📄 中文摘要
因果机器学习（ML）能够恢复图形结构，从而揭示潜在的因果关系。大多数进展集中在没有明确时间顺序的横截面数据上，而从时间序列数据中恢复因果结构仍然是因果机器学习研究的热点。除了传统的因果机器学习外，研究还评估了计量经济学方法，这些方法被认为能够从时间序列数据中恢复因果结构。计量经济学领域对因果关系，特别是时间序列的关注为这些方法的使用提供了理论基础。这为计量经济学与传统因果机器学习之间的因果发现性能比较提供了可能性。
计量经济学与因果结构学习在时间序列政策决策中的比较：来自英国 COVID-19 政策的证据
📄 English Summary
Econometric vs. Causal Structure-Learning for Time-Series Policy Decisions: Evidence from the UK COVID-19 Policies
Causal machine learning (ML) recovers graphical structures that reveal potential cause-and-effect relationships. Most advancements have focused on cross-sectional data without explicit time ordering, while recovering causal structures from time series data remains a key area of ongoing research in causal ML. In addition to traditional causal ML, this study evaluates econometric methods that are argued to recover causal structures from time series data. The significant emphasis that the field of econometrics has placed on causality, particularly in relation to time series, supports the application of these methods. This creates an opportunity to compare the causal discovery performance between econometric approaches and traditional causal ML.