LAUSR

A classic concept going back to work in the 1980s by Budge and Farlie (1983) and Petrocik (1989), issue ownership is currently booming in political science research (Walgrave et al., 2015). Issue ownership refers to the link between issues and parties in voters' minds. In other words, issue ownership relates to an evaluation voters make of parties with regard to their dealing with particular issues. Parties ‘own’ issues if voters consider them best able to deal with a given issue, voters then consider the party to be the most competent and/or committed to tackle the issue. Traditional examples are green parties who own the issue of the environment or left-wing parties who are perceived to be most competent/committed to deal with welfare issues (Seeberg, 2016). A recent debate in the issue ownership literature distinguishes the competence dimension of issue ownership from its commitment dimension. Scholars are increasingly qualifying their claims by talking about ‘competence issue ownership’ on the one hand and ‘associative issue ownership’ (referring to the commitment dimension) on the other hand (Walgrave et al., 2012). Applied to elections, issue ownership theory, both of the competence and the associative variant, states that parties see their vote share increase if the issues they own become more salient among the electorate. Numerous studies yielded direct or indirect evidence underpinning this claim (see for example: Budge and Farlie, 1983; Petrocik, 1989; Ansolabehere and Iyengar, 1994; Petrocik, 1996; Bellucci, 2006; Bélanger and Meguid, 2008; Walgrave et al., 2012; Egan, 2013; Lachat, 2014; Lanz, 2017). Issue ownership, both of the competence and associative kind, seems to matter for the vote. But how come that voters consider a certain party to be most competent and/or committed to an issue? How does the identification of parties with issues come about? While issue ownership is firmly established as a driver of the vote, the question where issue ownership itself originates from only started to enjoy scholarly attention recently (the two first papers probably being: Holian, 2004; Damore, 2004). The growing body of work on the origins of issue ownership points to several roots of issue ownership, the main two being parties' own attention to issues and voters’ party preference. While most of the work on the origins of issue ownership focused on the competence dimension of issue ownership, this paper deals with its associative dimension. Drawing on a novel research design confronting Belgian voters with a very large number of specific issues randomly taken from the news media, we revisit the two existing explanations of issue ownership (attention to issues by parties and party preferences of voters). The work on the origins of issue ownership is plagued by endogeneity issues. Most importantly, the key debate is whether issue ownership perceptions explain voting or whether the opposite is the case with voting explaining issue ownership perceptions. Our design working with specific media issues has the major advantage that parties have not been able to build a long-term reputation for commitment with regard to many of these specific problems and events. This allows to directly compare specific issues on which there is information available for voters about parties' attention to these issues with situations in which such information is largely absent. Concretely, we examine whether such a low-information context triggers citizens to rely more on heuristics, such as their party preference, when attributing associative ownership to parties. In other words, our design focusing on specific issues allows to tease out under which circumstances voters rely on heuristics instead of on real information about parties’ attention to issues. We find that associative issue ownership perceptions are partly driven by real information about parties’ actual attention to issues. Concretely, people tend to consider parties more as issue owners if those parties have been covered more in news coverage on those issues. This is good news for the usefulness of issue ownership as a predictor of the vote as it basically means that associative issue ownership is not entirely endogenous with party preference. Yet, instead of driving the vote as decades of research has argued, the exact opposite happens as well: people prefer a party first and then they attribute ownership to it when a new issue shows up. The partisan heuristic is powerful, and even more so when the issues are important to voters. Moreover, specific associative issue ownership is also inferred from more general, issue domain ownership perceptions. When being asked about their associative issue ownership perceptions with regard to specific issues, voters tend to use their ownership perceptions regarding broader issue domains and apply these general perceptions to specific issues. In other words, there is a second heuristic at play as well whereby the general is applied to the specific. The key contribution of the study is that we show for the first time that the use of the partisan and issue domain heuristic varies with the