slides_Ch2_W[1]

# 2b after some algebra we have e u jx 1 fa pb pa a

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

This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: n E [U ] = 0. Recall E [U ] = E [U jA] Pr (A) + E [U jB ] Pr (B ) = µ A fA + µ B ( 1 fA ) Conclude E [U ] = 0 i¤ µB = Melissa Tartari (Yale) µA Econometrics fA 1 fA (11) 73 / 93 Failure of SLR.2: An Example (Q.2.b) The question is: what does E [U jX = 1] = E [U jX = 0] imply for (µA , µB , pA , pB )? Before answering that question we need to take care of a little issue: we need to impose the normalization E [U ] = 0. Recall E [U ] = E [U jA] Pr (A) + E [U jB ] Pr (B ) = µ A fA + µ B ( 1 fA ) Conclude E [U ] = 0 i¤ µB = µA fA 1 fA (11) We use (11) to replace µB in expressions (10) and (9): in so doing we impose the normalization. Melissa Tartari (Yale) Econometrics 73 / 93 Failure of SLR.2: An Example (Q.2.b) After some algebra, we have E [U jX = 1] = Melissa Tartari (Yale) µ fA (pB pA ) µA fA (1 pB ) ; E [U jX = 0] = A Pr (X = 1) 1 Pr (X = 1) Econometrics 74 / 93 Failure of SLR.2: An Example (Q.2.b) After some algebra, we have E [U jX = 1] = µ fA (pB pA ) µA fA (1 pB ) ; E [U jX = 0] = A Pr (X = 1) 1 Pr (X = 1) We subtract E [U jX = 1] numerator 0 = (1 Pr (X = 1)) µA fA (1 = ... = µA fA [(1 Melissa Tartari (Yale) E [U jX = 0] and equate to zero the pB ) pB ) Pr (X = 1) (1 Econometrics Pr (X = 1) µA fA (pB pA ) pA )] 74 / 93 Failure of SLR.2: An Example (Q.2.b) After some algebra, we have E [U jX = 1] = µ fA (pB pA ) µA fA (1 pB ) ; E [U jX = 0] = A Pr (X = 1) 1 Pr (X = 1) We subtract E [U jX = 1] numerator 0 = (1 E [U jX = 0] and equate to zero the Pr (X = 1)) µA fA (1 = ... = µA fA [(1 pB ) pB ) Pr (X = 1) (1 Pr (X = 1) µA fA (pB pA ) pA )] Thus, for E [U jX ] = 0 it is su¢ cient that one of the following conditions holds (it is necessary that at least one holds ): Melissa Tartari (Yale) Econometrics 74 / 93 Failure of SLR.2: An Example (Q.2.b) After some algebra, we have E [U jX = 1] = µ fA (pB pA ) µA fA (1 pB ) ; E [U jX = 0] = A Pr (X = 1) 1 Pr (X = 1) We subtract E [U jX = 1] numerator 0 = (1 E [U jX = 0] and equate to zero the Pr (X = 1)) µA fA (1 = ... = µA fA [(1 pB ) pB ) Pr (X = 1) (1 Pr (X = 1) µA fA (pB pA ) pA )] Thus, for E [U jX ] = 0 it is su¢ cient that one...
View Full Document

## This note was uploaded on 02/13/2014 for the course ECON 350 taught by Professor Donaldbrown during the Fall '10 term at Yale.

Ask a homework question - tutors are online