This is a sample question from Jacquie in Nairobi.

Avatar2I’m preparing a story on “Africa’s World War”, the genocide in Eastern Congo. I’ve read that the ‘official’ death toll estimate (over 5 million) since the early 1990s is disputed. Where does this figure come from, how was it calculated and how can I come to a better estimate?

This is a sample answer from Mina in Johannesburg.

Mina100x100I found the source of the claim, source of the dispute, and a source that offers another estimate.
Here are the steps I followed.


1) Identify the source of the claim

Using the correct search terms, and knowing what to include or exclude, is necessary to get to the most useful results!

I use these terms: “genocide in Eastern Congo” OR “Africa’s World War” death toll estimate disputed -wikipedia -kibeho

Some notes about these search terms:

  1. Wrapping words in double quotes means that the words must appear in the same sequence as within the quotes.
  2. The OR is a special kind of phrase that tells Google, hey! Give me either what comes before or (pardon the pun) after the OR.
  3. Putting a dash or hyphen before a word tells Google to exclude results that have this word in them. I use these to get rid of results from Wikipedia and message boards.
  4. I put in the final exclusion Kibeho because it’s one of the massacres during the genocide and we’re looking for the total number.

We’re looking for search results from primary sources and reliable secondary sources.

Our first clue comes from here:

TAYLOR B. SEYBOLT, Major armed conflicts, in collaboration with the UPPSALA CONFLICT DATA PROJECT

PDF itself has little useful information otherwise. We can look later at the Uppsala Conflict Data Project.

Our second clue comes from the UN Refugee Agency report.

“The human cost of the conflict is tremendous. The widely cited figure of 5.4 million deaths between 1998 and 2007 has been challenged on methodological grounds, but numbers remain staggering. More conservative estimates claim a death toll of 3.3 million deaths in this timeframe.”

We need to find the source of this figure. The footnote here points to Tony Gambino, ‘Democratic Republic of the Congo’, World Development Report 2011 Background Case Study, World Bank, 2011, available at (last access 30 November 2011).

The URL to the WorldBank page seems to be dead, so I look for it on That website is refered to as ‘The Wayback Machine’. It allows to search through the “history” of the Internet to find a page. It’s basically the memory of the Internet! Here’s a short little video on how to get to the URL.

The PDF link is broken on the page, so we look for it on Google using these terms: “Tony Gambino Democratic Republic of Congo pdf” The PDF comes up as the first result.

The following extract starts to answer our question:

The International Rescue Committee‘s (IRC) estimate of 5.4 million excess deaths in the Congo between August 1998 and April 2007 has become a regularly cited figure in the press and by diplomats. Since 2000, the IRC has conducted a series of mortality surveys to obtain an estimate of ―excess mortality‖ in the DRC. The IRC method used rigorous, scientifically-valid methods to estimate mortality in particular areas, starting in eastern Congo. For the first survey, data were extrapolated in a manner that is not statistically representative, a limitation recognized openly from the start by the studies‘ authors and the IRC. … The validity of the IRC‘s methods and overall estimates have come under attack by some academics and others. Many of the criticisms are summarized in Chapter 3 of the Human Security Report 2009: The Shrinking Costs of War (HSR), entitled ―The Death Toll in the Democratic Republic of the Congo,‖49 which criticizes the IRC numbers on a number of grounds. This note will not discuss all the objections that have been raised. The methodology of the surveys appears quite sound, as was the decision to extrapolate. The IRC results have been published in three peer-reviewed medical journals.

We now need to look for the IRC and HSR reports. The former is the source of the claim and the latter source of the dispute.

The third search result is the IRC report with these terms: “International Rescue Committee Mortality Survey Congo 2000”.


2) Find the source of the dispute

The third search result is the HSR report with these terms: “Human Security Report 2009: The Shrinking Costs of War (HSR) pdf”.

We use the same technique to retrieve the subsequent reports by the committee. This is important to see if the numbers changed over time as more surveys were done. A summary is below:

IRC, 2000: 1.7 million excess deaths or more
IRC, 2001: 2.5 million. There was this note in the document: The overall death toll of 2.5 million fatalities was estimated, not measured. This estimation is fraught with potential for error.
IRC, 2003: ca. 3.3 million people
IRC, 2004: ca. 3.8 million people
IRC, 2007: 5.4 million excess deaths have occurred between August 1998 and April 2007. An estimated 2.1 million of those deaths occurred since the formal end of war in 2002.

On page 5 of the HSR report, we find the dispute to the IRC figure.


3) Find better estimates

Let’s look for alternative figures to check the claim by HSR report with these terms: “data death toll eastern congo genocide -wikipedia -answers -quora –news

This results show some promise, this one goes into the problems of counting and estimates, but don’t give alternative statistics.

I made the word data compulsory: “data” death toll eastern congo genocide -wikipedia -answers -quora -news

The fifth result is the most promising. It states on page 18 that 1.75 million is estimated by Leitenberg.

Try using the same techniques and remember to focus on reliable primary sources. (Hint: Wikipedia is not one of them.) Good luck!