发表机构
Charles University; Institute of Economic Studies, Charles University; Centre for Economic Policy Research; Meta-Research Innovation Center at Stanford; Anglo-American University（查理大学; 查理大学经济研究所; 经济政策研究中心; 斯坦福大学元研究创新中心; 英美大学）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
本文通过元分析1091个估计值，校正发表偏倚后，将本地与移民劳动力替代弹性从约13提升至约22，最佳估计约为17，表明移民对相对工资的压力比未校正结果低约40%。
AI 中文摘要
本文首次对本地与移民劳动力之间的替代弹性进行了全面的元分析，利用了来自41项研究的1091个估计值。我们发现了选择性报告的强有力证据：精度较低的估计值系统性地与较低的报告弹性相关联。使用元回归和选择方法对这一偏差进行校正后，隐含弹性从约13提高到约22，这意味着移民带来的相对工资压力比未校正结果所暗示的少约40%。贝叶斯和频率学派模型平均显示，异质性主要由地理尺度、数据粒度以及样本是否限于低经验工人所驱动，而对数平均工资与平均对数工资之间的选择则起次要作用。我们的最佳实践估计剔除了发表偏倚并优先考虑最细粒度的数据，在基线区域设定中隐含弹性约为17，低于简单偏差校正所暗示的值，但显著高于未校正的均值。
英文摘要
This paper presents the first comprehensive meta-analysis of the elasticity of substitution between native and immigrant labor, drawing on 1,091 estimates from 41 studies. We find strong evidence of selective reporting: less precise estimates are systematically associated with lower reported elasticities. Correcting for this bias using meta-regression and selection methods raises the implied elasticity from about 13 to about 22, implying about 40% less relative-wage pressure from immigration than uncorrected results suggest. Bayesian and frequentist model averaging show that heterogeneity is driven mainly by geographic scale, data granularity, and whether the sample is restricted to low-experience workers, while the choice between log mean wages and mean log wages plays a secondary role. Our best-practice estimates, which net out publication bias and prioritize the most granular data, imply an elasticity of about 17 in our baseline regional specification, lower than implied by a simple bias correction but substantially higher than the uncorrected mean.
Comments47 pages, 21 figures, 10 tables. JEL: J15, J61, C83. Data and code available at https://meta-analysis.cz/migrant/
