Election monitoring in austere contexts using ICTs

The following note was penned to the co-ordinator of the Centre for Monitoring Election Violence (CMEV), who asked me to go into some detail around the information security, gathering, archival and dissemination strategies I conceptualised and deployed around the historic Presidential Eleection in Sri Lanka, held on 8th January 2015.

CMEV’s information operations are a template for the monitoring of election violence similar austere circumstances, where the independent monitoring agencies, monitors and activists are at risk of violent pushback before, on the day of and after the elections.

Also read An election monitoring SMS template, from some years ago, where I designed a template through which election monitors could report violence in a structured manner to HQ over basic SMS.

###

Screenshot_2014-12-22-16-32-25

CMEV’s operations during the lead up to the Presidential Election on 8th January 2015 included the first time use of mobile technologies as well as, anchored to usual practice, the use of web and social media apps, platforms and tools to monitor, record and disseminate election related violence. 

The context of CMEV operations leading up to election day was one of heightened anxiety, both around the violence in the lead up to the election and the overarching pervasive architecture of surveillance, coupled with the high risk of violent pushback if Mahinda Rajapaksa, the then incumbent, was re-elected against those he and his Govt perceived as somehow in opposition, or supporting Maithripala Sirisena, the leading opposition candidate. Whereas CMEV operations are geared to face physical threats and harm, the context of the Presidential Election, given the known capabilities of the State, also raised fears over disruptive operations in cyberspace, aimed at CMEV’s information gathering and dissemination. 

This led to an unprecedented level of planning around the information security (infosec) and information dissemination capabilities and capacities of CMEV. CMEV used Google Maps to record violence leading up to election day, and incidents of violence on election day. CMEV is the only election violence monitoring body in Sri Lanka that uses web based mapping to highlight key incidents. 

The domain CMEV.org was registered (moving away from the domain associated with WordPress.com), making it easier to access the main website of CMEV. A complete website revamp was anchored to making the entire site responsive (enabling the viewing of content over any smartphone or tablet) and making it easier to get to key content in Sinhala, Tamil and English. The backend continued to be WordPress and with the domain registration, added security (in the form of two step authentication) was enabled, in order to prevent unauthorised access as well as attempts to hack into the site and disrupt CMEV operations. 

Mailchimp.org was used to send out emails including situation reports, eye-witness accounts, vital background information and conduct voter awareness campaigns. 

The cross-platform secure messaging app and service Telegram was recommended for use within CMEV HQ and field staff as a secure means of communicating information. For the first time  a secure means through which to communicate with CMEV was also opened to anyone from the general public, in order to encourage whistleblowing around election irregularities, malpractices and violence. 

Also for the first time, CMEV used WhatsApp over a dedicated smartphone (and associated number) in order to push out vital updates in the form of text, audio and video. Around the 7th of January, there were two WhatsApp groups in operation, pushing out updates to around 350 individuals both in Sri Lanka and outside, from the media, diplomatic corps, civil society, academia and other sectors (to the extent that could be discerned from the numbers provided). The service was opt in, and widely publicised over social media and the web as a means of getting CMEV updates around the clock to one’s mobile. WhatsApp also leverage the fact that mobile phone penetration in the country is extremely high, with most phones sold today being smartphones.

In addition to the above, CMEV and Groundviews also launched an Instagram account, through which images around CMEV operations were pushed out. Instagram growth in Sri Lanka is skyrocketing, and was chosen as a platform to push visual updates from CMEV because of its wide reach within the country.

CMEV’s website featured dozens of updates in English, Tamil and Sinhala, including podcasts (recorded in the field as well as in the HQ over smartphones) and video. CMEV operations were also exhaustively covered over Twitter (via @CMEV), which currently features 674 followers. A number of tweets were retweeted over Groundviews (via @groundviews) and CPA’s institutional Twitter account (@CPASL), ensuring the widest possible reach for the most vital updates pushed out by CMEV. 

In addition to Twitter, CMEV was also on Facebook, which at the time of writing had over 2,000 fans. Facebook is the most used social media platform in the country, and the content featured on the site ensured that it reached a demographic that included first time voters, which was key.

CMEV also, in collaboration with Groundviews, featured and promoted the #IVotedSL social media campaign – an unprecedented attempt over social media, which went viral, to encourage voters to exercise their franchise. A range of compelling infographics (which also went viral across social media) was also produced by CPA and featured on CMEV’s website in order to raise voter awareness around the powers and functions of the Executive President. 

CMEV’s information operations on the backend was across three ISPs, ensuring that if it was blocked or barred from one ISP, we always had another channel to access the web and publish information. In preparation for election day, access to the key CMEV accounts (Twitter, Facebook and website) were also given to a trusted source based outside of Sri Lanka, in order to keep information operations active to the extent possible in the event of a complete internet and web blackout from Sri Lanka, aimed at CMEV and other similar institutions. We also hardened security over key social media accounts by enabling two-step authentication, minimising the risk of unauthorised access and hacking attempts. Staff were provided with the latest smartphones running Android, and with WhatsApp and Telegram as the default mode of information exchange over SMS and voice calls. HQ operations were meticulously planned to run backups of critical data, with offsite data storage as well as cloud based operations all geared to ensure business continuity in the event of arson, internet blackout or worse. 

An election monitoring SMS template

I was speaking with David Kobia from Ushahidi today and remembered that my post on election monitoring in Sri Lanka (Election monitoring using new media: Notes from my experience in Sri Lanka) had forgotten to mention the SMS template that I had developed for election monitoring in Sri Lanka.

The template can be seen here.

It only works if the election monitoring agency (I work with the Centre for Monitoring Election Violence) has a robust and well defined set of categories the monitoring is anchored to, and trained election monitors who are well versed in this framework. In other words, this is not a template designed for, or will work in a context that untrained citizens report in election violations via SMS.

The template can easily be adapted to a range of contexts, and is Twitter friendly as well. One can also envisage other uses for such a template, such as monitoring human rights violations or the delivery of aid to IDP camps.

I developed this two years ago, in 2008, when Sri Lanka was still at war. Vast swathes of the country’s north and east were very badly connected, and so mobiles didn’t really work for election monitoring at the time. This is changing rapidly post-war, and I’m looking forward to working with CMEV in the future to develop this template further and design more robust backend, SMS / mobile based frameworks for election monitoring.

Election monitoring using new media: Notes from my experience in Sri Lanka

Ushahidi’s blog has a great post on the development of a tech toolbox for election monitoring. Unsurprisingly, it is anchored to the use of Ushahidi as a platform for election monitoring, which has been used quite effectively in India amongst other places.

Through the Centre for Monitoring Election Violence (CMEV), I introduced Web 2.0 and mapping to election monitoring in Sri Lanka around two years ago, realising at the time that the visualisation of election violence and real time monitoring could help media and civil society hold members of political parties and their supporters more accountable for their actions.

There is no other election monitoring agency in the country that uses maps, the web, or Facebook, in a comparable manner and the Government Department of Elections website remains stuck in the 90’s.

Writing in 2009 on the challenge of using new media for election monitoring, I noted that,

Right now, there’s no escaping the labour required for the task – each location and incident is entered into the map directly, no automated source from the web is used to populate maps. Helps us give as close to a real time image of the ground situation in the lead up to and on the day of election, more useful we are told by extensive feedback from local media, than a mashup that just puts unverified reports on a map along with other data streams.

At the time I started to work on using web media for election monitoring, and even today, the installation, configuration and customisation of Ushahidi requires a level of technical expertise that is well beyond every single election monitoring NGO I know of in Sri Lanka. It is therefore not easy to promote or use the system without a sufficient budget to get help in localising it and for staff training. Further, short code SMS’s for purposes of election monitoring are very hard to negotiate with all mobile service providers, given the risk averse nature of these business to monitoring that obviously clearly targets political parties including those which constitute government.

With these limitations in mind, I used the following set of tools to undergird CMEV’s monitoring operations, which continue to date.

  • Video production: Vimeo, YouTube, iMovie, Flip Mino HD
  • Mapping: Google Maps
  • Backend: WordPress
  • Podcasts: drop.io, Garage Band, Skype
  • Social networking: Facebook, Twitter
  • Photos: Flickr

Video

I use the Vimeo channel of the Centre for Policy Alternatives to host videos shot using a Flip Mino HD. The Flip records very high quality video, which I process and edit using iMovie on my Mac, saving it in a lower resolution before uploading it to Vimeo using its relatively new (and still buggy) Adobe Air based uploader. I have been forced to use YouTube (e.g. this video) when on the day of an election, Vimeo has temporarily halted uploading and site functionality for routine maintenance.

There is a high degree of interest in these videos. For example, 4 videos uploaded to the CMEV site during the course of the Presidential election day on 26 January 2010 were viewed over 1,400 times.

Mapping

From the get-go, I used Google Maps because of its ease of use and rich feature set. Two years ago, Google Maps didn’t have any of the street level information for key cities it now features. I used a laborious technique of getting lat / long data from one map via Google and entering this into the election monitoring map to plot incidents in cities and towns. Today, this process is made much faster and easier with the street level information that was first featured in 2009.

I have manually plotted well over 1,500 incidents over four elections since 2008, including Sri Lanka’s first post-war Presidential election earlier this year.

For each election, I have created two maps – one plotting election violence leading up to election day, and another plotting election violence on election day.

The map plotting election violence on the day of the Presidential election held on 26 January 2010 has been viewed nearly 26,000 times to date, proving my hunch that information visualised on Google Maps would be in high demand.

Plotting election violence on Google Maps also allows for new forms of advocacy against election violence through the visualisation of patterns and perpetrators, such as A map of shame: Clustering cities and regions with very high levels of election violence.

Backend

I had used WordPress for years, and when CPA’s own website in its previous avatar became too laborious to hold all the information the monitoring was generating, I created a new blog to be the hub for CMEV’s information dissemination. All the while, I kept in mind how CMEV could continue operations even if WordPress temporarily went down during a critical phase of monitoring.

CMEV’s website now holds all of its election monitoring reports, field reports, videos, podcasts and maps, with links to archived material on CPA’s own website.

Podcasts

I recorded all the podcasts in elections before 2010 using my Mac’s in built speaker, which was good enough for the job. I wasn’t aiming for broadcast quality in these recordings, but a clarity good enough to get the message across. I edited the podcasts using my Mac’s GarageBand, adding for example an intro and extro to give it a touch of professional flair.

For the Presidential Election in 2010, we did something completely new and revolutionary in Sri Lanka, using drop.io’s voice mail feature to record election monitoring updates via mobile phones. For the cost of a call to the US, we were able to use mobile phones to give out timely updates that media was able to use as well. As noted on the CMEV site and all of our press releases on election day,

These updates can be downloaded as MP3s for broadcast, listened to online, embedded on any website and social networking platform, emailed or easily linked to. Special incidents will also be covered in these updates, and to help reduce the burden on our monitors, journalists are strongly encouraged to use this feed as their primary channel of regular soundbites from CMEV.

At the time of writing, over 450 had listened to these podcasts.

Drop.io came into its own when on the day of election, just before polls closed, there was a serious question over the eligibility of a leading candidate. I was able to use Skype to record an interview with a leading constitutional lawyer (and friend) and immediately feature it on the drop.io channel for media to pick up on and use to dispel rumours and propaganda.

Social networking

CMEV is the only election monitoring agency to have a presence on Facebook and give out updates using its Twitter feed. But it is not only through CMEV’s Twitter feed that I have given critical updates during elections.

On the 27th of January, I tweeted on Groundviews from 3am throughout the day giving updates that at the time were some of the first eye witness reports on the situation unfolding on the ground, and served to dispel false rumours to boot. My article, Updates capturing aftermath of presidential elections, is a detailed look at how Twitter aided election monitoring and reporting and how its use by Groundviews was unprecedented in Sri Lankan media to report on, what was at the time, a very disturbing situation on the ground.

Photos

Instead of using a custom CMS or uploading them to WordPress, I opted to use Flickr to catalogue CMEV’s photo documentation of election violence and its work. This aspect of CMEV’s work need to be strengthened, but the photos received from the field are all catalogued here, and embedded as appropriate on CMEV’s WordPress blog (often linked to CMEV’s press releases).

Of course, that I haven’t or can’t use the tools mentioned in Ushahidi’s blog post, including the Ushahidi platform itself, is no reflection on their ability to efficiently nod effectively support election monitoring processes. It’s just that in the absence of any funding, and with my limited technical knowledge, I have almost single-handedly demonstrated in Sri Lanka without the use of any of these tools that the violence of political parties and some politicians in particular can be mapped and archived for posterity, to be leveraged by concerned citizens in civil society in on-going agitations for greater democracy.

But is technology alone enough? Even if all these tools and those in Ushahidi’s blog are used, it is any guarantee of a better, more transparent and accountable electoral process and democracy? As I note in Mapping violence during elections and voter education, albeit from a Sri Lankan perspective,

… unless awareness campaigns before an election, and advocacy campaigns after which bring to light, including name and shame, perpetrators of elections violence, these exercises alone, including my own, have little chance of really strengthening democracy. The problem with raising awareness before an election is that NGOs can never match the reach of an incumbent government’s propaganda, or even that of a political party, both of which have vested interests in keeping the public ignorant about the history of candidates and their violence.

The problem with post-election advocacy is that placing the violence of winners in public scrutiny will almost always be (a) seen as a conspiracy to undermine the legitimacy of their victory (b) cast as a rival party political bid to discredit the electoral victory and the ‘will of the people’ (c) be seen as some sort of NGO / civil society campaign to discredit the winners.

Technology alone then is no guarantee of cleaner elections. But technology can be part of the solution.

Update: An election monitoring SMS template is a blog post that expands on the points here by linking to a SMS template I designed for election monitoring.

The pros and cons of crowdsourcing election monitoring

Sharek

MobileActive.org’s Katrin Verclas has a great article looking at the pros and cons of crowdsourcing election monitoring, based on the experience of Lebanon recently. 

I agree that crowdsourcing anything leaves much to be desired in terms of accuracy and information fit to feed into critical decision support processes. This is why the ICT4Peace Foundation is working on a crisis information management demonstrator, built on top of Ushahidi, that has information qualification routines built in. The tool will not be for the masses, but for agencies with trusted networks of field personnel who will feed in information, with the system itself open to social media input that can be vetted by agency trusted personnel. This opens up the system to be wholly crowdsourced, à la the Lebanese model, or completely closed to those outside the trust network(s) of an agency / agencies working on a particular issue, in a certain region or towards a shared goal. The design also allows the system to be anything in between these two extremes, so that the key responders to a crisis can determine the best degree of openness.  The important point that even if different international and local agencies had different approaches to what degree the system should be made public (i.e. extend to untrusted, initially unverified crowdsourced information) the common underlying information management architecture and standards would make for far greater and easier interoperability and information harmonisation. 

I’m interested in how Ushahidi’s evolving Swift River concept tackles this problem, that Paul Currion has succintly and accurately expressed here.  

Until such time there’s a better solution, I’ll still be training election monitors at the Centre for Monitoring Election Violence (CMEV) how to enter data into Google Maps that are verified for accuracy in a timely manner. Right now, there’s no escaping the labour required for the task – each location and incident is entered into the map directly, no automated source from the web is used to populate maps. Helps us give as close to a real time image of the ground situation in the lead up to and on the day of election, more useful we are told by extensive feedback from local media, than a mashup that just puts unverified reports on a map along with other data streams.