本仓库聚焦合成控制法（Synthetic Control Method, SCM） 及其拓展方法的代码实现与应用指南，旨在为因果推断领域的研究者、学生及实务工作者提供一站式技术资源。
仓库核心内容涵盖三部分：
- 基础实现：包含经典SCM的核心代码（如Abadie经典框架），支持面板数据的处理、权重估计与反事实构建，附详细注释与示例数据集；
- 前沿拓展：更新近年主流改进方法，重点覆盖溢出效应缓解（如迭代合成法、预设结构法）、多处理组SCM、时间异质性SCM等热点技术；
- 应用工具：提供学术论文复现案例（如政策评估、区域经济研究）。
代码以Python（Pandas、Scikit-learn）和R（Synth包、CausalImpact）为主，兼顾易用性与可扩展性，可直接适配不同研究场景的数据格式。欢迎星标（Star）关注更新，也期待通过Issues或Pull Request交流改进建议，共同推动SCM方法的实践落地。
This repository focuses on the code implementation and application guidelines of the Synthetic Control Method (SCM) and its extensions, aiming to provide a one-stop technical resource for researchers, students, and practitioners in the field of causal inference.
The core content of the repository includes three parts:
- Basic Implementation: Core code for classical SCM (e.g., Abadie’s framework), supporting panel data processing, weight estimation, and counterfactual construction, with detailed comments and sample datasets;
- Cutting-Edge Extensions: Updated with mainstream improved methods in recent years, focusing on spillover effect mitigation (e.g., Iterative SCM, Preset Structure Method), multi-treatment SCM, and time-heterogeneous SCM;
- Application Tools: Functions for result visualization (e.g., placebo test plots, dynamic effect plots), code for robustness checks (e.g., permutation tests, weight sensitivity analysis), and reproducible cases of academic papers (e.g., policy evaluation, regional economic research).
Codes are mainly written in Python (Pandas, Scikit-learn) and R (Synth package, CausalImpact), balancing usability and scalability, and can be directly adapted to data formats in different research scenarios. Feel free to star the repository for updates, and we welcome suggestions via Issues or Pull Requests to jointly promote the practical application of SCM methods.
欢迎使用本合成控制法资源库。为帮助您更高效地检索与学习，我们强烈建议先阅读本仓库的核心索引文件：合成控制法文献梳理.xlsx 。
该表格系统梳理了合成控制法领域绝大多数方法类文献，具备以下特点：
-
内容分类清晰：按方法类型、发展脉络等维度对文献进行归类；
-
要点提炼精准：每篇文献均附有主要内容摘要与内容类型；
-
适用层次广泛：无论您是初学者还是进阶研究者，都能快速定位所需信息。
此外，我们还为文献整理了可复现的官方或社区代码，放置于对应文件夹中。您可结合文献原文与代码进行对照学习，深入理解各类合成控制方法的实现过程与技术细节。
我们衷心希望本仓库能成为您学习与应用合成控制法的得力工具，如有任何问题或建议，欢迎参与维护与讨论。
Welcome to the Synthetic Control Methods Resource Repository. To help you navigate and learn more efficiently, we strongly recommend starting with our core index file: 合成控制法文献梳理.xlsx (Synthetic Control Methods Literature Review.xlsx).
This spreadsheet systematically organizes most methodological literature in the field of synthetic control methods and offers the following features:
- Clear Categorization: Literature is classified by method type, developmental trajectory, and other dimensions.
- Precise Summaries: Each entry includes a summary of main content and content type.
- Wide Applicability: Whether you are a beginner or an advanced researcher, you can quickly locate the information you need.
In addition, we have compiled reproducible official or community code for the literature, stored in corresponding folders. You can compare the original papers with the code to deepen your understanding of the implementation process and technical details of various synthetic control methods.
We sincerely hope this repository becomes a valuable tool for your study and application of synthetic control methods. If you have any questions or suggestions, you are welcome to participate in maintenance and discussions.
| File Name | Authors | Year | Title | DOI |
|---|---|---|---|---|
| hcw-data | Hsiao et al. | 2012 | A Panel Data Approach for Program Evaluation -- Measuring the Benefits of Political and Economic Integration of Hong Kong with Mainland China | 10.1002/jae.1230 |
| Abadie-2015 | Abadie et al. | 2015 | Comparative Politics and the Synthetic Control Method | 10.1111/ajps.12116 |
| gsynth-master | Xu | 2017 | Generalized Synthetic Control Method: Causal Inference with Interactive Fixed Effects Models | 10.1017/pan.2016.2 |
| robust control-main | Amjad et al. | 2017 | Robust Synthetic Control | 10.48550/arXiv.1711.06940 |
| augsynth | Ben-Michael et al. | 2018 | The Augmented Synthetic Control Method | 10.3386/w28885 |
| Firpo-2018 | Firpo and Possebom | 2018 | Synthetic Control Method: Inference, Sensitivity Analysis and Confidence Sets | 10.1515/jci-2016-0026 |
| bscm-master | Kim | 2020 | Bayesian Synthetic Control Methods | 10.1177/0022243720936230 |
| Esposti-2020 | Esposti et al. | 2020 | Can synthetic controls improve causal inference in interrupted time series evaluations of public health interventions? | 10.1093/ije/dyaa152 |
| scul-master | Hollingsworth and Wing | 2020 | Tactics for design and inference in synthetic control studies: An applied example using high-dimensional data | 10.2139/ssrn.3592088 |
| Chernozhukov-2021-JASA | Chernozhukov et al. | 2021 | An Exact and Robust Conformal Inference Method for Counterfactual and Synthetic Controls | 10.1080/01621459.2021.1920957 |
| Distributional conformal prediction | Chernozhukov et al. | 2021 | Distributional conformal prediction | 10.1073/pnas.2107794118 |
| Ferman-2021-JASA | Ferman | 2021 | On the Properties of the Synthetic Control Estimator with Many Periods and Many Controls | 10.48550/arXiv.1906.06665 |
| Ferman-2021-QE | Ferman and Pinto | 2021 | Synthetic controls with imperfect pretreatment fit | 10.3982/QE1596 |
| Fernandez 2021 | Fernandez et al. | 2021 | Low-rank approximations of nonseparable panel models | 10.1093/ectj/utab007 |
| Imai-2021 | Imai et al. | 2021 | Matching Methods for Causal Inference with Time-Series Cross-Sectional Data | 10.1111/ajps.12685 |
| MASC | Kellogg et al. | 2021 | Combining Matching and Synthetic Control to Tradeoff Biases From Extrapolation and Interpolation | 10.1080/01621459.2021.1979562 |
| pensynth | Abadie and L'Hour | 2021 | A Penalized Synthetic Control Estimator for Disaggregated Data | 10.1080/01621459.2021.1971535 |
| RPCA | Bayani | 2021 | Robust PCA Synthetic Control | 10.48550/arXiv.2108.12542 |
| scpi-main | Cattaneo et al. | 2021 | Prediction Intervals for Synthetic Control Methods | 10.1080/01621459.2021.1979561 |
| synthdid-sdid-paper | Arkhangelsky et al. | 2021 | Synthetic Difference in Differences | 10.1257/aer.20190159 |
| treebased synthetic controls-main | Mühlbach and Nielsen | 2021 | Tree-based synthetic control methods: Consequences of relocating the US embassy | 10.48550/arXiv.1909.03968 |
| synthetic learner | Viviano and Bradic | 2022 | Synthetic Learner: Model-free inference on treatments over time | 10.48550/arXiv.1904.01490 |
| CAPPMx-main | Chandra et al. | 2023 | Bayesian Nonparametric Common Atoms Regression for Generating Synthetic Controls in Clinical Trials | 10.48550/arXiv.2201.00068 |
| Cattaneo-2023 | Cattaneo et al. | 2023 | Uncertainty Quantification in Synthetic Controls with Staggered Treatment Adoption | 10.48550/arXiv.2210.05026 |
| Klinenberg-2023 | Klinenberg | 2023 | Synthetic Control with Time Varying Coefficients A State Space Approach with Bayesian Shrinkage | 10.1080/07350015.2022.2102025 |
| Replication DSC ECMA | Gunsilius | 2023 | DISTRIBUTIONAL SYNTHETIC CONTROLS | 10.3982/ECTA18260 |
| shixu-2023 | Shi et al. | 2023 | Theory for Identification and Inference with Synthetic Controls: A Proximal Causal Inference Framework | 10.48550/arXiv.2108.13935 |
| synthetic-combinations-main | Agarwal et al. | 2023 | Synthetic Combinations: A Causal Framework for Combinatorial Interventions | 10.48550/arXiv.2303.14226 |
| DR Proximal SC-main | Qiu et al. | 2024 | Doubly robust proximal synthetic controls | 10.48550/arXiv.2210.02014 |
| Liu-2024-AJPS | Liu et al. | 2024 | A Practical Guide to Counterfactual Estimators for Causal Inference with Time-Series Cross-Sectional Data | 10.1111/ajps.12723 |
| Spillover-SCM-main | Melnychuk | 2024 | Synthetic Controls with Spillover Effects: A Comparative Study | 10.48550/arXiv.2405.01645 |
| Ylmaz-2024-J0M | Yılmaz et al. | 2024 | Causal inference under selection on observables in operations management research: Matching methods and synthetic controls | 10.1002/joom.1318 |