It is not Always the Same: The Effect of District Magnitude Throughout Various Kinds of Electoral Systems

The belonging of district magnitude within key variables of electoral systems is a commonly shared piece of knowledge in political science. The growing number of mandates divided among candidates during an electoral race is strongly associated with increasing proportionality of an electoral system and it decreases the strength of strategic incentives attacking voters in order to avoid wasting their votes (Cox and Shugart, 1996; Cox, 1997). This presumption is inherently present in the way we think about consequences of mechanical arrangements of electoral systems, even though we should at least be curious about the uniform applicability of this rule. The Kenneth Benoit’s analysis (2001) revealed that electoral formula is a relevant attribute in considering the effect of district magnitude, however, there still lacks this kind of a broad cross-national analysis revealing the clue on what to expect from variable amount of allocated mandates throughout various categories and variations of electoral systems. The presented paper will, using slightly extended data of the Constituency-Level Elections Archive (Kollman et al., 2014), analyze the variability of outputs generated by the same electoral mechanisms with varying district magnitude with special attention paid to the strategic coordination of voters, proportionality, and the reduction effect towards party system. The preliminary results have shown that the effect of district magnitude differs across the systems with proportional and majoritarian electoral formulas. When the results from low and medium magnitude districts are compared, the group, with majoritarian rules implemented, is exhibiting much larger differences in strategic coordination as well as reductive effect towards party system. The increase in proportionality of outcomes is similar for both groups. There are countries included in the analysis, where in different districts both formulas coexist revealing the tendencies between both groups, which supports the validity of these preliminary conclusions.