pysyncon

A python module for the synthetic control method that provides implementations of:

• Synthetic Control Method (Abadie & Gardeazabal 2003)
• Robust Synthetic Control Method (Amjad, Shah & Shen 2018)
• Augmented Synthetic Control Method (Ben-Michael, Feller & Rothstein 2021)
• Penalized Synthetic Control Method (Abadie & L'Hour 2021)

The package also provides methods for performing placebo tests and generating confidence intervals.

The implementation of the synthetic control method aims to be reconcilable with the R package Synth and similarly the implementation of the Augmented synthetic control method and the R package augsynth.

Installation

Install it from PyPI using pip:

python -m pip install pysyncon

Usage

Documentation is available on github-pages. In the examples folder are notebooks reproducing the weights from:

• The Economic Costs of Conflict: A Case Study of the Basque Country, Alberto Abadie and Javier Gardeazabal; The American Economic Review Vol. 93, No. 1 (Mar., 2003), pp. 113-132. (notebook here)
• The worked example 'Prison construction and Black male incarceration' from the last chapter of 'Causal Inference: The Mixtape' by Scott Cunningham. (notebook here)
• Comparative Politics and the Synthetic Control Method, Alberto Abadie, Alexis Diamond and Jens Hainmueller; American Journal of Political Science Vol. 59, No. 2 (April 2015), pp. 495-510. (notebook here)