Inverse probability weighted estimation of local average treatment effects: A higher order MSE expansion

• Propose a modified inverse probability weighted estimator for LATE.
• Use local polynomial regression to estimate the instrument propensity score.
• The modified estimator remains asymptotically normal and efficient.
• Provide a higher order expansion of its asymptotic mean squared error.

We consider a new inverse probability weighted estimator for the local average treatment effect parameter where the instrument propensity score is estimated by local polynomial regression. We derive its asymptotics and provide a higher order expansion of its asymptotic MSE.