{[ promptMessage ]}

Bookmark it

{[ promptMessage ]}

# hws07 - Economics 241B Answers to Exercise 7 1(Final...

This preview shows pages 1–4. Sign up to view the full content.

Economics 241B Answers to Exercise 7 1. (Final Examination, Economics 241B, Winter 2004) a. For the classic linear regression model, identi°cation of ° 1 follows from mulitiplying both sides of the regression by X t : X t Y t = °X 2 t + X t U t : Taking the expectation of both sides identi°es ° if E ( X t U t ) = 0 (because we have sample analogs for E ( X t Y t ) and E ( X 2 t ) . With measurement error, the resultant equation is X t Y t = °X 2 t + X t ( U t ° °V t ) : Because E ( X t V t ) 6 = 0 , ° = E ( X t Y t ) E ( X 2 t ) ° E ( X t V t ) . Because there is no sample analog for E ( X t V t ) , ° is not identi°ed. b. For Z t to be a valid instrument, it must be the case that n ° 1 n X t =1 Z t Y t P ! °Q ZX ; where Q ZX = lim n !1 n ° 1 P n t =1 Z t X t . To check n ° 1 n X t =1 Z t Y t = n ° 1 n X t =1 Z t ( °X t + U t ° °V t ) : Because Z t is a function of X t and X t is independent of U t : n ° 1 n X t =1 Z t U t P ! 0 : Next n ° 1 n X t =1 Z t V t = n ° 1 n X t =1 V t 1 ( X t ± med ( X )) ° n ° 1 n X t =1 V t 1 ( X t < med ( X )) ; which converges in probability to E ( V j X ± m ) P ( X ± m ) ° E ( V j X < m ) P ( X < m ) ;

This preview has intentionally blurred sections. Sign up to view the full version.

View Full Document
where m is the population median of X . Clearly P ( X ± m ) = P ( X < m ) = 1 2 . Thus Z is a valid instrument if E ( V j X ± m ) = E ( V j X < m ) : Such a result would hold if V and X were independent, but they are clearly dependent. Because V and X are dependent, the two conditional expecta- tions are not generally equal, and Z is not a valid instrument. In fact, if X is also symmetrically distributed around its mean (then mean=median) and E ( V j X ± m ) = ° E ( V j X < m ) , so that n ° 1 P n t =1 Z t V t P ! E ( V j X ± m ) 6 = 0 . c. The assumption that V t is symmetrically distributed yields additional mo- ment information. To exploit this, we set up the moment conditions. Because the model is in deviation-from-means form, the °rst moments are zero. The second moments yield EX 2 t = E ( X ± t ) 2 + ± 2 V EY 2 t = ° 2 E ( X ± t ) 2 + ± 2 U E ( X t Y t ) = °E ( X ± t ) 2 ; which is a system of 3 equations in 4 unknowns. For this system ° is not identi°ed. To estimate the parameters, we need further moment conditions. Because the distributions of U t and V t are symmetric, the third moments are zero, which yields EX 3 t = E ( X ± t ) 3 EY 3 t = ° 3 E ( X ± t ) 3 : We now have a system of 5 equations in 5 unknowns, so ° is identi°ed. We solve via the method of moments. In particular B MM = ° P n t =1 Y 3 t P n t =1 X 3 t ± 1 3 : 2. (Fall 1990 Final Examination) a) To correct for heteroskedasticity we transform the model to obtain an error that is homoskedastic. In the case at hand, we know V ( U t ) = ± 2 Z 2 t with Z t known, so the transformation is simple. From the de°nition of the variance, V ² U t Z t ³ = ± 2 , and so U t Z t is a homoskedastic error. The transformation is Y t Z t = ² + ° X t Z t + U t Z t :
Why the presence of an intercept? Because P n t =1 U t = 0 does not imply P n t =1 U t Z t = 0 . To remove the intercept, we would need to remove the mean of Y t Z t and X t Z t .

This preview has intentionally blurred sections. Sign up to view the full version.

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

{[ snackBarMessage ]}

### What students are saying

• As a current student on this bumpy collegiate pathway, I stumbled upon Course Hero, where I can find study resources for nearly all my courses, get online help from tutors 24/7, and even share my old projects, papers, and lecture notes with other students.

Kiran Temple University Fox School of Business ‘17, Course Hero Intern

• I cannot even describe how much Course Hero helped me this summer. It’s truly become something I can always rely on and help me. In the end, I was not only able to survive summer classes, but I was able to thrive thanks to Course Hero.

Dana University of Pennsylvania ‘17, Course Hero Intern

• The ability to access any university’s resources through Course Hero proved invaluable in my case. I was behind on Tulane coursework and actually used UCLA’s materials to help me move forward and get everything together on time.

Jill Tulane University ‘16, Course Hero Intern