2018/05/14 by Petr Polák · 25 citations
Economics, Econometrics and Finance · Mathematics · #Bayesian inference #Bayesian probability #Econometrics #Economic Policies and Impacts #Economics #Estimation #Global trade and economics #Mathematics #Meta-analysis #Monetary Policy and Economic Impact #Publication bias #Statistics
paper · doi:10.1111/joes.12264
published in Journal of Economic Surveys 33(1), 101-124 (Wiley)
openalex publication_date 2018/05/14 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/22
Abstract Many studies have estimated the trade effect of the euro, but their results vary greatly. This meta‐analysis collects 3323 estimates of the euro effect along with 28 characteristics of estimation design from almost 60 studies and quantitatively examines the literature. The results show evidence of publication bias, but they also suggest that the bias decreases over time. After correcting for the bias, the meta‐analysis shows that the literature is consistent with an effect ranging between 2% and 6%. The results from Bayesian model averaging, which takes into account model uncertainty, show that the differences among estimates are systematically driven by data sources, data structure, control variables, and estimation techniques. The mean reported estimate of the euro's trade effect conditional on best‐practice approach is 3%, but is not statistically different from zero.