This paper examines the effectiveness of partition in ceasing violence during ethnic conflict. Wigmore-Shepherd’s 2012 study argued that ethnic conflict is often due to the congruence between ethnic and political identity, allowing political conflicts to become ‘ethnicised’ and ethnic conflict to eclipse the original political dispute. Therefore this paper hypothesises that ethnic homogenisation via partition can allow the original political conflict to re-emerge in a potentially violent manner. The hypothesis is tested by an agent based model adapted from the model used in the 2012 study. The model finds that in the instances where there is not a perfect congruence between ethnic and political identity, politically motivated violence does persist in the ethnic enclaves.  It was found that a lower level of congruence would result in a higher level of post-partition violence. Furthermore the act of migration itself can encourage spikes of ethnically motivated violence and agents of different ethnicity cross paths to reach their enclave.

This thesis outlines and presents an alternative hypothetical process to the emergence of ethnic conflict. Ethnic conflicts, rather than being dependent upon pre-existing ‘ancient hatreds’, are instead the result of a congruence between ethnic and political identity which grants individuals the ability to use ethnicity to identify and eliminate political threats. This hypothesis is formed by the examination of three case studies of ethnic conflict: Lebanon, Northern Ireland and Croatia. This hypothesis is then formalised and tested using an agent based simulation in which agent interactions are dependent upon ethnic and political identity and the congruence between the two. As predicted there was a strong positive correlation between how accurately ethnic identity reflected political identity and the level of ethnically motivated violence in the simulation, although the relationship was not linear. Furthermore the effect of a shift in congruence was found to be roughly comparable to the effect of initialising agents with a moderate level of pre-existing ethnic antagonism.