Fragmented outlets, fragmented voters?

Do digital media fragment party systems? According to the fragmentation hypothesis, the increasing number of media outlets resulting from the digital revolution has forced media to offer specialized messages and carve new niche markets (Hamilton, 2005; Mullainathan & Shleifer, 2005; Coe et al., 2008; Peters, 2010). This may pave the way for new cleavages to appear or reinforce, isolating citizens from mainstream politics (Bimber 2008). As a result, issue-oriented groups and small parties take advantage of this situation to break official narrative and gain momentum (Masket 2009; Layman, Carsey and Horowitz 2006). The equalization hypothesis also suggests that digital media consumption benefits small parties, since the Internet allows web presence to all political organizations regardless of monetary investments (Mazzoleni & Shultz, 1999). Recent studies find some support for these expectations; they show that Internet use during political campaigns increases voters’ uncertainty (Sudulich, Wall &Baccini 2014) and polarization (Mancini 2013). Yet, Bimber (2008) posits that the right question it is not whether the Internet leads to political fragmentation, but what features of new media force for fragmentation, and which ones facilitate integration. This paper asks whether digital media contribute to vote fragmentation. At the individual level, this means that voters following an electoral campaign through digital media should be more prone to change their mind during the campaign, and to move from big national brands to small, specialized parties. They should also experience higher levels of uncertainty in the process. This paper uses data from the Making Electoral Democracy Work project and from a seven-wave Spanish online panel in order to test the fragmentation hypothesis. Our goal is tracking vote change comparing voting intention in pre-electoral surveys with vote reported in post-electoral surveys in national elections. We will also track voting change by means of within-individual evolution of parties’ voting probabilities. The goal is ascertaining if overall change and change benefiting smaller parties are related, ceteris paribus, to digital media use during the campaign. We will also address whether all digital media exert the same impact or some (i.e. social media) are more “fragmenter” than others (i.e. traditional newspapers’ webs). Finally, we will test the mediating role of uncertainty between digital media use during the campaign and voting intentions change.