Using our proprietary sample and process methododology, the 4th Estate Project has found an interesting correlation between campaign VoiceShare in the media until a day before the election and the electoral result.
Correlation between Candidate VoiceShare in Media and Popular Vote
(CLICK on chart for a lightbox view.)
Candidates’ VoiceShare in-state media contrasted with electoral results shows a clear correlation both in the net result and the order of the candidates. In the data, we found an average variation of 5.3% between electoral result and campaign media VoiceShare for four handpicked recent elections.
For example, Gingrich’s campaign was the leading content generator in the Georgian media with 56% VoiceShare which he won comfortably with 47% of the votes. The ranking of Gingrich’s VoiceShare in the 4th Estate data coincides with this electoral result. In the case of Florida the correlation is even clearer. Romney won and received a VoiceShare of 41% the day before the election while getting 46% of the votes. Gingrich, the runner up with a VoiceShare of 36%, got 32% of the votes. All candidates contrasted had a proportion of votes within 5% of the VoiceShare given by the media in each state.
This correlation has some variation that require further analysis. For example, there is a massive difference in Gingrich Campaign VoiceShare in Illinois when compared to its electoral result. He had much larger 21% VoiceShare in the media while only garnering 8% of the primary votes.
The case of Massachusetts (not shown in data) is also interesting. If any media would have given the winner a VoiceShare similar to the electoral result, it would have been the result of a biased and unfair process. Political history may explain better the result in Massachusetts.
In the 4th Estate project, we will continue to explore data that can inform the electoral process. It seems clear that media VoiceShare has a high correlation to electoral votes in the political primary process.