One of the largest parallel vote tabulation efforts is under way right now in Nigeria to observe and monitor gubernatorial elections. Project 2011 Swift Count is observing the national assembly, state assembly, and gubernatorial elections with 8000 trained election observers. The observers are also working across six high-priority states to complete parallel vote tabulations to verify the official results in six corresponding gubernatorial elections.
At around 9pm in Nigeria on election day, MobileActive.org spoke with Chris Doten, senior program officer on the ICT team for the National Democratic Institute (NDI). When we spoke with Doten, two-thirds of trained election observers had sent in via text message final vote counts, meaning the final tallying was complete at that particular polling place. A much higher percentage of observers have sent in summaries of total ballots cast.
NDI is providing technology assistance and is partnering with Project 2011 Swift Count, a coalition of 4 civil society organizations pulled together for this specific election. “They are all united in trying to make sure that Nigeria has a clean election this year,” Doten said. According to the web site, the Swift Count vision is a Nigeria where elections are free, fair and peaceful as well as viewed as credible and legitimate by its citizens.
Project 2011 Swift Count in Nigeria employs statistical principles and well as ICTs to observe the electoral process. Trained, accredited non-partisan observers are deployed to a representative sample of polling stations in Nigeria and transmit reports to the national information center in Abuja, via basic mobile phone and encoded text messages.
Speed, volume, and mobile technology
In Nigeria, the team needs speed and volume, Doten said. In terms of speed, paper wouldn’t work, as one goal is to have information in hand before the national election commission does. In terms of volume, a call center couldn’t be staffed to handle 8000 people calling in multiple times a day. “Text messages were a natural fit,” Doten said. “Everybody is very comfortable with texting. It’s very natural.” It’s also an approach NDI has used in other elections.
The trained observers use their normal phones, and send encoded messages along with polling place information. Each observer has an individually printed checklist with information on their specific polling place and a series of questions. At various points, the observer sends the answer to the question along with election reporting messages, all via an established shortcode.
A gateway provider pulls from the five different major providers in Nigeria, and the team’s servers pull from there using RapidSMS. The RapidSMS server is housed at a co-location facility in Europe to try and avoid connectivity issues in Nigeria.
After an observer texts in a report, a return receipt is sent with confirmation or potential errors. If the information parses successfully, this relay can happen in 10 to 15 seconds, but bottlenecks can occur. Doten said that today, there were a few bottleneck issues and 8 minute lags were observed between the time the observer sent a reporting message and received a return SMS receipt. Observers have been trained to wait up to two hours before re-sending a report to prevent a further bottleneck to the system.
Observers work in a tiered model, Doten explains. Individual observers report to a local government area (LGA) supervisor. Observers come from the areas they are monitoring and are a pat of the community they are observing. There are 774 LGAs in Nigeria and the Swift Count partners tried to get observers in each and every LGA across the country for a greater breadth of coverage. These LGA supervisors report to senate districts, state directors, and state coordinating committees. This, in turn, funnels up to the national information center in Abuja.
A human touch
Doten, NDI, and Project 2011 Swift Count have kept a team blog to track the process of the election observation and parallel vote tabulation. The live blog provides a very human, very timely, and very interesting account of what is happening on-the-ground and with the observers.
This account from today is about RapidSMS earning its name:
Once the messages hit our server we can watch them pretty closely, checking flow rates on our remote RapidSMS server (proving its name today) and bandwidth utilization from our VSAT. In the data center I’m sitting like a mother hen over our "servers" and making sure that they’re not running out of processing power, RAM, or bandwidth, and that our local network is staying up. Actually, everything has been pretty darn stable today.”
We’ll follow up with Doten and others and report back on results and lessons learned along the way.