CRITICAL FINANCE REVIEW · Forthcoming· 2017
经济学的研究可以复制吗？——发表在13个期刊的60篇文章表明“多半不能”
作者：Andrew C. Chang (the Federal Reserve System)
Phillip Li (U.S. Department of the Treasury)
摘要：我们尝试使用作者提供的包括数据和代码的文件复制发表在13个广受好评的经济学期刊的67篇宏观经济学论文。除去其中6篇文章使用了机密数据，另外有35篇被要求在出版时提供数据和代码，我们获得了其中29篇（83%）文章的数据和代码复制文件。在26篇不要求在出版时提供数据和代码的文章中，我们获得了其中11篇文章的数据和代码文件。将复制成功定义为使用作者提供的数据和代码文件得到原文中的关键量化性结论。在没有联系作者的情况下，我们成功复制了67篇中的22篇（33%）。除去使用机密数据的6篇和缺乏处理软件的2篇，在作者的帮助下，我们成功复制了剩下59篇中的29篇。由于我们能复制的文章数量占总样本数不到一半，我们断言经济学研究多半无法复制。我们建议提高经济学研究的可复制性。
关键词：数据和代码文件，国内生产总值，GDP，宏观经济学，国民收入和生产账户，预分析，发表偏移，复制
Is Economics Research Replicable? Sixty Published Papers from Thirteen Journals Say “Often Not”
Andrew C. Chang (the Federal Reserve System), Phillip Li (U.S. Department of the Treasury)
ABSTRACT
We attempted to replicate 67 macroeconomic papers published in 13 well-regarded economics journals using author-provided replication files that included both data and code by following a preanalysis plan. Aside from 6 papers that used confidential data, we obtained data and code replication files for 29 of 35 papers (83%) that were required to provide such files as a condition of publication, compared to 11 of 26 papers (42%) that were not required to provide data and code replication files. Defining replication success as our ability to use the author-provided data and code files to produce the key qualitative conclusions of the original paper, we successfully replicated 22 of 67 papers (33%) without contacting the authors. Excluding the 6 papers that used confidential data and the 2 papers that used software we did not possess, we replicated 29 of 59 papers (49%) with assistance from the authors. Because we were able to replicate less than half of the papers in our sample even with help from the authors, we assert that economics research is often not replicable. We conclude with recommendations on improving replication of economics research.
Keywords: Data and Code Files, Gross Domestic Product, GDP, Macroeconomics, National Income and Product Accounts, Preanalysis, Publication Bias, Replication
原文链接：
http://cfr.ivo-welch.info/readers/2018/chang-li-2018.pdf
翻译：吴雨玲
