Econ Papers Daily / 期刊论文
回复《关于加拿大北部营养补贴传导率的评论》
Reply to Comment on “On the pass‐through rate of the Nutrition North Canada subsidy”
摘要
en In Naylor, Deaton, and Ker (2025), we discuss an appropriate econometric method for estimating the pass-through of the Nutrition North subsidy in the presence of heterogeneous effects and endogenous sampling. Applying this method to the data set compiled by Galloway and Li (2023), who conclude that there is “large leakage” of the subsidy, we fail to find evidence that supports their conclusion. After our paper was published, Galloway and Li published both a corrigendum to their 2023 article and a comment to our paper based on that corrigendum. In their comment, Galloway and Li (2026) claim that after “minor but easily justified” changes to their original data, our econometric approach leads to either imprecise or incomplete estimates of pass-through. In our response we show that: (i) Galloway and Li (2026) choose a classification of market structure that differs from the one recommended by federal government officials and from the classification they themselves use in their most recent work, including Galloway and Li (2025b) and Li, Daley, and Watson (2025); (ii) the classification structure consistent with the recommendations of federal government officials and their aforementioned research supports full pass-through; (iii) Galloway and Li's (2026) removal of 75% of the original sample is not minor; and (iv) Galloway and Li's (2026) repeated emphasis on sub-sampling, robustness, and statistical power is incomplete and thus misleading. Importantly, our response emphasizes that the correct econometric approach does not turn on whether one can selectively find a data set in which the empirical estimate of pass-through changes.
中文摘要
在Naylor、Deaton与Ker（2025）的研究中，我们探讨了在存在异质性效应与内生抽样条件下，估算“营养北方”补贴传导效应的适当计量方法。将该方法应用于Galloway与Li（2023）编制的数据集（该研究得出补贴存在“大量渗漏”的结论）后，我们未能发现支持其结论的证据。论文发表后，Galloway与Li针对其2023年文章发布了勘误，并基于该勘误对我们的论文发表了评论。在评论中，Galloway与Li（2026）声称，对其原始数据进行“微小但易于论证”的调整后，我们的计量方法对传导效应的估算要么不精确，要么不完整。我们的回应表明：（i）Galloway与Li（2026）选择的市场结构分类，既不同于联邦政府官员推荐的分类标准，也不同于其自身最新研究（包括Galloway与Li（2025b）及Li、Daley与Watson（2025））所使用的分类；（ii）与联邦政府官员建议及上述研究一致的分类结构支持完全传导效应；（iii）Galloway与Li（2026）剔除75%原始样本的做法并非“微小调整”；（iv）Galloway与Li（2026）反复强调子样本分析、稳健性与统计功效的做法不完整且具有误导性。重要的是，我们的回应强调：正确的计量方法并不取决于能否选择性找到传导效应实证估计值发生变化的数据集。