Darfur Crisis:

Death Estimates Demonstrates Severity of Crisis, but Their Accuracy and Credibility Could Be Enhanced

GAO-07-24: Published: Nov 9, 2006. Publicly Released: Dec 11, 2006.

Additional Materials:

Contact:

David B. Gootnick
(202) 512-3149
contact@gao.gov

 

Office of Public Affairs
(202) 512-4800
youngc1@gao.gov

In 2003, violent conflict in Darfur, Sudan, broke out between rebel groups and government troops and government-supported Arab militias. While few would dispute that many thousands of Darfur civilians have died, less consensus exists about the total number of deaths attributable to the crisis. Estimates by the Department of State (State) and other parties report death tolls up to about 400,000 for varying populations and periods of time between February 2003 and August 2005. Based on the views of experts convened by GAO and the National Academy of Sciences, interviews with estimate authors, and a review of relevant literature, this report (1) evaluates six Darfur death estimates, (2) identifies general challenges to estimating deaths in such crises, and (3) discusses measures to improve death estimates.

The experts GAO consulted did not consistently rate any Darfur death estimate as having a high level of accuracy; moreover, they noted that all the studies had methodological strengths and shortcomings. Most of the experts had the highest overall confidence in estimates by the Centre for Research on the Epidemiology of Disasters (CRED) and had a slightly lower level of confidence in State's estimate. Many experts believed State's lower-end estimate was too low. Additionally, the published documents describing State's estimate lacked sufficient information about its data and methods to allow it to be replicated and verified by external parties. Estimating deaths in a humanitarian crisis such as that in Darfur involves numerous challenges. For example, in Darfur, difficulties in collecting mortality data, such as lack of access to particular geographical regions, impacted the data's quality and led to data gaps. Because of such data gaps, some Darfur death estimates relied on potentially risky assumptions and limited contextual information. Further, limitations in estimates of Darfur's population before and during the crisis may have led to over- or underestimates of the death toll. Finally, varying use of baseline mortality rates--the rate of deaths that would have occurred without the crisis--may have led to overly high or low death estimates. The experts proposed and rated a wide range of measures that U.S. agencies could take to improve the quality and reliability of death estimates for Darfur and future humanitarian crises. Among these measures, the most highly rated was ensuring that public documentation of the data and methods used contain sufficient information to enable external replication and verification of the estimates. Other very highly rated measures include collecting and maintaining data for specific periods of time and geographic areas and housing the responsibility for making estimates in a reputable independent body.

Recommendations for Executive Action

  1. Status: Closed - Implemented

    Comments: Since our report, efforts to support the improvements in data and methodologies have been delegated to USAID. The State Department has not itself published death estimates since we reported.

    Recommendation: To safeguard the Department of State's credibility as a source of accurate and reliable death estimates, the Secretary of State should promote greater transparency in any of its future death estimates for Darfur or other humanitarian crises by ensuring that publicly available documents contain sufficient detail on the estimates' data, methods, assumptions, and limitations to allow external researchers to replicate and verify the estimates.

    Agency Affected: Department of State

  2. Status: Closed - Implemented

    Comments: USAID continues to provide support to non-profit organizations and institutions that improve field studies that allow for reporting on the human impact of disasters. USAID has supported the collection and publication of the results of field studies in a database as well as improvements in data methodologies.

    Recommendation: To enhance the U.S. government's capacity to assess the dimensions of, and respond appropriately to, any future humanitarian crises, the Secretary of State and the Director of U.S. Foreign Assistance and USAID Administrator should consider the experts' other suggested measures to help address gaps in data and improve the quality and reliability of any future death estimates.

    Agency Affected: United States Agency for International Development

  3. Status: Closed - Implemented

    Comments: Since our report, efforts to support the improvements in data and methodologies have been delegated to USAID. The State Department has not itself published death estimates since we reported.

    Recommendation: To enhance the U.S. government's capacity to assess the dimensions of, and respond appropriately to, any future humanitarian crises, the Secretary of State and the Director of U.S. Foreign Assistance and USAID Administrator should consider the experts' other suggested measures to help address gaps in data and improve the quality and reliability of any future death estimates.

    Agency Affected: Department of State

 

Explore the full database of GAO's Open Recommendations »

Oct 21, 2014

Sep 24, 2014

Sep 10, 2014

Sep 9, 2014

Aug 28, 2014

Jul 24, 2014

Jul 21, 2014

Looking for more? Browse all our products here