Berto Jongman: Humanitarian Aid & Forgotten Conflicts

04 Inter-State Conflict, 05 Civil War, 06 Genocide, 07 Other Atrocities, 08 Wild Cards, Budgets & Funding, Civil Society, Cultural Intelligence, IO Deeds of Peace, Non-Governmental, Peace Intelligence, Policies
Berto Jongman

Some important connections drawn between aid, corruption, and positive change; and also important omissions — conflicts out of the news where paying attention could make a difference.

Singling Out Forgotten Conflicts

The ISN Blog, 15 March 2012

A popular method for identifying which conflicts necessitate more attention from the international community is to estimate the difference between supply and demand of humanitarian assistance in these conflicts. Supply and demand, however, are very hard to measure in emergencies. This has led to the development of several indicators used to measure ‘forgotten conflicts’.

These indicators are often applied on an annual basis and are intended to generate media attention (to increase donations) and/or support donor operations (to comply with impartiality). Have these efforts been successful? Have they effectively singled out and buttressed forgotten conflicts? Looking back on the past decade, in this blog post I’ll assess which conflicts received the least (and most) attention from international actors.

I would argue that, ideally, humanitarian assistance filed under the protection cluster (e.g. regarding child protection or gender-based violence programs) is disbursed in proportion with the severity of the conflict. If this is the case, we can then establish that protection-based humanitarian assistance was allocated on the basis of needs alone in the past decade. In this post, I’ll also identify which conflicts have drifted from the expected levels of humanitarian assistance and were either ‘over-’ or ‘under-funded’.

How to go about it?

Severity of the conflict is understood as the number of civilian deaths from 1998 to 2010 caused by one-sided violence of governmental or non-state actors (see my methodology and variables; and the original data for more information). These figures implicitly show the extent to which a government was unable (or unwilling) to protect its population — and could therefore provide a rough estimate for the demand of humanitarian protection-based assistance.

I found that the number of civilian deaths significantly and positively correlates with the amount of humanitarian assistance: More civilian casualties mean more financial aid channeled by international donors to that particular conflict. On the other hand, civilian deaths explain only 43% of the variation in the amount of humanitarian assistance; hence, it is premature to infer from this model alone. Let’s see what else could explain the varying amount of humanitarian assistance.

Firstly, the number of internally displaced persons (IDPs) within a country needs to be taken into account – the protection cluster’s creation was actually prompted by the lack of institutional arrangements for IDPs.

Secondly, in light of the attention give by the protection cluster to human rights, it would be reasonable to hypothesize that poor compliance with human rights would set off an increase in humanitarian assistance.

Lastly, levels of corruption in a country of concern might either deter donors to provide aid (as they see it as ‘lost money’) or would conversely suggest higher levels of humanitarian assistance, given the inefficient use of funds. (Please note that other possible factors, such as media coverage, are excluded now for the sake of parsimony and due to the difficulties in measurement.)


I found that when combined, the number of civilian deaths, the number of IDPs and the level of corruption produced a model which accounts for 76% of the variation in the amount of humanitarian assistance – stronger than when just based on civilian deaths alone. (Human rights indicators are excluded for they decrease the robustness of the model without adding to its predictive power).

So, what can we say about the disbursement of protection-based humanitarian assistance over the last decade?

Working from baseline values, my model suggests that in the period 1998 to 2010, a country with

  • one more IDP prompted an additional $55 of protection-based humanitarian assistance;
  • one more civilian death prompted an additional $5440 of protection-based humanitarian assistance; and
  • a one-place drop in the country’s (relative) ranking regarding the perception of corruption prompted an 18% increase in the amount of protection-based humanitarian assistance

Of course these computed numbers do not perfectly describe any real conflicts; nevertheless, they provide an understanding of how protection-based humanitarian assistance was disbursed in the past. Accordingly, we can now identify the biggest outliers.

By acknowledging that the model outlined above represents the consensual willingness of donors to provide humanitarian assistance, I found that:

  • Sudan, Afghanistan, Somalia, Sri Lanka, and Haiti are the five most ‘over-funded’ conflicts of the last decade; and
  • Colombia, Angola, India, the DRC, and Turkey are the five most ‘under-funded’, “forgotten conflicts” of the last decade.

If we accept the vital role that protection-based humanitarian assistance serves in strengthening civilians’ resilience against conflict, then the fate of “forgotten conflicts” needs to be emphatically addressed by donors.