Mapping disaster risk reduction

posted in: Uncategorized | 7

My thesis focuses on how international and national actors work with each other to reduce disaster risk— the relationships, governance structures, and power dynamics shaping risk reduction. Concern is my case study and I’m looking at its work in three countries: Kenya, Sierra Leone, and Bangladesh. I’m currently in the process of finalising my dissertation and I wanted to show something that I’m excited about: a few of the maps that I’ve produced for this project.

DRR is multi-sectoral and cross-scalular and multi-hazard, in other words, a rat’s nest of actors, interventions, and hazards. To top it off, actors each have their own interpretation of what DRR actually is, meaning that one person’s DRR is another person’s development or health.

This leads to the practical problem: how do you understand DRR? While some people  define DRR a-priori, I’ve turned to regime theory to make sense of this. Regime theory is about figuring out how ‘regimes’ (the confluence of actors around a given issue area) operate and the effect they have on actor behaviour. It’s similar to new institutionalism and polycentric governance, but with a special focus on delimiting the issue area and the actors within it. All this makes it a pretty good fit for my research, and there’s been quite a bit of work on it already so it provides a valuable platform for guiding the study.

I’ve approached understanding DRR interactions as a mapping problem.  I’ve collected a bunch of qualitative data (interviews, reports, policies) and quantified it using NVivo, a pretty awesome programme for qualitative analysis. I then plugged it in to Gephi, a tool for social network analysis (SNA), to map the regime. Since SNA is about actors and relationships I’ve had to code those actors and relationships. For example, the sentence local communities work to reduce flooding would be coded as local communities –> flooding; the sentence Concern works with the national government would be coded Concern –> national government. I’ve also categorised each actor by type (public, private, civil society) and level (local, national, international). While this is a meticulous way to analyse data and has involved thousands of codes, I’m pretty happy with the results.

Here are a few of the maps I’ve produced:

Sierra Leone





(here’s the PDFs: Sierra Leone, Kenya, Bangladesh)

The colours in these maps are linked to type: blue is civil society, orange is government, red is private sector, and white is other. The size of the node (circle) and edges (arrow) is dependent on the number of times coded (the more codes, the larger the size) and is an indication of importance. It is important, however, to note that size is relative to each regime and that for size these maps cannot be cross-compared.

From the maps, it is clear that for all three cases the regime, by number of actors, is dominated by public sector with some civil society and very little private sector. That said, in all cases core actors (actors occupying central positions and with a lot of big arrows) are a mix of public and civil society actors. The maps also show the relative scale of the regime, with Sierra Leone (the first one) having the fewest actors, Kenya (middle) a moderate number of actors, and Bangladesh (last) a massive number of actors. This stuff is useful for me because it provides a basic picture of the regime and the interactions, showing that the actors vary in size but are similar in type.

I’ve also simplified down actors to their essence:

Sierra Lone



(PDFs: Sierra Leone, Kenya, Bangladesh)

From these maps it becomes clear that while DRR is often described as multi-sectoral, the private sector is relatively absent from DRR activities. Interestingly as well, it can be seen that types of relationships and interactions are similar. I’ve also developed maps of how these actor types engage with various hazards, showing the hazards and interventions under the regime.

These maps provide me an overview of what the regime is. After this I need to understand how the regime works— the arrows between the actors. I’m doing this through qualitative data analysis to produce a narrative of actors and their interactions. The narrative is really interesting, particularly the use of mapping, training, and documentation to measure disaster risk and make it actionable (regime construction toward a Weberian bureaucracy).

Both these techniques are letting me figure out how DRR works, which is great, because, even with my years studying DRR, I am still deciding what I think DRR actually is. Results also have policy implications for NGOs and governments working to improve DRR, as they show gaps in the current DRR system as well as how that system operates.

7 Responses

  1. Mary Ndanyi

    This is quite informative. Can this be applied to livestock disease outbreaks as a disaster?

    • Hi Mary. This approach would be broadly transferable to a number of policy realms within and outside disaster management. It’s particularly useful in areas where there are a number of actors/types of interventions. For disease outbreaks you could map the actors and interactions, as I have done, but you could also map how those actors or actor types engage in different areas of disease outbreak management (vaccination, destocking/restocking, monitoring, etc). The maps are interesting on their own, but to really understand them you’d need a strong qualitative component focused on how interactions work and what they mean. If you’re interested in applying this approach to livestock let me know and I might be able to help you out.

      • Mary Ndanyi

        Hi Aaron.
        Thanks for the reply. I am interested in applying this approach in our project. We are working on development of disease control strategies for pastoralists in Kenya. I will appreciate your assistance on how to go about this. Looking forward to hearing from you.

  2. Francis Matheka

    Hi Aaron. Interesting map. Do you have a summary of what the linkages mean. It would be interesting for me at UNDP Disaster Risk Reduction Team and other partners in Kenya. Other than that I checked the Kenya map and it seems to be missing some key linkages. For example there’s is no inter-linkages between National Disaster Operations Center, National Disaster Management Unit, National DRR Platform, ISDR, Counties and UNDP. These are the key public actors in DRR in Kenya. Any explanations or insights on that?

    • Hi Francis. Thanks both for your comments about actor involvement; I agree that specific links between national/UN bodies might not be represented. This boils down to two issues, one related to methodology, one related to research objectives. Methodologically, I’ve taken a strict approach to coding because I’ve been worried about over-interpreting data. If a sentence reads, for example, ‘the government has received assistance form the UNDP’ I code it UNDP–>national public. I do not try and extrapolate to identify which agency. I’ve also only shown connections with 2+ mentions, otherwise the maps gets too cluttered with edges. The other is research objectives: I’m trying to show how the actors in my study (Concern and nat. regime) view each other and interact. This, for example, means that while certain agencies might play important roles with each other, they aren’t actually talked about that much. If you’re interested in the more formal dynamics of interaction you’d really want to do something like an organogram (which you could then compare against these discussions). If you’re interested in regime as a whole from a multi-actor perspective, you’d be much better off developing a survey targeted at all the main actors, or utilising an existing database with stakeholder interactions (e.g. )

  3. Alexander Mirescu

    Aaron, this is a very impressive piece of research that you have collected and processed. I commend you on your efforts, since DRR is such a rapidly developing field. As a DRR specialist with UNISDR and a professor, I find that both academia and global public actors are having a hard time keeping up with the newest developments. One of the newest aspect that the Sendai Framework captures is the agency of sub-national actors – university’s, community groups. I realize that civil society is in your research but it is so often a catch-all phrase for all things non-governmental. It is not a criticism, but with the rise of importance of sub-national actors and its emphasis in the post-2015 framework, it might be worth giving it a nod in your dissertation. Otherwise, as a political science professor, I would also love to have a read of your dissertation. Good luck with the defense!

    • Thanks for taking an interest Alex! Civil society is a huge term; beyond differentiating between civil society actors, I found it difficult to classify actors by sectors/levels. Community DMCs were often comprised of people from the government, the local population, and customary governance. International NGOs implemented government policy in absence of the state. And international best practice on DRR was often embedded within national government practices. A good narrative is a means of sorting through some of this, and is something I am definitely including in my dissertation.

Leave a Reply