict. Whereas most cross-national studies have found little evidence that patterns of ethnic fragmentation and dominance are strongly related to the onset of civil war, individual narratives often emphasize relations between ethnic groups as root sources of conflict.
What existing studies have shown, however, is that global measures of ethnic fragmentation and dominance at the level of states are not strongly related to conflict. However, if civil wars are local phenomena, then there are no inherent reasons to expect that national level attributes should capture or reflect the conflict-related local characteristics. Consider, for example, the case of Indonesia. The indigenous population of the Aceh province is a small share of the total population of the state (only 3 million out of 210 million), but a majority in parts of Sumatra, where the perceived threat of expansion from Malays mobilizes insurgents.
Likewise, studies on the relationship between deprivation and conflict that focus on inequalities among groups generally tend to find stronger evidence than studies that look at aggregate social inequality. Although minority groups may constitute a small share of total population (and as a result contribute little to overall inequality in a country), large social and economic differences that follow group lines may contribute to political polarization and mobilization. Whereas theory suggests that (local) differences between groups and actors primarily make a difference under specified contingencies, most of the existing research have looked at national aggregates and averages that are only loosely linked to the rationale for conflict.
Civil War Diffusion
Just as existing studies that focus on national characteristics fail to take into account local variation that is not easily aggregated or summed up to the level of states, they often also fail to reflect or adequately represent central aspects of the processes by which civil wars diffuse, expand, or evolve within countries and regions. Accounting for the onset, diffusion, and expansion of conflict requires understanding how individual attitudes change from the values, norms, and practices of peacefully coexisting communities, where conventional forms of political action dominate and violence is deemed to be an unacceptable political tool.
Existing studies of conflict have focused primarily on the strength of nation states and the "opportunities" or "willingness" of potential insurgents as a focus of country characteristics, but has explained less about the individual attitudes and characteristics that prevail in situations where individuals resort to the use of violence, and why attitudes change in some communities but not others. The civil war in Bosnia-Herzegovina, for example, was to a large extent a localized process that diffused in a "top-down" or imposed fashion to particular communities within regions. The war was initially brought to Bosnia by nationalist forces in Serbia, strongly connected to fighting in neighboring regions of Croatia, and to a large extent orchestrated by individuals from outside the area. There are numerous examples of local communities in Bosnia where members of the different ethnic groups initially tried to cooperate on preventing violence. Such efforts, however, became difficult to sustain as the consequences of fighting elsewhere in the region increased fears among individuals and changed perceptions from a situation where peaceful coexistence was possible to a situation where partisan allegiances dominated and the use of violence was seen as necessary.
Project Specifics
Project organizers believe that whereas cross-national studies of civil war may be approaching a state of diminishing marginal returns, studying civil war and transnational conflict in a more disaggregated fashion offers considerable promise of providing insights into the micro-level processes that make up the aggregate phenomena that we label civil war, state failure, or transnational violence. They hope to foster research disaggregating the study of civil war along three broad lines:
studies that examine local level attributes and how these are related to the onset, duration, and outcomes of violence;
studies that detail the micro-level process of interaction among actors that make up what we call violence and "peace" at the aggregate level; and
studies that examine the differences in individual values associated with "peaceful" communities and "violent" communities where neighbors engage in the use of force against one another.
The term ‘transnational conflict’ is used to emphasize that many intrastate conflicts between states and non-state actors are not confined to individual states, but often involve actors and resources that lie outside of the territory of an individual state. Researchers will start by looking at variation in conflict and peace geographically and over time in a certain state or region by combining local level social and economic data and geo-referenced data sources with survey data from geographically stratified samplings. Knowledge based on particular conflicts and regions can then be used to develop theory that can be applied more broadly. This will in turn be used to foster better informed comparative empirical studies, as well as provide input for computational modeling efforts and simulations that can accommodate both observed trajectories and counterfactual worlds.
Related Research
The DSCW project is related to various initiatives that either have received funding or are in the development process. There is a potential for great synergies among these projects and their participants.
Organizers Kristian Skrede Gleditsch and Barbara Walter, both of UC San Diego, have funding from the National Science Foundation for a 2004–2006 project on the "International Dimensions of Civil War." The project has collected new data on the attributes of non-state actors in civil war and the transnational linkages of civil, which will be useful for more disaggregated studies of civil war. In addition, a group of researchers at UC San Diego (Miles Kahler, David Lake, Gleditsch, and Walter, as well as Jodi Nelson from the International Rescue Committee) are working on "Failed States and the Reconstruction of Political Authority: International Strategies and the Role of NGOs," where the key themes outlined above will be highly relevant.
Gleditsch is involved in NSF-funded project "The Dynamics of Civil War Outcomes" with John O’Loughlin (Colorado University), Jeremy L. Mennis (Temple University), Gerard Toal (Virginia Tech), and Michael D. Ward (University of Washington), which will examine the effects of civil wars on public health, the environment, infrastructure, political attitudes, refugee movement and settlement, and economic adjustments, focusing on Bosnia and the North Caucasus. The project will run from 2004 to 2007.
Gleditsch is also involved in "Geographic Representation of Wars" (with Lars-Erik Cederman, Håvard Hegre, Simon Hug, and Jan Ketil Rød), a cooperative project between UC San Diego, the Swiss Institute of Science and Technology, the Norwegian University of Science and Technology, and the Center for the Study of Civil War (CSCW), International Peace Research Institute, Oslo. The project combines GIS-based empirical information on conflict-related features and computational modeling efforts.
Finally, DSCW organizers hope to rely on links to CSCW, with which UC San Diego has a cooperative agreement. The CSCW includes a working group on "Conflict and Values" that has carried out surveys in Bosnia-Herzegovina on attitudes and their relationship to conflict and peace.
IGCC
is a non-profit, nonpartisan institute with official 501(c)(3) status. We welcome
your tax-deductible donations to help support our work, and encourage you
to contact
us about our programs and activities.
Copyright 2001–2008 by the Regents of the University of California on
behalf of IGCC.
