This paper establishes electricity consumption as an indicator for tracking economic fluctuations in Bangladesh. It presents monthly data on national electricity consumption since 1993 and subnational daily consumption data since February 2010. Electricity consumption is strongly correlated with other high-frequency indicators of economic activity, and it has declined during natural disasters and the COVID-19 lockdowns. The paper estimates an electricity consumption model that explains over 90% of the variation in daily consumption based on a quadratic trend, seasonality, within-week variation, national holidays, Ramadan, and temperature. Deviations from the model prediction can act as an indicator of subnational economic fluctuations. For example, electricity consumption in Dhaka fell around 40% below normal in April and May 2020 during the first COVID-19 lockdown and remained below normal afterwards. The later lockdowns, in contrast, had much smaller impacts, in line with less stringent containment measures and more effective adaptation.

本文将用电量作为跟踪孟加拉国经济波动的指标。它提供了自 1993 年以来的国家电力消耗月度数据和自 2010 年 2 月以来的地方日常消耗数据。电力消耗与经济活动的其他高频指标密切相关，并且在自然灾害和 COVID-19 封锁期间有所下降。该论文估算了一个用电量模型，该模型基于二次趋势、季节性、周内变化、国定假日、斋月和温度来解释超过 90% 的日常消耗变化。模型预测的偏差可以作为地方经济波动的指标。例如，在第一次 COVID-19 封锁期间，达卡的用电量在 2020 年 4 月和 2020 年 5 月下降了约 40%，此后一直低于正常水平。相比之下，后来的封锁产生的影响要小得多，这与不那么严格的遏制措施和更有效的适应相一致.