Title: Non-parametric methods for doubly robust estimation of continuous treatment effects

Abstract: Continuous treatments (e.g., doses) arise often in practice, but many available causal effect estimators are limited by either requiring parametric models for the effect curve, or by not allowing doubly robust covariate adjustment. We develop a novel kernel smoothing approach that requires only mild smoothness assumptions on the effect curve and still allows for misspecification of either the treatment density or outcome regression. We derive asymptotic properties and give a procedure for data-driven bandwidth selection. The methods are illustrated via simulation and in a study of the effect of nurse staffing on hospital readmissions penalties.

This week, we will take around 10 minutes for pre-discussion before the main presentation. I list some of the questions so that people can be better prepared for this pre-discussion:

• What examples of continuous treatments are there?
• What are the key ideas of the proposed non-parametric methods for continuous treatments?
• What theoretical accomplishments/limitations the authors have?
• Are there any practical advantages/issues of the proposed non-parametric methods for continuous treatments?
• What are potential future research directions?

In addition, we will take two break times for a few minutes to discuss main proofs (probably just sketches) during the main presentation.

We encourage everyone (from junior graduate students to senior faculty members), who is interested in Causal Inference, to participate in this group in any ways.