Lead Investigator: Dr. Melvyn Weeks, Faculty of Economics, University of Cambridge, UK
Coauthor: Eoghan O’Neil, Faculty of Economics, University of Cambridge, UK
The investigation of treatment effect heterogeneity is of interest in many fields including economics and personalised medicine. In some applications, there are many available covariates and functions of covariates that are plausibly associated with treatment effect heterogeneity. In this paper we apply and compare a number of recently developed methods for describing the heterogeneity of treatment effects. We utilise these methods to investigate how household demand response to Time-of-Use (TOU) tariffs varies with aspects of past consumption data and other household demographics.
Heterogeneous treatment effects will be described by estimating Conditional Average Treatment Effects (CATE), which are the expected differences between treated and control households for subsets of the population defined by covariates. First, standard CATE estimates are produced for covariates believed a priori to be informative. Secondly, the method of causal trees is applied to search across many potential conditioning variables for aspects of heterogeneity that are diffcult to hypothesize a priori. Thirdly, we obtain individual specific estimates from a causal forest.
This application illustrates a number of advantages and disadvantages common to the application of machine learning methods in econometrics, These issues include: the choice of interpretable conditioning variables; the choice and interpretation of variable importance measures; and how to present individual treatment effect estimates. Issues related to the chosen methods include the trade-off between the interpretability of a single causal tree and the stability of estimates produced by a causal forest.