Conference Paper SOEP 2010, 9th International German Socio-Economic Panel User Conference, held June 20-July 1, 2010, Berlin
Persuasion effects in electoral campaigns: a comparative analysis of household panel data
30 Jun 2010
This paper studies campaign effects of national elections using household panel surveys from Germany, Great Britain and Switzerland. Because household panels question every year the same individuals about their party preference, they offer a unique opportunity to study individual dynamics over the electoral cycle. More specifically, they allow a better distinction between activation and persuasion effects of campaigns than it is possible with ordinary electoral panels, which are conducted during campaigns. Campaign effects are captured through the time difference between the interview and the election date. We add evidence to the debate whether campaigns do persuade voters to change between parties, or whether changes mostly reflect activation effects. Using random effects models with different transitions as the dependent variable, we find strong evidence for activation and persuasion effects in all three countries. Furthermore, we study the impact of political awareness on activation and persuasion, relying on theories on memory based and on-line information processing. Mostly in line with expectations, we find that highly aware citizens are less likely to be (de)activated and persuaded than little aware citizens. In Germany and Switzerland, there is no additional heterogeneity in the reaction to campaigns according to levels of political awareness. For persuasion in Great Britain however, campaigns have the strongest effect on the little aware.