How Election Polls Shape Voting Behaviour

Abstract

This article investigates how election information such as opinion polls can influence voting intention. The bandwagon effect claims that voters ‘float along’: a party experiencing increased support receives more support, and vice versa. Through a large national survey experiment, evidence is found of a bandwagon effect among Danish voters. When voters are exposed to a news story describing either an upwards or downwards movement for either a small or large party, they tend to move their voting intentions in the according direction. The effect is strongest in the positive direction – that is, when a party experiences increased support, more follows. Consistent effects are found across two different parties for a diverse national sample in a political context very different from earlier research on the bandwagon effects. Considering previous research and the fact that evidence is not found that suggests that the effect of polls vary across sociodemographic groups, the results imply that bandwagon behaviour is based not on social or political contingencies, such as media or political institution, but on fundamentals of political cognition.

The Winner-Loser Spiral in Political News Coverage: Investigating the Impact of Poll Coverage on Subsequent Party Coverage

“…Moreover, the finding that it appears to be positive, and not negative poll coverage that affect subsequent coverage is also in line with a number of recent studies of exposure to poll coverage on voters conducted in multi-party systems. For example, both Dahlgaard et al (2017) and Van der Meer et al (2016) have found that framing of positive poll stories increase vote intention for parties, while there were no corresponding effect for negative stories. Both these two explanations should, however, be viewed as provisional, and something that would need to be evaluated on a new empirical material.…”

“…It is worth noting that evidence of the bandwagon effect has been mixed (Moy & Rinke, 2012). However, considering the extensive and persistent interest in how poll results might influence opinion formation (Rothschild & Malhotra, 2014;Toff, 2018) and political behavior among voters (Dahlgaard et al, 2017;Stolwijk et al, 2016;Van der Meer et al, 2016), it is somewhat puzzling that so few studies have considered how the media's publication of poll results might influence subsequent political coverage (Toff, 2016).…”

“…Over the years, there has been ample work on how journalists use, and often misuse, horse race polls by, among other things, not adhering to the statistical limitations in the underlying data (Oleskog Tryggvason & Strömbäck, 2018) and by providing disproportionate attention to deviant results (Searles et al, 2016). There is also an extensive and growing literature examining how poll coverage can influence opinion formation (Rothschild & Malhotra, 2014;Toff, 2018) and electoral behavior (Dahlgaard et al, 2017;Hardmeier, 2008;Van der Meer et al, 2016).…”

¿Quién se expone a los sondeos preelectorales en campaña? Voto estratégico y competencia electoral en España (2011-2019)

“…Hodgson y Maloney (2013) o Gelman y King (1994) han analizado cómo los votantes aprenden del pasado: los efectos podrían ser menos influyentes en su comportamiento electoral si anteriormente las encuestas han fallado considerablemente acerca del resultado final o si los votantes perciben intenciones partidistas o ideológicas para alterar su comportamiento electoral. Además, Van der Meer et al (2016) o Dahlgaard et al (2017) demuestran que los estudios acerca del bandwagon no presentan efectos a largo plazo en el comportamiento electoral de los votantes o en el sistema de partidos.…”

Think Twice before Jumping on the Bandwagon: Clarifying Concepts in Research on the Bandwagon Effect

“…This argument, considered elsewhere in proportional representation contexts (Dahlgaard et al, 2017; Irwin and Van Holsteyn, 2000; Stolwijk et al, 2017) is worth comparing with the dominance of the static approach, which is arguably more applicable in ‘winner-takes-all’ contexts (Irwin and Van Holsteyn, 2000: 9). It is apparent that the context in which these effects are studied is central to determining which type of comparison is salient.…”