Scribd
Upload
18 views1 page

Ec1 hw7

Uploaded by

bb747
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOC, PDF, TXT or read online on Scribd
18 views1 page

Ec1 hw7

Uploaded by

bb747
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOC, PDF, TXT or read online on Scribd
You are on page 1/ 1

Rauli Susmel

Econometrics 1
Homework 7

1. Unbalanced design for random effects. Suppose that the random effects model is to be estimated
with a panel in which the groups have different numbers of observations. Let T i be the number of
observations in group i.
(a) Show that the pooled least squares estimator is unbiased and consistent in spite of this
complication.
(b) Show that the GLS estimator based on the pooled least squares estimator of β (or, for that matter,
any consistent estimator of β) is a consistent estimator of σε2.

2. What are the probability limits of (1/n)LM, where LM is defined as in class


under the null hypotheses that σu2 = 0 and under the alternative that σu2 ≠ 0?

3. For this question, you will analyze a panel of airline industry cost data. The data are available from
Greene's data table F7.1 located at
http://pages.stern.nyu.edu/~wgreene/Text/tables/tablelist5.htm
Table F7.1: Cost Data for U.S. Airlines, 90 Observations On 6 Firms For 15 Years, 1970-1984.
Source: Data constructed by Christensen Associates of Madison, Wisconsin.

I = Airline,
T = Year,
Q = Output, in revenue passenger miles, index number,
C = Total cost, in $1000,
PF = Fuel price,
LF = Load factor, the average capacity utilization of the fleet.

For this exercise, you will estimate both a fixed effects model
ln(Cit) = αi + xit’β + εit
and a random effects model
ln(Cit) = α + xit’β + ui + εit

where xit= β1 ln(Qit) + β2 ln(PFit) + β3 LFit + β4 (LFit )2 + β5 T

Assume that εit is iid following a N(0;σε2), ui is also iid following a N(0;σu2), and Cov(ε;u) = 0.

3a. Given the loglinear model above, show the implied nonlinear relationship between C and the
independent variables for the FE model.

3b. Estimate a FE model with the LSDV approach using Gauss (Matlab or R are OK) and (also in
Gauss) test whether a fixed effect model or a pooled model (with one intercept) is appropriate.
Provide your Gauss code, your LSDV coefficient estimates and their associated t-statistics; your F-
statistic to test pooled versus FE, and its associated P-value, and explain your F-test result and
conclusion.

3c. From your (Gauss) LSDV and pooled regressions, calculate estimates σu2 and σε2 and calculate
Random Effects model parameter estimates. Report your code and the variance and
parameter estimates.

You might also like