FWL Theorem: Meaning, Concept, and STATA practical [Video embedded]

In STATA

clear all

set obs 1000

set seed 12345

// Generating variables

gen y = rnormal()*10+10

gen x = y+runiform()*5

gen z = y*2+runiform()*y

// Main regression

reg y x z

// Regression 1

reg y z

predict resid1,r //Residuals 1

// Regression 2

reg x z

predict resid2,r //Residuals 2

reg resid1 resid2 // The coefficient must be same as Main regression for "x".

reg y resid2 // The coefficient must be same as Main regression for "x".

---