Nonstationary Time Series
Often time series data are not stationary. This
chapter covers nonstationary time series, and the
implications for estimators and for testing.
Please print out and attach your do file.
1.(6) For your airport, regress ln(passengers) on trend and a dummy for 911 defined as
(ts911/(1+ts911) where ts911=time since 911 for quarters since 911 (trend=36-end of data, and =
1.(10) Suppose you are examining the supply and demand of chicken in the Eugene area. The supply of
chicken depends only on the price of chicken. The demand for chicken depends on the price of chicken
in addition to the pric
EC 421, Fall 2013
HW4 : Suggested Solutions
An important note: This solution was don as quickly as I could to get something up
tonight. A complete solution included the work showing how each case eects the estimators
from our OLS formulas for b1 and b2 .
1.(7) In blackboard, there is a data file that contains costs for railroad firms from 1983 through 2003.
One firm, Burlington Northern(rr=BN) had a major merger in 1995. The data for 1983-1995 are p
1.(8) Take your airport data, merge with the quarterly income and population file and cpi. For quarter 4
of 2011, run a regression of passengers per captia income. Save the residuals, and first plot
The strawperson principle is the idea that you should always make your null hypothesis the idea you
wish to disprove. That is to say, always seek to prove your argument by using the data to disprove the
1.(20) True/False circle the correct answer
. g T
ZSLS is an instrumental variable approach that corrects for heteroscedasticity.
A random walk with drift is
OPTIONAL: This homework can be completed and used to replace a homework score already received
(including a missing homework score). But, if you turn in, it will replace a pre-specified homework score
regardless of the score. It is due during t