This paper has a distinctive empirical focus, namely to map various aspects of electoral volatility among the five Nordic countries. Different kinds of volatility estimates (votes at national (net) and local level, seats, survey data (gross)) are presented. At issue is to find out whether different estimates tend to agree or not. The selection of the relatively similar Nordic countries with their relatively stable party systems has implications for the analysis conducted in this paper; if the estimates tend to agree, it gives some support to the expectation that reliable volatility estimates are possible to arrive at; on the other hand, if there are serious problems accounted for among the Nordic countries, then the possibility of arriving at reliable volatility estimates seems to be difficult to reach. Preliminary findings do however suggest a large amount of agreement for different estimates, with the exception for a few elections in Iceland at the end of the 1990s.
