Elite Hacking: Public Opinion Formation and Elite Fragmentation

Abstract

In the 2016 U.S. General Election, agents of the Russian Federation attempted to influence the election through a multi-faceted campaign, which included the spread of of fake news as well as cyber attacks on campaigns, individuals, and even electoral infrastructure. Elite reaction was swift. Unlike on most issues, however, elites in the Republican Party remain divided. Leveraging this rare and enduring intra-party disagreement in today's age of polarized parties, we are able to begin to understand how people react to conflicting elite cues that are free from usual partisan considerations. This enables us to present a theory of elite-led opinion formation that includes intra-party polarization, expanding on the current literature. We theorize that in a scenario where there is a divided party and a unified party, those who identify with the unified party will form crystallized opinions that are reflective of those held by party elites. Members of the divided party will have perceive conflicting elite cues, and form opinions that are more malleable and susceptible to change from future elite cues. We test this theory by examining the impact of salience of Russian hacking as well as different levels in elite cohesion on belief that Russian hacking actually took place heading into the 2018 election. This is carried out through a two-pronged approach, leveraging a nationally-representative online survey (N=1000) and two survey experiments. We find that different levels of elite agreement on Russian hacking, rather than the information about Russian hacking itself, caused respondents to update their perceptions of Russian hacking. This finding is particularly strong for the Republican subsample in our experiment. We also find strong evidence that supporters of Trump are resistant to information, going as far as to demonstrate the opposite effect when presented with information that directly contradicts the President's position. Furthermore, we find similar results regarding elite influence on opinion for trade policy, a much less salient issue in comparison to Russian hacking. In doing so, we not only provide new evidence on the impact of disourse around Russian hacking, but add to a growing literature on elite discourse to include intra-party discourse and attachment to groups of the elites within a single party.