Saenger Till Raphael, Kapstein Ethan B, Sircar Ronnie
Operations Research and Financial Engineering, Princeton University, Princeton, New Jersey, United States of America.
Empirical Studies of Conflict Project (ESOC), School of Public and International Affairs, Princeton University, Princeton, New Jersey, United States of America.
PLoS One. 2024 Dec 13;19(12):e0315337. doi: 10.1371/journal.pone.0315337. eCollection 2024.
The Taliban's takeover of Afghanistan in August 2021 is associated with a rapid collapse of the Afghan economy. However, assessing the scale of this collapse is proving difficult as official data are scarce. To complement qualitative measures obtained through rapid surveys of the population, we employ monthly nightlights data as a proxy measure for changes in economic activity. By combining a synthetic control approach with nightlights data from neighboring countries, our analysis reveals a significant shift in Afghanistan's economic trajectory: from positive growth to a deep recession, even considering the impact of the Covid pandemic. Our estimations suggest that Afghanistan's GDP has declined by approximately 16% from 2020 to 2022, notably less than the World Bank's current survey-based measure of a 28% decline in 2021 alone. In contrast to other available estimates, our reporting includes confidence intervals to convey the uncertainties surrounding these point estimates. This study showcases the potential applicability of our methodology and the use of appropriately processed monthly nightlights data in scenarios where administrative data is limited or unreliable.
2021年8月塔利班接管阿富汗，这与阿富汗经济的迅速崩溃有关。然而，由于官方数据稀缺，评估此次崩溃的规模颇具难度。为补充通过对民众进行快速调查获得的定性指标，我们使用月度夜光数据作为经济活动变化的替代指标。通过将合成控制法与来自邻国的夜光数据相结合，我们的分析揭示了阿富汗经济轨迹的显著转变：即使考虑到新冠疫情的影响，也从正增长转向了深度衰退。我们的估计表明，阿富汗的国内生产总值在2020年至2022年间下降了约16%，明显低于世界银行目前基于调查得出的仅2021年就下降28%的衡量标准。与其他现有估计不同，我们的报告包含置信区间，以传达这些点估计周围的不确定性。本研究展示了我们方法的潜在适用性，以及在行政数据有限或不可靠的情况下，使用经过适当处理的月度夜光数据的可行性。
