21 March 2005

To whom it may concern:

With this letter I want to recommend Mirsad Tokaca and the Research and Documentation Center (RDC) of Sarajevo, Bosnia and Herzegovina. I have known Mirsad since September, 2004. It is perhaps more relevant that I have been working with the RDC database since last year.

Since 1991, I have designed and implemented information management systems and conducted quantitative analysis for large-scale human rights data projects for truth commissions, non-governmental organizations, tribunals and United Nations missions in El Salvador, Ethiopia, Guatemala, Haiti, South Africa, Kosovo, Sierra Leone, Chad, Sri Lanka, Perú, Colombia, and East Timor.

Much of my work has been finding ways to use human rights databases to calculate the total documented events, including killings, disappearance, forced migration, torture, and detention. The major complexity in this work is not to determine what is known through the databases: that part is relatively easy. The hard part is figuring out how much of the relevant phenomenon has never been documented. The unknown part of the social history of mass violence is often greater than the part that is recorded in a society's historical memory.

The unreported violence can overwhelm any attempt to make sense of the known patterns. If 10,000 killings happened in a certain place, but only 100 of those killings are reported, statistical analysis done on the reported killings will have little chance of properly representing the true patterns. There are statistical techniques that can be used to estimate the undocumented deaths, and the RDC database is ideally designed for such calculations. Partisan positions in a post-conflict society often argue about who has been responsible for the majority of violations. It is in each side's interest to portray themselves as victims and their opponents as the aggressors.

Great powers are occasionally complicit in the creation of this false moral equivalence: in the interests of short-term agreements, some mediators recommend that each side accept some responsibility and then put aside the questions of truth about the past and accountability in the present.

In the context of mass atrocities, where hundreds of thousands of people were killed and disappeared, the contending positions cannot be reconciled while the argument is made only at the anecdotal level. Even if true responsibility is grossly asymmetrical, where one side committed dozens of times more violations than the other, in an anecdotal debate, each side can continue to raise new cases long after anyone has stopped listening. Only rigorously and transparently calculated statistics offer a way out of this dilemma.

A statistical understanding of mass violence is the way to avoid both problems: good statistics avoid analytic artifacts resulting from underreporting and they transcend the limitations of anecdotal argument. This is the achievement that Mirsad and the RDC are on the brink of reaching in Bosnia-Herzegovina. I have built databases around the world, including seven truth commissions. The database that Mirsad and his colleagues at the RDC have built is, quite literally, the largest ever collected. It includes 259,969 individually documented reports. When the duplicate reports are noted, there are more than 86,068 unique, fully identified individual victims.

The richness and depth of this database is astonishing. What I find even more impressive is the meticulous care with which they have assembled the reports. All of the reports about each unique victim are available individually. There are over fourteen independent streams of information combined in this database. That means that an individual victim could, in theory, appear in all fourteen sources.

I presented evidence of this kind as an expert witness in the trial of Slobodan Milosević using databases that documented deaths and migration in Kosovo in 1999. (Our databases were much smaller than the one RDC has compiled) My argument was that this evidence refutes claims by the defense that the killings and migration could have been caused by NATO or the KLA guerrillas. While we are still waiting for the judges' decision, they engaged the evidence carefully, and seemed to understand its import.

In a very preliminary analysis, I have grouped the RDC database sources into four types (military, press, family or witness testimony, and administrative sources). More than 69% of the records include more than one of these sources, which is a higher density of reporting than I've seen in any previous project.

Organizing the data in this way means that the RDC data could be used to calculate definitive estimates of the total deaths in in the war in Bosnia- Herzegovina. This would include combat deaths in all the major forces, as well as victims from each of the various ethnic groups, considered over time, space, and type of perpetrator. This data is richer than anything currently available to the International Criminal Tribunal for the Former Yugoslavia. The estimates that could be produced by this database would consequently be more reliable than existing analysis.

The RDC database could have a significant impact on how the history of the Bosnian war is understood. From my conversations with Mirsad and the staff of the RCD, I think that this potential is largely the result of their vision. For more than ten years, this team has been steadfast in their focus on the importance of high-quality documentation.

On behalf of my organization, I have committed the Human Rights Data Analysis Group to work with RDC on analysis of their database. We believe that this analysis will be an enormous contribution to the social history of the Balkan wars of the 1990s, to the prosecutions in The Hague, and to the historical memory of the victims.

Please do not hesitate to contact me for more information about Mirsad Tokaca or the database of the Research and Documentation Centre.

Sincerely,

Patrick Ball, Ph.D.

Director, Human Rights Program

The Benetech Initiative

480 S. California Ave, Suite 201

Palo Alto, CA 94306

pball@martus.org