评《温度与决策：来自207,000个法院案件的证据》
Comment on “Temperature and Decisions: Evidence from 207,000 Court Cases”
American Economic Journal: Applied Economics · 2022 · Holger Spamann
中文摘要
Heyes和Saberian（2019b）利用2000—2004年的数据估计得出，室外气温会降低美国移民法官批准庇护申请的倾向。这一估计结果是由编码错误、数据错误以及样本选择造成的。修正这些错误后，点估计减少了三分之二，95%置信区间较宽且跨越零。将样本扩大至1990—2019年后，点估计的符号发生反转，并以极高的置信度排除了Heyes和Saberian所报告的效应大小。对美国联邦地区法院法官在1992年至2003年间作出的所有刑事量刑决定进行分析，也未发现温度或其他天气效应的证据。（JEL K37, K41, Q54）
Abstract
Heyes and Saberian (2019b) estimate from 2000–2004 data that outdoor temperature reduces US immigration judges’ propensity to grant asylum. This estimate is the result of coding and data errors and of sample selection. Correcting the errors reduces the point estimate by two-thirds, with a wide 95 percent confidence interval straddling zero. Enlarging the sample to 1990–2019 flips the point estimate’s sign and rules out the effect size reported by Heyes and Saberian with very high confidence. An analysis of all criminal sentencing decisions by US federal district judges from 1992 to 2003 yields no evidence of temperature or other weather effects either. (JEL K37, K41, Q54)