Econometric vs. Causal Structure-Learning for Time-Series Policy Decisions: Evidence from the UK COVID-19 Policies
arXiv:2603.00041v1 Announce Type: new Abstract: Causal machine learning (ML) recovers graphical structures that inform us about potential cause-and-effect relationships. Most progress has focused on cross-sectional data with no explicit time order, whereas recovering causal structures from time series data remains the subject of ongoing research in causal ML. In addition to traditional causal ML, this study assesses econometric methods that some argue can recover causal structures from time series data. The use of these methods can be explained by the significant attention the field of econometrics has given to causality, and specifically to time series, over the years. This presents the possibility of comparing the causal discovery performance between econometric and traditional causal ML algorithms. We seek to understand if there are lessons to be incorporated into causal ML from econometrics, and provide code to translate the results of these econometric methods to the most widely
arXiv:2603.00041v1 Announce Type: new Abstract: Causal machine learning (ML) recovers graphical structures that inform us about potential cause-and-effect relationships. Most progress has focused on cross-sectional data with no explicit time order, whereas recovering causal structures from time series data remains the subject of ongoing research in causal ML. In addition to traditional causal ML, this study assesses econometric methods that some argue can recover causal structures from time series data. The use of these methods can be explained by the significant attention the field of econometrics has given to causality, and specifically to time series, over the years. This presents the possibility of comparing the causal discovery performance between econometric and traditional causal ML algorithms. We seek to understand if there are lessons to be incorporated into causal ML from econometrics, and provide code to translate the results of these econometric methods to the most widely used Bayesian Network R library, bnlearn. We investigate the benefits and challenges that these algorithms present in supporting policy decision-making, using the real-world case of COVID-19 in the UK as an example. Four econometric methods are evaluated in terms of graphical structure, model dimensionality, and their ability to recover causal effects, and these results are compared with those of eleven causal ML algorithms. Amongst our main results, we see that econometric methods provide clear rules for temporal structures, whereas causal-ML algorithms offer broader discovery by exploring a larger space of graph structures that tends to lead to denser graphs that capture more identifiable causal relationships.
Executive Summary
This article presents a comparative analysis of econometric and causal structure-learning methods for time-series policy decisions, focusing on the UK COVID-19 policies. The study assesses four econometric methods and eleven causal machine learning algorithms in terms of graphical structure, model dimensionality, and their ability to recover causal effects. The results suggest that econometric methods provide clear rules for temporal structures, while causal ML algorithms offer broader discovery, leading to denser graphs that capture more identifiable causal relationships. The study provides code to translate econometric results into the bnlearn library, enabling comparison and integration of the two approaches. The findings have implications for policy decision-making and highlight the potential benefits of incorporating econometric methods into causal ML frameworks.
Key Points
- ▸ Comparison of econometric and causal machine learning methods for time-series policy decisions
- ▸ Assessment of four econometric methods and eleven causal ML algorithms
- ▸ Investigation of graphical structure, model dimensionality, and causal effect recovery
- ▸ Code provided to translate econometric results into the bnlearn library
Merits
Strength in Temporal Structure Analysis
Econometric methods provide clear rules for temporal structures, enabling more accurate policy decision-making.
Broader Discovery with Causal ML
Causal machine learning algorithms offer broader discovery, leading to denser graphs that capture more identifiable causal relationships.
Comprehensive Comparison
The study provides a comprehensive comparison of econometric and causal ML methods, enabling researchers to understand the strengths and limitations of each approach.
Demerits
Limited Generalizability
The study focuses on the UK COVID-19 policies, which may limit the generalizability of the findings to other contexts.
Computational Complexity
Causal ML algorithms can be computationally intensive, which may be a challenge for large-scale policy decision-making applications.
Expert Commentary
The article presents a timely and relevant contribution to the field of causal inference in time-series data. The comparison of econometric and causal ML methods highlights the strengths and limitations of each approach, providing valuable insights for researchers and policy makers. However, the study's focus on the UK COVID-19 policies may limit the generalizability of the findings. Furthermore, the computational complexity of causal ML algorithms may be a challenge for large-scale policy decision-making applications. Nonetheless, the study's findings have significant implications for policy decision-making and highlight the potential benefits of integrating econometric methods into causal ML frameworks.
Recommendations
- ✓ Researchers should investigate the generalizability of the findings to other contexts and consider extending the study to other policy domains.
- ✓ Policy makers should consider incorporating econometric methods into their decision-making frameworks to improve temporal structure analysis.
- ✓ Researchers should explore ways to reduce the computational complexity of causal ML algorithms to enable their use in large-scale policy decision-making applications.
