[8] Shanto Iyengar similarly argued in 2005 that while campaign strategies can have an effect, "The results of presidential elections can be predicted with a high degree of accuracy from indicators of economic growth and public approval of the incumbent administration.
"[9] Mark Pack, a British politician and former campaign manager, noted that in 14 of the 16 United Kingdom general elections from 1964 to 2019, the party leading most polls in the previous January subsequently won the most votes.
[10] A 2018 study in the American Political Science Review found that campaigning methods do not usually influence an election outcome, and can only do so under specific conditions.
For example, if a poll shows a third-party candidate with low support, some voters might avoid backing them to prevent a spoiler effect.
Horse-race spin, the degree of media coverage implying a candidate is gaining or losing political support, is associated with the 1988 presidential election.
[15][16] Horse race coverage, and election forecasts in particular were cited as a potential factor in Donald J. Trump's surprise victory over Hillary Clinton.
Horserace coverage, forecasters, and polling in general drew criticism from many different sources in the wake of the 2016 election.