Python packages
(Last checked: 08 Sep 2026)
Only dedicated DiD packages are listed below. Add the package version and Python requirement to reproducible analyses; GitHub activity does not imply a stable release.
| Name | Installation | Package by | Reference paper | |
|---|---|---|---|---|
| csdid | GitHub | Alexander Quispe; Alexander Fischer | Callaway and Sant’Anna (2021) | |
| csa-py | GitHub | Jonas Skjold Raaschou-Pedersen | Callaway and Sant’Anna (2021) | |
| differences (0.3.0) | pip install differences (Python >= 3.10) |
PyPI and documentation | Bernardo Dionisi and contributors | Callaway and Sant’Anna (2021); Sant’Anna and Zhao (2020); Rambachan and Roth (2023) |
| diff-diff | GitHub | Isaac Gerber and contributors | Callaway and Sant’Anna (2021); Arkhangelsky et al. (2021) | |
| did-imputation | GitHub | Jonas Skjold Raaschou-Pedersen | Borusyak, Jaravel, Spiess (2024) | |
| did-stepwise | GitHub | Jonas Skjold Raaschou-Pedersen | Harmon (2024) | |
| moderndid | GitHub | Jordan de Klerk and contributors | Modern staggered/adoption DiD estimators | |
| pysynthdid | GitHub | Masa Asami | Arkhangelsky et al. (2021) | |
| synthdid.py | GitHub | d2cml-ai contributors | Arkhangelsky et al. (2021) | |
| diffindiff | GitHub | Geo Wieland and contributors | Difference-in-differences workflows in Python | |
| LPDiD | GitHub | P. J. Lambert | Dube, Girardi, Jordá, Taylor (2023) |