disco: Distributional Synthetic Controls

TL;DR

The paper introduces 'disco', a new method and software for constructing synthetic distributions to evaluate policy impacts at a distributional level, enhancing traditional synthetic control techniques.

Contribution

It presents the 'disco' command implementing Distributional Synthetic Controls, enabling distributional impact analysis with new aggregation, inference, and visualization tools.

Findings

Successfully replicates empirical results from prior studies

Provides a flexible framework for distributional causal inference

Enhances interpretability of policy effects at different distribution points

Abstract

The method of synthetic controls is widely used for evaluating causal effects of policy changes in settings with observational data. Often, researchers aim to estimate the causal impact of policy interventions on a treated unit at an aggregate level while also possessing data at a finer granularity. In this article, we introduce the new disco command, which implements the Distributional Synthetic Controls method introduced in Gunsilius (2023). This command allows researchers to construct entire synthetic distributions for the treated unit based on an optimally weighted average of the distributions of the control units. Several aggregation schemes are provided to facilitate clear reporting of the distributional effects of the treatment. The package offers both quantile-based and CDF-based approaches, comprehensive inference procedures via bootstrap and permutation methods, and visualization capabilities. We empirically illustrate the use of the package by replicating the results in Van Dijcke et al. (2024).