Names and numbers: how the victims of the Kosovo war are counted

The question "how many people were killed in the Kosovo war?" looks simple. It is not. The answer depends on the period counted, on who is considered a victim of the war, on how the missing are treated and on which sources are used. This article explains how the victims have been counted, why the figures differ and why, despite all of this, every figure must lead back to names.

Why numbers matter — and why they are not enough

Accurate numbers are essential for justice, for memory and for countering denial. Inflated and deflated figures do equal harm: the former hand arguments to deniers, the latter erase real victims. For this reason the archive does not round figures for effect and does not publish unverified lists.

But a number, however accurate, does not tell us who a person was, what they left behind or how their family lived after the loss. That is why the remembrance space in this archive starts from a simple principle: names are not numbers.

The Kosovo Memory Book: documentation by name

The most systematic project to count the victims of the Kosovo war is the Kosovo Memory Book, led by the Humanitarian Law Center together with its sister organization in Kosovo [1] [2]. Its principle is simple and demanding: a person is counted only when they can be identified by name and when their case is supported by documented sources.

According to HLC's February 2015 report, the database recorded 13,535 people killed or missing in connection with the war between 1 January 1998 and 31 December 2000 [1]. The figure includes people of all ethnic backgrounds, both civilians and members of armed formations. The database is updated over time: a later, undated HLC page shows 13,625. When we cite a figure, therefore, we cite it with its year and source.

Definitions: who is counted, and how

Three distinctions are essential. The first: killed or missing. A missing person is not recorded as dead until their remains are found and identified; for the family, this uncertainty is a distinct wound in itself. The second: civilian or member of an armed formation. The distinction matters in law, but it is often difficult to establish for people killed away from the fighting. The third: the connection between the death and the war.

A good database makes these categories visible and does not mix them. When two sources give different figures, the first question is not "who is lying?" but "are they counting the same thing?".

The period: war and aftermath

The period covered by the Kosovo Memory Book, 1998–2000, includes the months after June 1999. After the withdrawal of Serbian and Yugoslav forces and the arrival of KFOR, Serbs, Roma, Albanians and members of other communities were also killed and went missing; many of these crimes remain unsolved [3]. An overall figure for 1998–2000 should therefore not be read as a figure for the war alone, nor as a figure for a single group.

This archive separates wartime events from those of the postwar period precisely so that the two histories are not confused and neither is used to deny the other.

An independent evaluation of the methodology

A casualty database becomes more credible when its methodology is examined from outside. The Kosovo Memory Book was evaluated by the Human Rights Data Analysis Group (HRDAG), an organization specializing in the statistical analysis of human-rights violations [4]. The evaluation examined how records were collected and verified and how far analysis could rely on them.

For the reader, the importance of this step lies not only in the evaluation's conclusions but in the principle itself: casualty figures must be checkable. HRDAG's detailed conclusions should be read directly in its document, which the archive links in the source registry.

Why the figures differ

During and immediately after the war, many estimates circulated from governments, international organizations, researchers and local organizations. Many were made before bodies had been exhumed, while families still did not know the fate of their relatives, and before concealed mass graves, such as those later discovered at Batajnica, were known. Early estimates were not necessarily dishonest; they simply could not be verified at the time.

The main reasons for the differences are different time periods; whether the missing are included; whether members of armed formations are included; different methods, such as counting by name versus statistical estimation; and when the research was done. When the archive presents differing figures, it states the definition, period and source of each.

The missing

For more than a thousand families, the question is not how many people died, but where their own person is. The number of people still being sought changes with every new identification, so the archive always cites it with its source and date of publication. The search for the missing is part of justice, not only of statistics.

How this archive uses these figures

• Every figure is cited with its source, year of publication, period and definition.
• No lists of names are published without a licence, verification and care for families.
• Victims of the war and of the postwar period are presented separately, without mixing categories.
• Figures are not used to compare the suffering of peoples or to assign collective blame.
• Memorial profiles are built only with consent and sources.