Ilkka Ruostetsaari and Mikko Mattila
The purpose of this chapter is to examine the factors that influenced the electoral success of candidates in the Finnish parliamentary elections of 21 March 1999. The Finnish electoral system uses open lists, where a candidate's election is determined on the basis of a personal vote (a preferential system). In other words, the party organizations cannot prioritize the rank order of candidates; the voters alone decide their fate (e.g. Gallagher et al. 2001:314). Thus the system emphasizes candidates' personal qualities, available resources and campaigning efforts. In fact, the open list system combined with the d'Hondt method of allocating the seats means that the most serious competitor for a candidate may well be from the same party, not from other parties. Compared with systems with closed lists, the Finnish system encourages individual candidates to invest much more time, effort and money in their election campaigns to secure their success. This is why Finnish election campaigns can be called candidate-centred as opposed to party-centred campaigns.
The main research question in this chapter is how much do campaigns matter in candidate-centred elections? We use survey data from an electoral district in the Tampere region of Finland collected during the 1999 election campaign. We analyse factors influencing the votes for individual candidates. Three rival models of possible influence on a candidate's personal vote are tested, two of which place stress on the candidates' campaigns. The first model emphasizes the candidates' prior political experience in determining their electoral success. Political experience is based on such things as candidates' status as incumbent MPs, or municipal councillors, or their participation in previous elections. The remaining two models focus on candidates' campaigns. While the second model concentrates on organizational aspects of the campaign, the third model's focus is on the candidates' access to advertising resources. We investigate how candidates' advertising expenditure in newspapers, on television and on radio are related to their election result.
We start with a discussion about the Finnish electoral system. Then we introduce our three models and evaluate their merits based on descriptive univariate and bivariate analysis. In the final part of the chapter we use