It does not seem unreasonable to think that if during an electoral campaign a set of premises regarding the use of Twitter were met, monitoring mentions of candidates and political parties could give us clues about voting intentions.
The premises would be:
1/ Select “twitterers” representative of different agents and population groups that transmit and/or receive messages related to the campaign: political parties, media, candidates, political officials, journalists, communication advisors, and obviously civil society.
2/ In selecting “twitterers” from civil society and their followers, ensure that we cover the different profiles of voters. It is not an easy task since the necessary information is not always available on Twitter.
3/ Select tweets that talk about topics related to the electoral campaign, “twitterers” also have other things to talk about during an election. A good selection of keywords is basic at this point.
4/ Capture the tweets in a minimum period of 8-10 days during the campaign, since the actions of the parties and their communication can change the state of opinion of the voters as the elections approach. Even external factors can have an impact on voting intention, let's imagine how the 2004 Spanish general elections would have been followed on Twitter.
Following the previous premises, we have analyzed the use that has been made of Twitter in the electoral campaign of the general elections in Andorra on April 3, and the results have been the following:
The buzz generated on Twitter during the electoral campaign anticipated the results of the April 3 elections. Comparing the results of the elections with the mentions of each party and each head of the list in the analyzed tweets, a very high correlation was observed, especially for the two parties with the most votes:
‒ The Demòcrates per Andorra coalition and its leader Toni Martí, who were the winners of the elections, were also the party and the candidate that obtained the most mentions on Twitter during the electoral campaign.
‒ The Partit Socialdemòcrata and its leader Jaume Bartumeu, who came in second position in number of votes, were also the second party and the second leader of the list most mentioned on Twitter.
‒ Furthermore, the difference in votes between the first and the second, 37% fewer votes for the Partit Socialdemòcrata compared to Demòcrates per Andorra, is very similar to the difference in mentions on Twitter, 34% fewer mentions of the Partit Socialdemòcrata compared to Demòcrates per Andorra, and 32% fewer mentions of Jaume Bartumeu compared to Toni Martí.
‒ The rest of the candidates Eusebi Nomen (Andorra pel Canvi) and Isabel Lozano (Verds d'Andorra) obtained a higher percentage of mentions on Twitter than the percentage of votes in the elections, but in both cases well below Toni Martí and Demòcrates per Andorra, as in the results of the elections.
But despite the previous results, can it be firmly stated that whoever “tweeted” the most and best won the elections?
It is still premature to say, in our opinion it will depend on the growth in the use of this social network and its representativeness. In any case, with respect to the last general elections in Andorra, the results were very similar.
Below we attach the complete study on the use of Twitter during the electoral campaign for the general elections in Andorra on April 3, 2011.
Francesc Camp
Hamilton Business Intelligence Unit Director
