This study seeks to empirically measure the moderating impact of PR-STV in ethnically divided societies by comparing levels of moderation in the manifestoes of Northern Irish parties at Northern Ireland Assembly elections, which take place under PR-STV, and U.K. House of Commons elections, which operate using First Past the Post. Northern Ireland presents a unique electoral situation, where ethnic parties regularly contest elections under two different systems: one which is designed to foster moderation, and one which is not. Whereas most studies on the impact of electoral engineering in divided societies compare elections before and after significant constitutional changes, the present study mitigates the confounding impact of social and attitudinal changes that occur over time on party moderation levels by limiting the time-lag between PR-STV and FPTP elections. Because these two electoral systems continue to operate alongside one another in an ethnically divided society, the study compares elections under different electoral systems but in largely similar political contexts, thereby isolating as much as possible the electoral system as the independent variable. In doing so, this study will contribute to our understanding of the effectiveness of electoral engineering in divided societies, and specifically the use of PR-STV to foster moderation.