Can SMS Influence Voter Turnout? 30,000 Hand-filled Polling Forms Hold the Answer
September 24, 2013 by Jennifer Cobb
Captricity extracts data for researchers at MIT on voter turnout.
In 2007, post-election violence in Kenya resulted in more than 1,200 deaths and several thousand displaced people. In the six years since 2007, the Kenyan government undertook significant reforms by promulgating a new constitution that encouraged decentralization, judicial reforms and the formation of the IEBC, an independent commission tasked with overseeing a free and fair election process.
These new institutions offered a powerful context for researchers at the Abdul Latif Jameel Poverty Action Lab (J-PAL) at MIT to explore how text messaging could inform voters about these new institutions and whether they have effects on voter turnout. The team designed a field experiment to study this and ultimately sent out 6 text messages to more than 1.8 million individuals in the country. The team included Benjamin Marx, Vincent Pons, both PhD students at MIT, and Tavneet Suri, Associate Professor at MIT, supported by research manager Asman Suleiman and analyst Eleanor Wiseman. More detail on the research project itself can be found here or here.
As a core part of this research, the team needed data on voter turnout at the polling station level. This information was contained on hand filled forms, one for each polling stream, totaling to just over 30,000 forms. These forms therefore needed to be digitized. As Eleanor Wiseman, an analyst for the team, commented: “We generally use surveyors in Kenya. And we have to enter all the data and then enter it again to get double entry quality. But we do not have the facilities and the infrastructure to manage that many forms and that quantity of data; especially as the team on the ground is trained to handle survey work, which is quite different.”
Captricity was recommended to them as a solution that could help and team lead Tavneet Suri reached out.
Suri and Wiseman were very happy with the results. The system was able to extract the information from the 30,000 forms in about a week, compared to the months it would have taken with manual entry. As Suri commented, “Captricity was very quick and flexible. We were able to tweak the forms and extract the data we needed to complete our research quickly and efficiently. One feature that has been particularly useful is the ability to click on a cell and view the original data.”
The team expects to have access to a lot more data for this research project and Captricity will be ready to help again. We are proud to be a part of this important research.
To read more about our work with social science research, click here. Captricity is pleased to offer a 15% discount for all our non-profit clients.
 Report of the United Nations Fact-Finding mission, available at: http://www.ohchr.org/Documents/Press/OHCHRKenyareport.pdf