arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2107.02637econ.EM
收藏

连续处理的差异-差异法

Difference-in-Differences with a Continuous Treatment

University of Georgia(佐治亚大学);Federal Reserve Bank of Minneapolis(明尼阿波利斯联邦储备银行);NBER(美国国家经济研究局);Emory University(埃默里大学)

Brantly Callaway, Andrew Goodman-Bacon, Pedro H. C. Sant'Anna

AI总结:

本文探讨了连续处理差异-差异法的识别问题,提出替代假设以消除选择偏差,并介绍了新的估计方法以改进处理效应的分析。

摘要:

本文分析了具有连续处理的差异-差异设计。我们证明,在类似于二元处理情况下的平行趋势假设下,处理-处理型参数是可识别的。然而,由于平行趋势不能排除选择偏差,跨处理比较这些参数具有挑战性。我们讨论了替代的、通常更强的假设,以消除选择偏差。我们进一步表明,流行的双向固定效应估计量允许多种解释,取决于底层因果构建块,所有都有重要的局限性,作为处理效应的有意义总结。最后,我们介绍了一种替代的估计程序,以避免这些缺点,并在实证应用中展示了它们。

英文摘要:

This paper analyzes difference-in-differences designs with a continuous treatment. We show that treatment-on-the-treated-type parameters are identified under a parallel trends assumption analogous to the binary treatment case. However, comparing these parameters across treatments is challenging because parallel trends does not rule out selection bias. We discuss alternative, typically stronger, assumptions that eliminate selection bias. We further show that popular two-way fixed effects estimands admit multiple interpretations, depending on the underlying causal building block, all having important limitations as meaningful summaries of treatment effects. Finally, we introduce alternative estimation procedures that avoid these drawbacks and demonstrate them in an empirical application.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑