MobileActive.org

On November 8, 2011, the Liberian President Ellen Johnson Sirleaf won her re-election campaign following a contentious runoff vote. In the October 11 general election, neither of the top two presidential candidates secured a majority vote –Johnson Sirleaf received 43.9% of votes and opposition candidate Winston Tubman received 32.7% of the nation’s votes. Johnson Sirleaf and Tubman were scheduled to participate in a November 8 runoff election; however, Tubman boycotted it saying that the first elections had been unfair; a claim international election observers dispute. As the only candidate, Sirleaf won the runoff despite a low 37.4% of eligible voters coming out for the second round (compared to more than 70% for the first round).

In light of the election’s tumult, MobileActive.org spoke to the National Democratic Institute and Ushahidi Liberia to learn more about their respective work in the country encouraging transparency and fairness through election monitoring and citizen reporting. 

The National Democratic Institute and Ushahidi in the 2011 Liberian Elections

Elections can be rigged in many ways, and voter fraud is varied. For instance, ballots can be changed or manipulated, voters can be influenced through intimidation or bribes, violence can shut down polling stations, or ballots can be changed after the election before the results are announced. Technical difficulties can also influence an election by preventing voters from casting their votes or having those votes accurately counted; difficulties could include long lines, failure to open a polling place on time, or a lack of necessary supplies.

Collecting data on the election process both in the pre-election period and during electon day itself is needed to ensure that elections are indeed conducted in a free and fair manner. There are two main approaches to gathering reports around elections these days: systematic election monitoring and crowd sourcing citizen reports. In the 2011 Liberian presidential election, the National Democratic Institute (NDI) and Ushahidi worked on systematic election monitoring and citizen reporting projects respectively in Liberia.  

Election monitoring can effectively assess the quality and conduct of an election using a set of established standards and by using trained local monitors to record and report on the election process. Election monitors are stationed at a statistically significant and randomly selected number of polling stations all around the country. Using SMS, they send throughout the day frequent structured reports based on a checklist of criteria. These include, for instance, whether polls opened on time, whether there were enough ballots, whether voters on the voter list were allowed to vote, and whether there were any security-related instances. In some countries, election monitors also report on the actual election results as reported in the end-of-day count at the polling station in a process known as 'Parallel Vote Tabulation."  Because monitors are located at thousands (or more - a statistically significant number) of randomly selected polling stations, results indicate with a high degree of confidence the actual conduct of an election and the actual results if a parallel vote tabulation is conducted as well.  

Crowdsourced citizen reports typically come in from untrained sources who comment on what they are seeing and experiencing. In the case of Liberia, Kate Cummings, project manager of Ushahidi Liberia, describes the Ushahidi Liberia approach to the elections as “crowd seeding,” as the organization collected information from many sources, including from trained and trusted partners - eleven local NGOs.

NDI provided support and training to a coalition comprised of two local Liberian NGOs, which then, in turned, trained domestic election monitors to send to polling stations and report back on what they saw. Nate Evans, NDI’s Senior Program Manager for Liberia, explains that that the local NGOs [the Civil Society Organizations National Committee for Election Monitors (CSO-NEC) and the National Coalitions for Elections Monitors (NACEM)] formed the Elections Coordinating Committee (ECC) to have one cohesive group doing election monitoring across the country. The group deployed 1725 Liberian election observers to all 15 counties, as well as 76 additional observers to 19 electoral magistrates. He explains the difference between crowdsourcing citizen reports and systematic election monitoring, and why the ECC chose to use election monitoring rather than crowd sourcing by saying:

They're very different. Crowd sourcing is a good tool for collecting a lot of information from a lot of different places, but the people aren't necessarily trained on what to look for or they may not have a system in place as to what's important and what's not important. Our election observers were trained by our partners, the ECC […] For the October 11th elections, they recruited [roughly] 2000 observers who were placed throughout the country at polling places and they held trainings for each of them. So what they did is they would bring together 32 coordinators who we trained and went through a list of about 28 questions of what to look for while in the polling station. […]

Throughout the day the election observers filled out that checklist and then at the end of the day they call in and report those results so that they next day the Elections Coordinating Committee can hold a press conference and actually have data to back up their claims. That's the difference. Crowd sourcing is more about people calling in and telling whoever is collecting the information what they saw, whereas election monitoring is a bit more rigorous and systematic and it's just focused on specific issues that are already tested out.

Since its launch after the 2008 Kenyan elections, Ushahidi has been used to map crowdsourced citizen reports around the world. User-submitted reports are logged on a public, online map. Ushahidi Liberia was founded in March of 2010 as part of a project to focus on early warning and conflict resolution. Although the original plan was to stay for six months, that early project turned into a long-term, in-country project focused on developing a technology lab to offer resources (such as access to computers and Internet connectivity) to domestic partners to use Ushahidi.

In the Liberian elections, the Ushahidi team took a different "crowd seeding" approach and worked hands-on on the ground. Cummings explains that the use of Ushahidi in the recent elections was a departure for the program saying, “This is a real anomaly in the way that Ushahidi works, normally we work on the tool from afar and are privileged to watch how others implement it and deploy it, so this is kind of a special case where we're on the ground as active deployers of the tool and it's been so informative for us.”

The different organizations verified the information among themselves, then posted the information to an online map, with many of the organizations collecting very different types of information from their field reporters. For example, IFES used it as a monitoring and evaluation tool for their small civil society partners in the field to make sure they were doing awareness-raising campaigns and activities on how to vote and register. Those reports were very specific to their organizational programming and for tracking information.

Other organizations like the African Elections Project used the platform to gather reports from their journalists in the field. And the Elections Coordinating Committee used checklists to report on events at each of the country’s polling stations (asking questions such as how registration rolls matched up to the voters, if elderly, disabled, and/or pregnant voters were provided assistance, and if votes were counted in full view of party agents and monitors), more akin to the way systematic election monitoring is conducted. 

Kate Cummings explains that the decision to use a crowdseeding rather than crowdsourcing approach was done to simplify the process for participating partners. She explains:

Our choice to crowdseed rather than crowdsource was something that came from our partners and was an effort to simplify the verification process to some degree; because trust is such a big part of communication here especially with new technology, the trust networks are really the way to go to make something feel more comfortable. So, we thought that we'd start there and see maybe in the future if folks would like to expand and open it up a bit more beyond trusted reporters, then we're very open to that.

Most of our partner organizations, if they're getting information from the field or from somewhere outside their office, felt much more comfortable that that was coming from their own staff or organizations that they work with that they trust. And they would rather go through a verification process with that information coming from trusted reporters than they would to have that information coming from just the crowd at large that there's already an inherent distrust in Liberia towards. So that was the reason for doing crowd seeding, to make our partners feel more comfortable with it.

Mobile Technology and Liberia

Liberia is still a difficult country in which to deploy tech projects – in 2009 it was ranked 193rd out of 216 countries for lowest Internet use per capita by the CIA World Factbook, and in 2010 the number of mobile subscriptions was estimated at 1.57 million out of a population of 3.87 million. Landline phones are rare outside of the capital. These limitations shaped how the ECC, NDI, and Ushahidi Liberia developed their systems to gather data from field reporter.  They used a combination of mobile phones, landline connections, and online communication to collect the most information quickly and reliably.

NDI's Evans explains,

“There was a big technological component to this program. Liberia is a challenging environment. It's a very poor country, infrastructure is bad, roads are bad, communication networks are difficult, there's not cell phone coverage everywhere around the country – it's in most places but not everywhere. So it's a tough country to do this kind of work in because this work requires people to be able to call in from everywhere and report results."

The ECC trained 1801 total observers for election day and gave each of those observers phone credit. 300 of them also called in the data, choosing the 300 from areas with strong mobile phone coverage. The organization used the sample from the callers to get an idea of what happened on election day for immediate use, and then continued to collect reports from the field workers during the rest of the week to build a complete picture of the election, even from areas without mobile coverage.

The ECC and NDI also used Google Spreadsheets to create an online repository of the reports that came in and used Ushahidi’s iLab for 20 data entry clerks who used the Google Spreadsheet to collect the checklists and enter in the data. On election night, they also set up a call forwarding system in the iLab in a partnership with local telecommunications company Cellcom to receive calls from election monitors and automatically forward them to any available phone in the room.

Says Evans, "the reason for that is we didn't want that data entry clerk to step outside the polling place and spend twenty minutes trying to get a call to go through."

Ushahidi partnered with local telecoms (Lonestar, Cellcom and LiberCell) to offer free SMS services via a shortcode (2011 from Lonestar and Cellcom) and a regular number through Libercell so that field workers could send in updates at no cost to be added to the map. The variety of reports include security issues, police actions, polling station issues, ballot issues, voter issues, positive events, and violence. The map has been running online since before the first elections and currently has over 5000 reports on the map, the largest contributor being the ECC.

On the first election round, Evans and Cummings say that most of the reports came in smoothly. However, Cummings said that a lesson they learned from the first round of election reports was that many of the field workers were waiting to only report problems. As the original election went fairly smoothly, there wasn’t a lot of information coming in on election day because the unlike the ECC that reports on the conduct of the election systematically, Ushahidi partners were only looking for problems.

After the election, civil unrest grew in the subsequent political tussle. Cummings says, “The reports we've gotten on the site since the election have mimicked what's been happening here on the ground. There's unfortunately been a devolution into a few hotspots of violence and real discord between the parties. So we've seen that reflected in the messages.”

Challenges, Lessons Learned, and Future Plans

Between the first election and the runoff, the Ushahidi blog responded to criticism about a lack of real time updates on the site. After the first round of elections, Ushahidi Liberia added a disclaimer to the map explaining many of the issues, the main two being the decision to crowdseed information from partners rather than crowdsource it from the public in real time, and the lag time between where reports were filed and then verified before going live on the map.

A disclaimer added to the Liberia map explained both the decision to crowdseed (“Given the potential for unreliable information and volatility from the crowd in Liberia, we have intentionally sought information from partner organizations whom we have trained in reporting to the platform – a crowdseeding approach rather than crowdsourcing. This is a different approach than is often taken in other deployments of the Ushahidi platform.”) and the dependence on verification through partner organizations and limited connectivity issues that affected some of the reports (“Reports are not always published in real-time because partner organizations are responsible for their own messages. Given the lack of consistent access to the Internet in Liberia, not all partners are able to submit their reports to the map in a timely manner. Thus, reports may not be published quickly after the incidents described therein.”). 

In Liberia, ICT4D and M4D projects are still finding their niche as the country works to rebuild its infrastructure. Cummings says that a major challenge for ICT4D and M4D projects in Liberia is the lack of technological infrastructure and a lack of access to tools in rural areas. She says that bridging this gap is a big goal for Ushahidi Liberia in the coming months as they work to reach out to citizens who can’t access Ushahidi through the traditional web portal. She says:

We've been working, slowly, on ways to try and get this map back out into the field, because it really does live online and within these organizations. So we've created a "print map" function that allows us to print out any customized view of the map with any combination of categories or reports. And already groups like IFES and the peacebuilding office here at the Ministry of Internal Affairs have been using those to take out in the field and show field reporters for the first time 'this is what your reports look like on the map.'  So that's more of the direction we want to move in, so that information can be distributed to the people who are spending really valuable time sending it in can see what that looks like. […]

I think they want it to relate to their work, and that's a real challenge because even just simply using mobile phones as a way to interact with this tool is so new and does not have immediate benefits in their minds because they don't really have interactions with the map very much. So I think it feels like just one more information being sent off into a black hole and one more top-down 'let's gather information and send it away' type of initiative, so we'd really love to change that and to increase the number of field trainings. So I'd like for them to feel like we are not in some distant land and are with them in this project and we want to know how to improve it with them.

For NDI and its local partners, systematic election monitoring and gathering in-country reports are a vital part of the voting process. Explains Evans, “After an election, it's hard for any one voter or any few voters to know if it went well all over the country. Even if it went well in their polling places doesn't mean it went well on the other side of the country. So the Elections Coordinating Committee uses its data and it can serve as the eyes for the population.”

Cummings said that she found that other eyes were turned on Ushahidi as well, despite the limitations of in-country Internet access. Google Analytics on the Ushahidi Liberia page found that most of the page views on the map were coming from areas with high concentrations of the Liberian diaspora, showing that although many in-country citizens can’t access the map online, out-of-country Liberians are looking for information about their country.

As re-elected President Ellen Johnson Sirleaf works to help the country heal from the tumultous election, having multiple reports that the election was free and fair will help maintain legitimacy for the democratic process.

Edited by Katrin Verclas.