Rationale

Civil wars, often related to ethnic conflicts, have become increasingly common throughout the world in recent years. They impose enormous and long lasting human, social and economic costs, and have repercussions well beyond the main theatre of combat. However, estimating the full costs of such conflicts is a difficult, if not impossible, task. The human and social costs of death, disability, dispossession, and the psychological trauma associated with violence and terror are not really quantifiable. On the other hand, at least in principle, the “purely economic costs” are amenable to quantitative measurement.

Objectives

The primary aim of this study was to assess the economic costs of the protracted ethnic conflict in Sri Lanka. However, the study also presents a critical review of the analytical approaches adopted in previous studies, and clarifies some important conceptual and methodological issues. The study attempted the following: to present a brief background on the Sri Lankan ethnic conflict; to discuss issues pertaining to the methodology of estimating the cost the war; to offer a critique of previous studies on the subject; to provide a methodological framework for Sri Lanka; and to estimate the cost of the war in Sri Lanka.

A war affects not only the current level of output, but also the level and efficiency of investments, and, hence, it lowers the long-term growth rate. The paper calculates the direct and indirect costs both in the short- and the long-term. Estimates of direct costs include military costs borne by the government and the LTTE, and costs of damage to physical and social infrastructure. Indirect costs are calculated by comparing the actual output levels for different affected areas with that of the counterfactual output levels that would have been attained in the absence of war. On conservative assumptions, the study suggests that the war may have cost the equivalent of 1.7 times Sri Lanka's 1996 GDP.