Estimation Methods with Instruments

This section contains methods for estimating (heterogeneous) treatment effects, even when there are unobserved confounders (factors that simultaneously had a direct effect on the treatment decision in the collected data and the observed outcome). However, the methods assumes that we have access to an instrumental variable: a random variable that had an effect on the treatment, but did not have any direct effect on the outcome, other than through the treatment.

The package offers two IV methods for estimating heterogeneous treatment effects: deep instrumental variables [Hartford2017] and the two-stage basis expansion approach of [Newey2003].

• Deep Instrumental Variables
• Sieve 2SLS Instrumental Variable Estimation
• Orthogonal instrumental variables
  • What is it?
  • What are the relevant estimator classes?
  • When should you use it?
  • Class Hierarchy Structure
  • Usage Examples