‘People are running, but where are they heading?’ Disentangling the sources of electoral volatility

Abstract:

Electoral volatility as measured by the Pedersen index is probably one of the most popular indicators in political science, but its interpretation is far from clear. Volatility is produced by a mix of party-switching, differential turnout and generational replacement. However, there is virtually no empirical research on which, if any, of the main mechanisms leading to volatility tends to have the stronger net impact on election results. Furthermore, the presence of generational replacement and its relative impact on election results have received little attention in studies of volatility.

This article develops several theoretical expectations concerning the strength that each of the three components of volatility is expected to exert on the latter. Subsequently, it estimates the net impact of each component on the results of 73 elections in six West European democracies using survey data. According to these estimates, party-switching produced 75 per cent of the total amount of volatility, with differential turnout and generational replacement producing 17 and 8 per cent, respectively. Although the effect of these components may work against each other at times, on average only 11 per cent of volatility was cancelled out this way.

Findings provide, for the first time, a map of the components of volatility in established democracies and set the ground for further research on the topic.