How do applied researchers use the Causal Forest? A methodological review of a method

TL;DR

This review analyzes how applied researchers utilize the causal forest method, highlighting common practices, variations, and areas for future development based on 133 peer-reviewed studies.

Contribution

It provides a comprehensive overview of current best practices and deviations in applying causal forests, guiding future methodological improvements.

Findings

Researchers predominantly use causal forests with low-dimensional data.

Common visualization methods include distribution and variable importance plots.

Deviations from standard practice may be beneficial or harmful.

Abstract

This methodological review examines the use of the causal forest method by applied researchers across 133 peer-reviewed papers. It shows that the emerging best practice relies heavily on the approach and tools created by the original authors of the causal forest such as their grf package and the approaches given by them in examples. Generally researchers use the causal forest on a relatively low-dimensional dataset relying on observed controls or in some cases experiments to identify effects. There are several common ways to then communicate results -- by mapping out the univariate distribution of individual-level treatment effect estimates, displaying variable importance results for the forest and graphing the distribution of treatment effects across covariates that are important either for theoretical reasons or because they have high variable importance. Some deviations from this common practice are interesting and deserve further development and use. Others are unnecessary or even harmful. The paper concludes by reflecting on the emerging best practice for causal forest use and paths for future research.