Estimating Conflict-Related Mortality in Timor-Leste, 1974-1999: A Comparative Review of Demographic and Statistical Estimation Methods

Romesh Silva, Human Rights Data Analysis Group, Benetech
Patrick Ball, Human Rights Data Analysis Group, Benetech

When we consider mass human rights violations, among the first questions is always “How many?”. Scientifically-defensible answers about the statistical patterns and overall magnitude of violence can depoliticize arguments about the past. However, quantitative analysis of the magnitude and pattern of conflict-related mortality in protracted conflicts is often difficult due to limited data and the non-stationary nature of populations during violent conflicts. This paper draws on a multiplicity of data sources in the form of survey data, administrative data and convenience sample, testimony data and estimates conflict-related mortality in Timor-Leste between 1974 and 1999. Using these data, we apply survey-based estimates, multiple-systems estimation and indirect demographic estimation using the sibling-survivor method. The results of the different estimation techniques are presented and a detailed discussion of their respective strengths and weaknesses is presented.