Guest post by Peter Mair (posted by David Farrell, Feb 28, 2011)
This election goes down not only as the most volatile in Irish democratic history, but also as one of the most volatile elections in postwar Europe. Aggregate electoral volatility is conventionally measured with the simple Pedersen index, which adds the absolute values of the aggregate gains of all winning parties to the aggregate losses of all losing parties, and divides by two. This measure of net volatility obviously underestimates the total amount of change, since party A’s losses to party B can be offset by its gains from party C. But we need individual level panel data to measure this type of gross volatility, and these data are not always easily or quickly collected. Measuring aggregate (net) volatility with the Pedersen index has the advantage of allowing for a more or less instant assessment, and of being able to compare levels of change in contemporary elections with those in the more distant past. Continue reading →