Research on negative campaigning has so far only dealt with few multiparty systems and has rarely examined dynamic aspects of this particular campaign strategy. Based on data from the Austrian National Election Study (AUTNES) the proposed paper puts in an additional case of a multiparty system as well as a new method that uniquely captures the dynamics of negative campaigning in the three most recent Austrian elections of 2002, 2006, and 2008. It is based on a relational content analysis of parties’ press releases, which is a type of source scholars of negative campaigning so far have only rarely used. Contrary to other data sources (e.g. media reports) press releases are under the direct control of parties. Using a dyadic data structure of daily party interactions we are able to examine the long-term dynamics of negative campaigning as well as the immediate reactions to rival parties’ attacks and other campaign events. Apart from the question of timing, the proposed paper examines which parties use this type of strategy (governing vs. opposition parties; parties expecting victory vs. parties expecting defeat) and whom they attack (e.g. rival parties’ top candidates vs. competing parties as collective organizations). Preliminary results based on data from the 2008 election (N=2,594) indicate that about 40 per cent of the press releases primarily include attacks on opponents and that party differences are significant. Opposition parties significantly send more attacking press releases but the relative share of such releases decreases over time for all parties. About 20 per cent of the attacks aim at rival parties’ top candidates, especially at the leader of the strongest party. Furthermore, negative campaigning is interactive: attacks immediately lead to counterattacks.
