Lecture+6+Regression+with+panel+data

Fortx ytxt01xtxt2wtxutxt or 02wtx1xtxtutxt

Info iconThis preview shows page 1. Sign up to view the full content.

View Full Document Right Arrow Icon
This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: = β0 + β1Xi1 + β2wi + ui1, i =1,…,n (1) Yi2 = β0 + β1Xi2 + β2wi + ui2, i =1,…,n (2) Yi3 = β0 + β1Xi3 + β2wi + ui3, i =1,…,n (3) One can difference the adjacent two time periods, i.e., (2) ‐ (1) and (3) ‐ (2), so Yi1 – Yi2 = β1(Xi1 ‐ Xi2) + ui1­ ui2, i =1,…,n Yi3 – Yi2 = β1(Xi3‐ Xi2) + ui3­ ui2, i =1,…,n One can estimate the above equations using OLS. So for t = T where T > 2, we have Yit – Yi(t­1)= β1(Xit ‐ Xi(t­1)) + uit­ ui(t­1), i =1,…,n; t=2,…,T 5 • We can estimate the regression by first differencing between adjacent time periods and then doing OLS. • Alternatively, can estimate the regression by including entity dummies (dummies for states in this...
View Full Document

Ask a homework question - tutors are online