⇓ More from ICTworks

Citizen election reporting in Kenya was a breakthrough in online-offline collaboration

By Guest Writer on May 13, 2013

Map Kiberia

The Kenyan elections were more than a month ago, but a debate continues in the crisis mapping community about whether the various technologies deployed to track and respond to outbreaks of violence were a confused and possibly dangerous mess, or a successful contribution to what was ultimately a peaceful (if disputed) process.

DO WE REALLY NEED ALL OF THOSE PROJECTS??? Do we really need 3 maps, 7 phone numbers, and several web-forms? Is that really such a crazy bad idea to have one coordinated number/web-form that could then have in the back-end multiple responders and organizations working together? – Anahi Ayala  (ICT Works blog post)

The criticism goes on to describe this duplication as irresponsible and dangerous, especially supposing that the submitted information has no real response mechanism.

While it’s true that having multiple public numbers for submitting information about one election is not ideal, I believe that behind the scenes was a much more encouraging process that has only just begun. Here’s why I think that in the balance, technology was part of the solution, not the problem, during Kenya’s elections:

1) Unprecedented collaboration among technologists, at least at a pilot level.

Map Kibera took part in elections monitoring by mapping and reporting through SMS, blogs and video throughout the election period on our multimedia sites, Voice of Kibera and Voice of Mathare in two slums. Kepha Ngito, our executive director, offers this extensive writeup of how the process looked behind the scenes, definitely worth a read. Map Kibera already had been working with Ushahidi-based websites and video news for three years in Kibera, and with blogging and video in Mathare for about two years; therefore neither project was created specifically for the election. Given our status as an established community-based group focusing on reporting and local information, we were ready to take on this event without creating any new technology.

As elections neared, more organizations began to set up temporary programming around election reporting. In particular, our team joined events held by Ushahidi around their Uchaguzi platform, and we began to think about how to collaborate: they as a national scope project and ourselves in-depth in two key communities. Ultimately, some members of our team worked throughout the election at the iHub headquarters of Ushahidi to monitor their reports coming in from Kibera and Mathare, and share our more detailed and verified reports with their system. This meant both reports and response could be tightly coordinated.

Would it have made sense to have used only one reporting number, that of Uchaguzi for instance? No, because Uchaguzi is no longer active, while Map Kibera is building up a long term citizen engagement including this SMS number. It made more sense to work on interoperability and coordination.

2) Unprecedented coordination among community-based groups

What was the key to all the information coming in, and the verification process?

At the community level, there was unprecedented coordination among a variety of agencies who normally do not work together. Concerned groups put aside whatever challenges normally keep them operating separately, and pledged support to each other in order to serve the good of the community. This kind of network is a real breakthrough in coordination locally given the challenges that often prevent such teamwork, and it naturally came from desire for security in the slum not any outside impetus. As Kepha writes:

KCODA, Pamoja FM, Map Kibera, Kamukunji Pressure Group, CREAW, the Langata District Peace Committee, Community Policing groups and the office of the District Commissioner joined efforts to create a network called the Kibera Civic Watch Consortium, a body that would respond to and coordinate the community’s efforts to maintain peace and provide interventions where possible. (Kepha Ngito, How slum communities came together to help prevent election violence)

3) Response and verification

It was this background offline coordination that would allow for immediate election-day verification and constant liaising with security groups, both official and community led, in case of any problem. The online-offline coordination often involved both SMS into the system, and phone calls to security or other key people to keep tensions down.

…We stationed our trained citizen reporters in each polling station to be relaying SMS news to our verification team to be verified and approved before being posted on our Voice of Kibera and Voice of Mathare websites.The Map Kibera verification team dealt with every information that came in, calling back and forth to establish the facts and figures about every report sent in. Our video teams rushed to scenes, most of which were not known or easily accessed by the mainstream or foreign press to capture instant news which they edited and uploaded on Youtube.

Members also took photos and posted them to our blogs and Facebook group. In this verification process, the team succeeded in dismissing several false alarms, wrong information and propaganda for violence. In addition and in response, security organs and emergency service providers enhanced their presence in these areas highly reducing chances of violence. In one instance, when many reports were sent about youth gathering around in groups in one area of Kibera, after several phone calls with the security organs, the Police Commissioner authorized a chopper to fly around conducting a security check, the crowds soon dispersed and calm returned.

(Kepha Ngito, How slum communities came together to help prevent election violence)

I heard a number of claims floating around about technology platforms being directly responsible for police or security response. I would advise that these be investigated closely. In the case of both Kibera and Mathare, if anything it was this type of “online” or SMS-based reporting in conjunction with offline personal and official networks, connected often through ordinary phone calls, which activated response, not a pure technical system.

Bringing together the various kinds of technical reporting options with great local networks can create response processes that are effective, but not overly sensitive to false alarm reports.

4) Good citizen election monitoring is not “pure” crowdsourcing. We and others relied on establishing networks and offline meetings and coordination to build participation.

Another misconception is that every SMS number was targeted for the general public. Actually, the Map Kibera SMS line was and is primarily used by our volunteer reporters to send information from prearranged locations like polling stations. For the election, aside from our members, the Kibera Civic Watch Consortium sent in many reports during the election. The numbers are publicized to general residents, but operate mostly through carefully cultivated users.

I think there is a basic misconception that “crowdsourcing” something like election violence will happen with anonymous individuals. In tight-knit communities, this is simply not the case. They may text into a nationally publicized number, but those reports are not always the ones you want or can rely on. Verification is needed, which means local networks are still key. My hunch is that when people report through a local system the information is more likely to be accurate, because they know the faces behind the tech.

Particularly in more marginalized, insecure, or informal communities, people come together based on relationship networks, and being known and trusted as a leader in the community is an earned privilege that does not come easily nor is it taken lightly. People do not often trust something new that is introduced from “outside”.

National scale projects targeted to single events like elections should heavily support existing community level initiatives, and community-level initiatives around information need to be long-term investments into the community fabric. This means not just new technology projects which are still rare, but also traditional community media or data-driven local planning groups.

5) We’re still working out what works best in the space, so multiple projects are important for narrowing it down. A top-down model won’t work for citizen-based technology in emergencies, at least not for a long while.

While coordination and duplication avoidance is good, we are talking about places where the normal emergency response functions need supplementation and should be supplemented. I don’t know if a single top-down system for emergencies is ever going to work in Kenya, but it certainly hasn’t yet. I’ve seen way too many everyday crises happen with absolutely no response at all save for neighbors helping neighbors (and literally saving each others’ lives on a daily basis). In that sense people need access to options and a variety of ways to draw attention to and publicize an urgent situation. It’s also in the spirit of the technology world to keep trying out new ideas and iterating.

On the other hand, it’s true that some might irresponsibly publicize reporting channels that seem to promise response they can’t deliver, and technology is most certainly something we are now seeing organizations use to make themselves look good even at the expense of the public. We should ask whether there were unnecessary institutional barriers or unethical motivations to any lack of coordination or collaborative spirit, and direct our transparency lens that way. If competition for funding or funder requirements inhibited the social benefit of working together, as it usually does, then these incentive systems should be exposed. Also, the opportunism of pop-up and parachute technology projects just a week or so before the election is a distraction, and there were several of these as well. But what I think we’ve seen here is a partial triumph of civic collectivism over the usual silos created by the donor marketplace, which is why I’m seeing the glass as half-full. It could stand to be filled up all the way.

But here’s the most important trend that gives me hope:

7) A sea-change is underway in terms of how people engage with information in Kenya: they feel it’s their right and responsibility to speak out and to protect peace by countering rumor; and they increasingly feel they have tools with which to do so.

This is my hunch, but since 2009, I believe that the positive side of social media has had an impact in Kenya – or at least in Nairobi. I noticed during the election that people expected to be able to counter incorrect news and information sources by using their Facebook account or Twitter or one of the projects referenced here – they have grown used to reporting themselves and no longer rely exclusively on traditional media sources. When they see something happening, they expect someone local – if not themselves, then a neighbor – to be able to take a photo, send a message, somehow communicate.

This means that rumors can be countered more quickly, and leaves room for peace activists (most Kenyans) to organize and amplify accurate and helpful messages and at least contribute to the conversation, creating a more balanced discussion.

During these elections there was a new sense of the importance and responsibility that citizens have for being information collectors, transmitters, publicizers, verifiers, and the inkling that the standard channels of information aren’t the only ones that exist anymore – and that citizens have the responsibility to not only voice opinions but keep the rumors down, to participate in peace. Real crisis prevention has much more to do with local leadership, coordination, and behind the scenes response than the information that’s necessarily visible online. But that isn’t to say that creating visibility, keeping track of the truth and bringing information out of the dark in close to real time isn’t extremely valuable – it connects and inspires those who want to keep the peace and provides the opportunity for a local counter narrative to the dominant media.

Don’t risk missing the bigger story here: the simple act of residents recording actual ground level events themselves will have a long-term transformative impact on society – nowhere perhaps, as profoundly as in places like informal settlements.

Erica Hagen originally posted this as “Citizen election reporting in Kenya: A failure of technology duplication, or a breakthrough in online-offline collaboration?” and its republished here with her permission.

Filed Under: Solutions
More About: , , , , , , , , , ,

Written by
This Guest Post is an ICTworks community knowledge-sharing effort. We actively solicit original content and search for and re-publish quality ICT-related posts we find online. Please suggest a post (even your own) to add to our collective insight.
Stay Current with ICTworksGet Regular Updates via Email

One Comment to “Citizen election reporting in Kenya was a breakthrough in online-offline collaboration”

  1. Anahi says:

    Dear Ericha,

    Thanks for the insightful post and for the sincere answer to my blog post. I have two main comments to make, strictly related: the first one is that you seem to have completely misunderstood my post, and the second one is that, because of this reason, you are actually not answering to my question.

    But first, let me clarify a couple of things:

    I never said that Technology was part of the problem during the Kenyan elections, in fact, if all, it was kind of irrelevant for what we know. If it was part of the solution as you stated, still need to be proven with real data. The anecdotal evidence you are providing here is great, but not enough. As people that work in technology we (all of us) need to stop evaluating ourselves, we need to start doing responsible and transparent independent evaluations that collect and disseminate real data on the impact of what we do.
    I do know as a fact that organizations where coordinating in between themselves, in fact I never doubted that. I was there during the Constitutional Referendum in 2010 when already some of those organizations where coordinating and working together. I do not doubt that organizations where coordinating in between themselves, what I am critiquing is the fact that this coordination never brought to a coordinated interface and also that there was a very poor communication with communities regarding this coordination. (I will elaborate on this later on)
    On the verification processes: I am super happy to learn what was the system used and looks like it was not just very effected but really community driven. I also never doubted that.
    “Good citizens election monitoring” is not “pure crowdsourcing”: I don’t disagree, but I personally think that both systems can work. One will give you less and more reliable data, the other one more data and less reliability. It all depends on the resources that you have and the system you build. I would not go as far as to decide which one is the “best”. I have seen both systems working in different occasions.
    I never advocate for “top-down” system. I talked about coordination in the way technology projects were interfacing with the local population. This does not have to be top down it can be easily bottom-up.
    Your points about community ownership, offline-online systems and the increased coordination among actors in the past 5 years in Kenya are absolutely great points, which I agree totally.
    Your point on neighbors helping neighbors is actually super spot in. I guess then they could help each other by calling each other and not necessarily using an online map? Especially in a place like Kibera where not a lot of people have a smart phone?
    ‘whether the various technologies deployed to track and respond to outbreaks of violence were a confused and possibly dangerous mess, or a successful contribution to what was ultimately a peaceful (if disputed) process.” While I do think this is valid debate, this is not what I argued. I have no proof to say that what happen was on or the other, and I never talked about a dangerous mess. I talked about the duplication and multiplication of interfaces with the communities that those projects were supposed to serve, or to be created by. I think that we (and I count myself with you and other in here) should stop being touchy and oversensitive over criticisms and start talking more about what is clearly not working and what is clearly less then an ideal situation and start working on possible solutions, not give each other pat-pat on the shoulders on how cool we are. The truth is that, since nothing happens we will never know if this “groundbreaking” system has been at all part of the solution or a possible disaster. What we do know is that still an enormous duplication of services was provided.

    Now lets come to my criticism to your post:

    You are arguing very eloquently about the coordination in between your organization and other and in between your organization and the community you have been working with, but the truth is that my post is not about the lack of any of the two. My point is that in this “groundbreaking” system still the interface with the community was multiple and duplicative. Again the question is, I understand why you needed to have your own number – to differentiate your on going project from Uchaguzi, which is a temporary one – but my question is, did the communities involved needed two numbers? Why? Also, if they did, why would they choose one of the other? Based on who they know? Or who they trust? If they trust you, and you trust Uchaguzi, and your work is community lead, shouldn’t the community also trust Uchaguzi?
    You make a great example that explain my point actually: once several reports where sent about youth gathering “after several phone calls with the security organs, the Police Commissioner authorized a chopper to fly around conducting a security check, the crowds soon dispersed and calm returned”. Now, if the community would have worked directly with the authorities, and if the report would have been sent to them directly through their channel, would this have been much faster? Why there is a need of one primary number (MapKibera) where one person would answer and then call another number (the police) and report the same information? In the case o something much more serious than this, wouldn’t the time lost in having one intermediary be actually potentially extremely dangerous?
    This is the most terrifying statement in your post: “It’s also in the spirit of the technology world to keep trying out new ideas and iterating.” Ericha, yes, we should always test out new ideas and iterate. But allow me to say, this is not what you do in potential emergencies. You never do that. Local communities during emergencies are not the technologists rats lab where we can test and play around we what we like. You don’t either test coordination in possible emergencies. This is not only dangerous, it is the assumption that “we” can test things on the shoulder of people that can actually die if we fail testing is about potential failure, and if we can avoid it, we should never do it.

    Now let me conclude this long answer. I agree with most of the point you make, but in your long response, while addressing very good points, you do not answer to my question: despite the fact that organization deploying those tools were coordinating in the back end, and verifying their information, and working together, what was the actual need for multiple interfaces for the local population? What was the advantage for the local population in having multiple interfaces?

    In addition to this, you say something extremely important: “Real crisis prevention has much more to do with local leadership”. Why couldn’t some of those efforts be directed to create more local ownership instead of creating multiple channels? Why the MapKibera could not work with the community in Kibera to support the responders on the ground – like the police you guys then ended up talking at the phone with? Could we use technology to create a more cohesive environment where we enable more local trust and ownership of those systems?

    Now let me clear on one thing that I think some of the organizations I mentioned in my post misunderstood as you did: I know almost all of you. I have lived in Kenya for 3 years and worked there since 2005. I have also worked with some of you. Believe me when I say that I know the challenges and the incredible achievements that a lot of these organizations have faced and achieved in working on these election. I am in no way doubting any of that.

    On the other side I reiterate my point: we really need to be careful when setting up those type of system in not creating a multiplication of interfaces that “can” result in complications and difficulties for the people using the system in understanding the situation or even getting the help they need. We need to be very clear in managing expectations and in talking about what we can do, and what we CANNOT do. We need to take into consideration the fact that ultimately a stronger and more transparent government is the product of a stronger and more aware population, and accountability have to result also in better relationships between the two, possibly without intermediaries. We also need to know and keep in mind that sometimes what it is good for us, it is not good for the local communities, and the comparative advantage may not be worth it.

    And lastly, we need to stop claiming “groundbreaking” events, game changing projects, life saving results without backing them up with real data and evaluations. Anecdotal evidence is good, but it is not enough anymore. As people with technology and trying to provide solution to real problems, we need to make sure that what we are doing is really having an impact. To do so we need to really start doing serious M&E and invest money into that. It is about our credibility as crisis mappers but also about really learning from our mistakes and understanding how to get better. The “testing” and “trying” and counting heads is not enough anymore. We need to get serious and start answering to the real questions: are we being transparent and accountable and open?