IV Wald Estimate

Problem

Compute the instrumental variable (Wald) estimate of a treatment effect: cov(z, y) / cov(z, x), where z is the instrument, x the endogenous regressor, and y the outcome.

Return 0.0 if cov(z, x) is 0 or input is empty.

Hint

The instrument must affect the outcome only through the endogenous variable for this estimator to be valid.

Test cases

case 1

in: [[1,2,3,4],[2,4,6,8],[1,3,5,7]]

expected: 1

case 2

in: [[1,2,3,4],[0,0,0,0],[1,2,3,4]]

expected: 0

case 3

in: [[1,2,3],[2,3,4],[5,6,7]]

expected: 1

case 4

in: [[],[],[]]

expected: 0

Starter codePython

def iv_wald_estimate(z, x, y):
    # Your code here
    pass

4 test cases · run your code against them in the editor