leebounds_unknown_sign: Compute Lee (2009) bounds with unknown monotonicity direction

In vsemenova/leebounds: Nonparametric bounds on Average Treatment Effect in the presence of Endogenous Selection

Compute Lee (2009) bounds with unknown monotonicity direction

Description

This function computes basic Lee (2009) upper and lower bound on the Average Treatment Effect . Its input argument is a dataframe consisting of d=treat (binary treatment), s = selection (e.g., employment, test participation), and outcome = sy observed only if s=1 (e.g., wage, test score). Lee (2009) bounds make two assumptions: (1) Treatment is randomly assigned and (2) Treatment either cannot hurt or cannot help selection, but the direction is the same for all individuals: S_1 >= S_0 a.s. or S_0>=S_1, where S_1,S_0 are potential selection outcomes.

Usage

leebounds_unknown_sign(leedata)

Arguments

leedata

data frame with treat, selection, outcome

Value

Lee (2009) lower and upper bound

vsemenova/leebounds documentation built on Sept. 30, 2023, 8:30 a.m.