3. Methodology 39
3.1 Quantitative or qualitative methods 39
After the Cold War, existing theories in conflict studies were reconstructed in order to explain the phenomena of civil wars. Simultaneously, new theories emerged, such as the greed theory
that focused on the causes of wars. Many studies used a statistical analysis, which does not provide a deeper understanding of the conflicts (Duyvesteyn, 2012, p. 602). Many scholars used Paul Collier's theory on greed as a trigger to war (Kalyvas, 2001; De Soysa, 2002; Regan & Norton, 2005). Collier's method and theory have been criticised. David Keen argued the weaknesses of Collier’s work in his article greed and grievance in civil war (2012). His criticism focused in particular on the quantitative method that Collier used. The quantitative method in conflict studies includes the use of databases that are, especially in poor and conflict-ridden countries, not completely reliable. Keen argued that the problem of missing data has been the exception rather than the rule (2012, p. 767). Duyvesteyn added to this point that research cannot rely on data from the state-level only, as it cannot explain sub-state violence. Therefore, a scholar needs reliable data from, for example, the local level to investigate the role of resolution, negotiation and settlement in the final stages of armed conflict (Duyvesteyn, 2012, p. 602). Collier even admitted himself in his book Understanding Civil Wars, which he wrote together with Nicholas Sambanis, that the main shortcoming of the Collier-Hoeffler model is the lack of data. Despite this shortcoming, the model would still give information about the dynamics that cause violence or peace. The advantage of the comparative case study design is that it enables to look at the interaction of variables over a specific timespan within a case and compare it to other cases in order to find out more about the processes that drive conflicts. This way, similarities and differences between cases and even patterns of frequent occurrence can be identified (Collier & Sambanis, 2005, p. 1-2).
Mats Berdal (2005) also questioned the use of the quantitative method in conflict studies. In his opinion, the truly interesting questions regarding violent conflict remain unanswered when quantitative method is used. A statistical analysis offers limited insight in the identification of factors that trigger and sustain civil war, the interaction between these factors, and the mutation of these factors over time. In order to gain a deeper understanding of conflict dynamics, one can use the qualitative method. A quantitative approach can be interesting for research looking for a precise percentage, such as the chance the returning of a conflict. Unfortunately, it offers limited knowledge on the specifics of individual cases or the actual conflict dynamics. Another important limitation of the quantitative method that Berdal (2005, p. 690) mentioned is the requirement to choose empirical data that might leave out highy relevant factors of the case. In order to gain a deeper understanding of a conflict and its
dynamics, one should select a small N and consider many variables. It calls for a qualitative approach.
Karen Ballentine and Jake Sherman (2003) confirmed that statistical methods can be used to identify key variables across cases, which only highlight part of the situation. It proves a correlation between two factors, such as economic factors and intra-state conflicts. The explanatory weight of statistical methods is limited, as to whether and how specific factors promote or sustain conflict remains unclear (Ballentine & Sherman, p. 5). Qualitative methods can be used to go beyond the revelation of correlation and identify causality. In contrast to the quantitative approach, which assesses the occurrence of specific factors in an isolated manner, these qualitative methods can be used to ascertain the causal impact on conflicts of specific factors (for instance economics) in relation to and combination with other possible significant factors (for instance culture, ideologicy or politics) (Ballentine & Sherman, p. 6). While a case-specific approach might lack the rigor of large-number quantitative studies, it does offer a more precise understanding of how specific factors ignite a conflict, the extent to which the factors explain a conflict, and the influence of other factors.
The latter is also confirmed in Tarrow’s article Inside Insurgencies: "[…] not quantities but interactions are the key to the dynamics of violence in civil wars" (2007, p. 596). In this article, Tarrow argued why he thinks the quantitative method is not right in conflict studies. He stated that much of the used data in quantitative conflict studies is over aggregated and that analytical distinctions are ignored, such as the distinction between civil wars as a whole and violence in civil wars as a variable. Secondly, the operationalisation of variables is often incomplete, using only crude, incomplete or indirect indicators. Thirdly, a variable can have different effects in different phases of the war or in other wars, which is often not taken into account in many of the conflict studies using the quantitative method. The fourth comment expresses the same concern as Keen (2012) and Duyvesteyn (2012) arued about the missing data from the subnational actors in civil war, since most the data is coming from state actors. Finally, Tarrow stretched the lack of extensive research into mechanisms or processes of civil war insurgencies, which is hard to measure using quantitative approach. Opposed to the quantitative study, a qualitative study enables better assessment of contextual factors, which gives more insight into the mechanisms and processes of irregular wars (Tarrow, 2007. p. 589).