The fiscal cost of conflict: Evidence from Afghanistan 2005–2017. (September 2022)

• Record Type: Journal Article
• Authors: Barrett, Philip
• Is Part Of: World development. Volume 157(2022)
• Publication Date: 2022-09
• Subjects: Conflict -- Peace dividend -- Fiscal policy -- Afghanistan; Economic history -- 1990- -- Periodicals; Economic assistance -- Developing countries -- Periodicals; 330.9
• DOI: 10.1016/j.worlddev.2022.105940
• Languages: English
• ISSNs: 0305-750X

Abstract

Highlights: Collects novel data on monthly provincial fiscal performance in Afghanistan Matches to conflict deaths to measure within-country impact of conflict. Extends a synthetic control method to address omitted variables bias. Provides summary measures of total the fiscal loss from conflict in Afghanistan. Abstract: I use a novel monthly panel of provincially-collected central government revenues and conflict fatalities to estimate government revenues lost due to conflict in Afghanistan since 2005. Headline estimates are large, implying total revenue losses of $3bn since 2005 and that gains from peace would have been about 6 percent of GDP per year. That this is larger than estimates in cross-country studies reflects the uncommon intensity of the conflict in Afghanistan. The key challenge to identification is omitted variable bias, which I address by extending Powell's (2021) generalized synthetic control method to a dynamic setting. This allows estimation of impulse response functions robust to a broad class of omitted variables bias.