Social Media Dialogue Predicts Election Outcome

Posted June 4, 2015

The advent of social media opened the doors to a wealth of possibility. Specifically, in the world of politics, social media allows the public to engage in dialogue with politicians. In a study conducted through the Nunn School of International Affairs and the School of Interactive Computing, Michael L. Best and Amanda Meng determined that social media platforms can help determine the results of an election. 

Recently, Best and Meng presented their paper, entitled "Twitter Democracy: Policy versus identity politics in three emerging African democracies," at the International Conference on Information and Communication Technologies and Development (ICTD) in Singapore.

However, this initiative goes beyond an academic paper. Best, Meng, and several other students formed the Social Media Tracking Center (SMTC), a collaboration between Georgia Tech and civil society organizations in several African countries. Ph.D. graduate Thomas Smyth and Best developed a software called Aggie which functions as a social media aggregator and collects data based on keywords from the party platforms and ethnicities.

The local organizations use Aggie to watch the reports from social media. Often, the organizations notice that action needs to be taken, such as a poll running out of materials, and can tweet a response or send someone who's on the ground nearby. Thus, this initiative has lasting benefits in the home countries.

For their paper, Best and Meng analyzed the Nigeria 2011, Ghana 2012, and Kenya 2013 elections. Meng personally traveled to the Ghana 2012 election. She arrived a week early and worked personally with the group of journalists, called Penplusbytes, who immersed her in the local culture.

"They had me copyediting, going to the political rallies, and really familiarizing myself with the political landscape and the election landscape," Meng recalled. "I was an outsider, but it was important to them for me to understand the context. It definitely helped me with the analysis in the paper."

Overall, Meng and Best found that the winning candidate had the most discussed policy material on Twitter, even in highly contested elections. Moreover, they learned that the public responds favorably when a political party correctly captures their desires, which the candidates were able to do using Twitter to promote individual self rather than solely identifying with the political party.

With regards to future implications, Best and Meng are planning on running a similar analysis on the Nigeria 2011 election and the recent 2015 election to see if anything has changed.

All in all, the collaboration with the local people who work directly with the politics is vital to the success of the project. As Meng said, "We play a role in the bigger picture and the fact that we work together as partners is so important."

Should anyone wish to run a similar analysis of another election or event in another country, Aggie works beautifully for that. As long as the important issues are identified, Aggie can track data from any specific social media platform.

Michael L. Best is associate professor at the Sam Nunn School of International Affairs and the School of Interactive Computing at Georgia Institute of Technology where he directs the Technologies and International Development Lab. Dr. Best is director of the Ph.D. Program within the Sam Nunn School. His research areas include conflict research, information and communications policy, science and technology policy, and telecommunications.

Amanda Meng is a Ph.D. candidate in the Sam Nunn School of International Affairs. Amanda is interested in how Information Communication Technology (ICT) is a tool for enhancing development and democracy outcomes. She investigates the open data and open government movements.