期刊名称：Proceedings of the National Academy of Sciences
印刷版ISSN：0027-8424
电子版ISSN：1091-6490
出版年度：2020
卷号：117
期号：12
页码：6312-6313
DOI：10.1073/pnas.1922152117
出版社：The National Academy of Sciences of the United States of America
摘要：The recent paper by Coronese et al. (1) reports a rise in economic damages due to extreme natural disasters reported in the Emergency Events Database (EM-DAT) (2). While Coronese et al.’s (1) paper is timely and relevant, we have serious concerns regarding the analysis and the interpretation of the results. First, the dependent variable of Coronese et al.’s (1) main model is, according to the definition in EM-DAT, a measure of capital stock (CS) damages, and CS data should therefore be used as a control variable. While CS data are contained in the Penn World Table (PWT) (3), Coronese et al. (1) instead used gross domestic product (GDP) data from the same source. This is concerning because the trends estimated by ref. 1 could be due to temporal trends in CS/GDP ratios (Fig. 1). If CS data were used instead of GDP, Coronese et al.’s (1) main model would be D a i = α + β t i + γ CS c ( l i ) , t i + δ t i × CS c ( l i ).
