How much information should we share in peacebuilding and humanitarian operations?

This is a response to a post made by Nigel Snoad here on the Strong Angel III website. Nigel highlights vital concerns in the design of humanitarian aid systems, but my interest is also from the perspective of ICT4Peace, which shares many of the same concerns.

How do we balance the need to be open and share information with operational partners (known and unknown) and the community, with the need to minimise risks to staff, equipment and the populations we serve? How do we define, and agree on, what constitutes a realistic and acceptable risk? Hiding too means we may miss opportunities to help, chances we may not ever realise we failed to grasp.

The problem of humanitarian aid within ethno-political conflict is that the aid effort is quickly framed within ethnicised perspectives. Each ethnic narrative seeks to build a case for their marginalisation in the aid effort as opposed to the bias towards the Other. In States with a clear division between ethnic majorities and minorities on a national level, the case can often be made for verifiable bias in the delivery of aid. The situation however is rendered more complex at the grassroots, where national majorities can be regional minorities and vice versa. In this light, agencies have to be extremely careful in the delivery of aid to affected communities – so as to ensure equity and, most importantly, communicate strategically to ensure that the perception of biased aid delivery is mitigated to the greatest extent possible.

This in turn requires the open dissemination of information on aid delivery – who gets what, where and when and why. As I noted in an article I wrote a month after the Boxing Day tsunami of 2004:

As civil society organizations have also pointed out , one needs to address the complex dynamics of sustainable development in a holistic manner. This may not lie in the creation of wholly new frameworks and institutions to deal with the tsunami relief efforts, but more critically, in strengthening existing institutions (and processes) to augment their capacity to address the needs of the social fabric affected by the tsunami. With accountable and transparent frameworks, aid should also go to legitimate, proven civil society organizations that have a demonstrable capacity to address the ripple effects of the tsunami on a number of levels – from grassroots to the levels of policy making. The government and LTTE must come together to forge a covenant that eschews bickering and instead builds frameworks for the sustainable development of regions affected by the disaster. The trust relationships created in this exercise would be invaluable in the larger processes of peacebuilding. Conflict sensitive approaches must be mainstreamed into every aspect of long term relief. Consonant with a renewed call for the introduction of Freedom of Information legislation, relief efforts must always be open to the rigour of public scrutiny. It is only by the creation of accountable and transparent structures that one can avoid further erosion of ethnic and communal harmony, and counter perceptions of favouritism or bias in aid delivery and relief work. All communities in Sri Lanka, especially the Muslim and Tamil communities in the North-East, must be equal partners in the long term relief efforts to ensure that partisan bias does not creep and undermine the sustainability of relief efforts. Long term relief needs to be looked at holistically – from a media that acts in the public interest to enabling legislation that strengthens the accountability of relief mechanisms the myriad of ways in which Boxing Day 2004 can change, for the better, the contours of the larger peace process remain uncharted to date.

The emphasis on accountability, transparency, trust, right to information legislation, equity and holistic, inclusive frameworks I believe under gird any appreciation of information security in humanitarian aid systems. As I note in a monograph written a few weeks after the tsunami that captured InfoShare’s information architectures for the humanitarian response, the first days & weeks of the relief efforts brought to light the following information needs:

  1. Information on the type of the disaster – what a tsunami was, how it formed, the dangers of further tsunamis during the severe after shocks that continued for many days etc
  2. Information on missing persons, including foreign nationals. This included details of those internally displaced by the tsunami
  3. Information on immediate needs of survivors (shelter, food & medicine)
  4. Information of resources available to deliver aid – from 4WD vehicles, to trucks and helicopters
  5. Information of organisation to give money and donations in kind to – collection centres, bank account details, wire transfer instructions
  6. Information on contact numbers for emergency services, relief agencies, regional offices of large NGOs, country representatives of INGOs and donor agencies, number for key agencies in the UN
  7. Dissemination of requests for help, channelling aid to appropriate locations, mapping resources and taking inventories of aid received
  8. GIS data on Sri Lanka post tsunami and pre tsunami, including accurate and up-to-date maps of affected regions and satellite imagery to pin point where aid was needed in communities which had been isolated after the tsunami.
  9. Coordination of local and international volunteers involved in the relief efforts – what their skills were, where they were needed, what they were doing once assigned to a particular area
  10. News reports on key developments in the affected regions, including the details of money pledged for relief efforts and how to access this money
  11. Database of various NGOs operational after the tsunami across the affected regions who could be mobilised for aid and relief operations
  12. Information on the actual ground situation in the worst affected areas – with dysfunctional mobile communications, the national telecom provider’s PSTN infrastructure badly affected, transport infrastructure washed away, there was an urgent need to ascertain the status of survivors

As the reader will recognise, some of this information is extremely politically sensitive – that which was captured in the relief effort could be used to target communities and ethnic groups in a renewed war effort, and given the Sri Lankan’s state’s pathological inability to engage in a serious peace process, we were faced with the acute problem of having on the one hand the need to collect, store, analyse and disseminate sensitive information and on the other hand the need to maintain control of who and where this information was used.

It is not a problem that one maintains complete control of. In fact, it is a process of an increasing loss of control – the longer the relief effort takes, and the more complex the relief operations, the greater the need for cross cutting information resulting in an exponential growth of ways through which such information finds its way into the public domain, or at the very least, into domains that are not necessarily part of the relief efforts. One also never knows exactly when this occurs, or to what degree.

Looking towards the future, the development of an open standards, secure, pervasive, PC and mobile based solution that allows humanitarian agencies to collaborate effectively for the limited goals and purposes of immediate relief may offer greater information security. Anything beyond the immediate relief efforts, and humanitarian agencies themselves compete for limited funding, necessarily impeding their ability and willingness to share information between themselves on operations on the ground. In this limited first phase, information gathered will be under a broad non-disclosure agreement that clearly stipulates how the information can be used, for what purposes and in what manner. Data export can only take place after consultations with the affected populations or their direct appointees (not even local politicians must be mandated to speak on their behalf). The affected populations have open access to monitor relief aid, but are also bound by restrictions in the manner they can use this information. The contest between building trust and maintaining information security versus the imperative of public scrutiny is not something that can be hard coded into a humanitarian aid system since it is so rooted to the specific socio-political and cultural context, but broad provisions can be made in the design phase to accommodate demands for transparency and legitimate concerns about information leakage to military actors.

As Nigel says:

In the context of civil-military deconfliction/coordination/cooperation information sharing becomes a particularly sensitive issue. Because of training, doctrine and culture security and military staff (and the systems within which they operate) instinctively protect information. Similarly humanitarian organisations withhold from the military instinctively. This can cause enormous waste and missed opportunities in humanitarian operations. It is usually justified as a necessary force protection (or community protection) method. Such with-holding may be passive, by failing to divulge that some information is available, or avoiding being accessible to civilians/military who are actively seeing it; or active, by denying that information exists when it does, or, that great catch-all, it’s “sensitive”, when in fact there are no capabilities being revealed.

Future technologies may also look into something akin to information DNA – invisible yet system wide meta-tags that clearly indicate when records were gathered, by whom, for what purpose. These tags can then be tracked, so as to ensure that information gathered for humanitarian relief is never used for active combat operations, however valuable such information may be for offensive / defensive operations.

We may also look at ways through which a common application eco-system, such as that which is made possible even today by Groove Virtual Office, can bring together civilian and military actors in shared spaces to exchange, in a controlled manner, information on relief efforts.

At the end of the day, the issue is deeply human. While the systems themselves must be secure and resilient to external threats, adult education on how information can be used for harm by actors disinterested in humanitarianism, basic security protocols, common sense and adaptive learning are all important to ensure that a common understanding on information security is constantly honed amongst those involved at all levels of relief work.

Nigel ends his post with a few questions:

c) what is the “resolution” of information that individual organisations are willing to make public?

d) is there a reasonable “threshold” of registration or identity verification that should be required to receive information?

Rather than see it as a resolution, I would submit that it is more useful to see information security as a process. A resolution assumes that there will come a time when all systems and those involved in using them would be on the same page when it comes to information security. I think not. The transformation of the current state of affairs in humanitarian response and information systems design for relief and aid work will require global policy leadership, commitment of the largest NGOs including the UN, acquiescence of the world’s largest militaries, notably the US, development in the technology itself, visionaries who are able to think of systems that address the failings that plague them at present, a concert of thinkers from peacebuilding, humanitarian aid, military and tech, and requisite funding to develop more satisficing systems.

As for a reasonable threshold, I think that’s an easy one to answer for humanitarian aid and ICTPeace.

Do no harm.

Any individual, organisation or process that seeks to do harm using information gathered, collated, disseminated or analysed during relief work or peacebuilding needs to be locked out of systems with sensitive information. Ironically, they may well use the same systems in closed networks of their own to gather this information, but this is not our concern – we cannot prescribe how technology is used, but we can show, by example, how technology in the hands of committed individuals and organisations can better address the significant challenges of humanitarian aid and peacebuilding that lie ahead of us.

3 thoughts on “How much information should we share in peacebuilding and humanitarian operations?

  1. Sanjana,

    thanks for the thoughtful comments. That was a late night sleep-deprived post. One of my continuous frustrations with international aid and assistance is the lack of transparency and accountability you mention. In a disaster response this applies first and foremost to the affected population, but also applies to governments, other aid organisations and donors. I’m continually astonished that aid groups place little emphasis on communicating with local populations and on engendering trust – it’s almost as if many individuals believe that the aid itself is enough. In Aceh immediately after the Tsunami there was an immense thirst of the local community to know what was going on and what all these foreign groups were doing. It’s also their right. Citizen media and media development groups like Internews and your own can play an invaluable role in asserting and fulfilling these rights for the most affected.

    Just a small thing, I don’t think I was as clear as I should have been in my comments that you critique above, c) where I mention “resolution” of information, I was actually referring to the level of detail available. In most cases more is better, though this can be overwhelming if not presented well. But in some cases, for instance in conflict areas, or where there’s the possibility of abuse, I know that we are under significant pressure not to release too much. Aggregate data is often the best policy, so our challenge then is to decide what level of detail, or resolution, is safe but still useful. For example: in Southern Sudan map-making was always compromised by the story of a map that showed the location of temporary schools, which were bombed not long after the map was published. It was simply easier to make a map with the detail, than to go to the minimal effort of aggregating it. The real challenge was in thinking through the consequences. Myth or not this anecdote held a strong sway on local and international information policy in Southern Sudan, and points to some of the ways that data can be abused if the longer term consequences are not considered and “do no harm” isn’t at the core of our approach. Which provides us with a challenge: in many cases “do no harm” is about starting by assuming that you’re going to be releasing less information, or less detail.

    It’s obvious, but sometimes we have no idea of how information will be used, or by whom. Authoritatively udging “do no harm” in this context is impossible. What you’re left with perhaps is a “best effort” and “best intent” to do no harm. This requires a due diligence when examining the consequences. It requires a consideration of not only short term consequences, but an examination of the long term implications – an extremely difficult task in most cases.

    So yet again, we arrive at a flaw in much humanitarian information management: there needs to be real thought about the stakeholders and users of information, and what their needs are. So much effort is wasted on situation reports destined only for filing cabinets. So little applied to communicating with recipients and the affected. Paul Currion’s ECB study did some great work looking at how agency sitreps might be reformed.

    As for our (always moving) threshold for authentication… yes “do no harm” is our starting criteria, but as for the details and principles about identity, access, and privacy, I think it’s still open, something that I’ll comment on more soon.

    Damn, this should have been posted on my blog. And it will be.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s