ABSTRACT
The outcomes of European elections in terms of party choice differ much less from what would have occurred in concurrent national elections than does turnout. Nevertheless, some parties do gain and some parties do lose compared to what would have occurred had the elections been general elections-often by as much as 6% of total votes. These gains and losses occur for two reasons. Some parties gain (lose) from the fact that they are more (less) successful in getting their normal supporters to the polls. The low turnout in European elections helps (hurts) some more than others. Other parties gain (lose) because they are supported (abandoned) in European elections by voters who in national elections would have voted differently. For some parties these two processes cancel out, leaving them with about the same level of support that they would have enjoyed in national elections; for others the two processes reinforce each other, leading to quite dramatic divergences from national election outcomes. However, by far the greater of the two effects is the effect of people voting differently which is generally more than twice as great as the turnout effect, and can involve more than 40 % of those voting (at the 1994 and 1999 elections in France and Denmark, and at the 1999 election in Italy), as shown in Table 10.3. Note, however, that this total amount of what has been called ‘quasi-switching’ (van der Eijk, Franklin et al. 1996) involves a great many contrary movements that cancel out in aggregate, leaving net effects of much lower magnitude (seldom more than 6%, see above).
