Voters’ Response to Parties Breaking Their Coalition Promises

During election campaigns, parties often announce which coalitions they might or might not form after the election. If voters take these signals seriously, they should compare the actual coalition behavior of parties during the government formation process with pre-election coalition signals. Voters then should condemn coalition promise breaking. I empirically investigate voters’ response to parties breaking their coalition promises (1) by a quasi-experiment as well as (2) by an survey experiment. For the quasi-experimental study, I employ the 1996 New Zealand election. In this election, the party “New Zealand First” clearly signaled that it would not form a coalition with the “National” but then did not keep this promise. My quasi-experimental research design exploits the fact that the day on which these two parties announced to form a coalition falls within the fieldwork period of the 1996 New Zealand Election Study. This allows me to make use of an Unexpected Event During Surveys Design. In addition, I conducted a survey experiment within the government formation period following the 2020 Irish general election. In this experiment, respondents were confronted with government formation scenarios of which some implied coalition promise breaking. Most notably, the results of the quasi-experiment and of the survey experiment suggest that voters respond to a party breaking its coalition promise by being more dissatisfied with the way democracy works. Furthermore, voters' responses seem to be conditioned by voters’ coalition preferences.