Question

For the following instrumental variables (IV) regression model Yt = ?T xt + ut, xt = ?T wt + vt, (4) where E(ut|xt) ? 0 and E(ut|wt) = 0. Here, it is assumed that wt is l × 1 and xt is k × 1 with l ? k for the identification. If l > k, model (4) is over-identified. (a) Please derive the IV estimator for ?, denoted by ??IV. (c) Show all steps on tests of over-identifying restrictions and the DWH test for model (4). Durbin-Wu- Haussman test

          For the following instrumental variables (IV) regression model
Yt = ?T xt + ut,


xt = ?T wt + vt,
(4)
where E(ut|xt) ? 0 and E(ut|wt) = 0. Here, it is assumed that wt is l × 1 and xt
is k × 1 with l ? k for the identification. If l > k, model (4) is over-identified.
(a) Please derive the IV estimator for ?, denoted by ??IV.
(c) Show all steps on tests of over-identifying restrictions and the DWH
test for model (4).
Durbin-Wu-
Haussman
test

For the following instrumental variables (IV) regression model
Yt = ?T xt + ut,

xt = ?T wt + vt,
(4)
where E(ut|xt) ? 0 and E(ut|wt) = 0. Here, it is assumed that wt is l × 1 and xt
is k × 1 with l ? k for the identification. If l > k, model (4) is over-identified.
(a) Please derive the IV estimator for ?, denoted by ??IV.
(c) Show all steps on tests of over-identifying restrictions and the DWH
test for model (4).
Durbin-Wu-
Haussman
test

Added by Jessica B.

Elementary Statistics a Step by Step Approach

Allan G. Bluman 9th Edition

Thanks for your feedback!

For the following instrumental variables (IV) regression model: Yt = 3T xt + ut Xt = TTWt + Vt (4) where E(ut|xt) = 0 and E(ut|wt) = 0. Here, it is assumed that wt is l x 1 and xt is k x 1 with l > k for identification. If l > k, model (4) is over-identified. (a) Please derive the IV estimator for Î², denoted by Î²Ì‚. (b) Show all steps on tests of over-identifying restrictions and the DWH Hausman test for model (4).