Examining Biases in the Single-Member District Tier of Taiwan's Electoral System

Two elections to the Legislative Yuan have been held under a mixed parallel system. While there have been criticisms that this new set of rules leads to a considerable disparity between parties' vote and seat shares in the district tier, in so far as the new electoral system has been accepted and therefore treated as given by both parties and voters, its fairness should be assessed not by the degree of proportionality, but rather by examining various sources of potential bias. These include differences in electorate sizes and turnout rates across districts, and the efficiency by which votes for the main parties' candidates are distributed. The present study investigates how 'fair' the functioning of Taiwan s new mixed parallel system was in the 2008 and 2012 legislative elections by simulating equal and reverse vote scenarios at the district level, and measuring the magnitude of each component of electoral bias. The results show that the operation of the electoral system entails no marked partisan bias, since it does not consistently confer an advantage to either of the main parties or camps.