Advantages and limitations in the use of transfer learning for individual treatment effects in causal machine learning

TL;DR

This paper explores how transfer learning can enhance the estimation of individual treatment effects in causal machine learning, especially in small sample scenarios, by leveraging knowledge from larger source datasets.

Contribution

It demonstrates the effectiveness of transfer learning in improving ITE estimation accuracy and reducing bias in small or different target environments using TARNet-based models.

Findings

Transfer learning reduces ITE error in simulations.

Transfer learning decreases bias in empirical application.

TL-TARNet outperforms standard TARNet in small sample settings.

Abstract

Generalizing causal knowledge across diverse environments is challenging, especially when estimates from large-scale datasets must be applied to smaller or systematically different contexts, where external validity is critical. Model-based estimators of individual treatment effects (ITE) from machine learning require large sample sizes, limiting their applicability in domains such as behavioral sciences with smaller datasets. We demonstrate how estimation of ITEs with Treatment Agnostic Representation Networks (TARNet; Shalit et al., 2017) can be improved by leveraging knowledge from source datasets and adapting it to new settings via transfer learning (TL-TARNet; Aloui et al., 2023). In simulations that vary source and sample sizes and consider both randomized and non-randomized intervention target settings, the transfer-learning extension TL-TARNet improves upon standard TARNet, reducing ITE error and attenuating bias when a large unbiased source is available and target samples are small. In an empirical application using the India Human Development Survey (IHDS-II), we estimate the effect of mothers' firewood collection time on children's weekly study time; transfer learning pulls the target mean ITEs toward the source ITE estimate, reducing bias in the estimates obtained without transfer. These results suggest that transfer learning for causal models can improve the estimation of ITE in small samples.