Skip to main content

Protecting People in Cities: the Disturbing Case of Haiti


Note: Edited version to be published by Disasters (Overseas Development Institute and Blackwell Publishing)


Disaster struck Haiti on 12 January 2010 in the form of a 7.0 earthquake which left some 223,000 people dead, 300,000 injured, and 2 million homeless. This 60-second earthquake, occurring in Léogâne, near the capital city, Port-au-Prince, had a particularly devastating impact on the Haitian government, with nearly 30 percent of its civil servants killed, all but one government ministry building destroyed, and basic infrastructure wiped out. The United Nations (UN) experienced its greatest loss of life on a single day ever, when 102 staff members died (Active Learning Network for Accountability and Performance (ALNAP) 2011). By any standards, it was a mega-disaster.

The international community mobilized rapidly and massively to assist Haiti. UN and other international agencies deployed staff quickly, thousands of NGOs rushed to the scene, donor governments and military forces sent personnel and some $3 billion was pledged in relief and recovery efforts. Indeed almost two-thirds of all international funds mobilized for natural disaster response in 2010 went to Haiti (Ferris & Petz 2011). Clusters, the international mechanism for coordinating humanitarian response, were set up, international staff arrived by the hundreds in Port-au-Prince, programs were established and aid poured in. 

It was a monumental effort and there have been many efforts to evaluate the international humanitarian response to the Haitian earthquake. Indeed, as of February 2011 ALNAP counted 45 evaluations of response to the earthquake (ALNAP 2011). In summarizing the results of these evaluations, ALNAP identified several commonly-identified shortcomings:

While the evaluations will undoubtedly continue and lessons will hopefully be learned to guide future humanitarian response, this study looks at one particular issue—protection—with a specific focus on the impact of Haiti’s unprecedented urban emergency on existing tools and definitions of protection. This study is based on field research carried out in Haiti, particularly the involvement of one of the authors as staff of the International Organization for Migration with particular responsibility for Camp Management and Coordination.

It is important to underline that the failure of protection in Haiti was largely due to the inability of the Haitian government to redress a chaotic and difficult reality. It is unlikely that international actors, even if the system had worked perfectly, would have been able to fully protect Haitians in this environment. The international actors managed to provide basic services and relief items and were at least moderately successful in staving off the worst effects of a cholera epidemic and hurricanes, but they were unable to provide physical safety to Haitians—arguably the most basic of human rights.

 This paper argues that there were two principal reasons for this failure:

  • The definition of protection as used in Haiti was simply too broad. International actors spent far too much time trying to define what protection was and the definition that emerged—full respect for all human rights—was not helpful in setting priorities in an urban setting characterized by immense need. Consequently, the clusters working on protection issues defined their tasks and priorities differently and ultimately were not effective in protecting the majority of Haitians.
  • The standard international protection tools which have developed over the past five to ten years simply did not work in Haiti.

While there were many Haitians in need of protection following the earthquake, this study only examines protection in camps and settlements. Protection of other groups, such as orphans and detainees, was reportedly carried out more successfully than for those living in settlements and camps, but that issue lies beyond the scope of this study.

Download the full article »

Get daily updates from Brookings