***************************************************************************
Student #:
***************************************************************************
Question 1 (16 points)
(a) What assumptions are we making in the context of Time Series data
when using OLS regressions? (7)
(b) What is seasonality? Explain what it is and give a di↵erent example
than what was covered in class. This is where you can use a friend. (4)
(c) State all multipliers of a temporary 2-period shock on Xt in the
following model: Yt = 2 + 3Xt � 2.5Xt�2 + 1.5Xt�3 + “t.(5)
Question 2 (10 points)
You have two equations:
Bi = ↵1 + �1Pt + �1Inct + �1PLRt + “1
Bt = ↵2 + �2Pt + �2Xt + �2PLRt + “2
where B is the number of beagles (a breed of dog. Think of Snoopy), P
is their price, Inc is the average income of a family, PLR is the price of a
labrador retriever (another breed of dog) and X is the average size of a litter
of beagle puppies.
(a) What would you call this? Why? What does each equation repre-
sent? (3)
(b) Can you estimate this model? Why or why not? How? (6)
Question 3 (25 points)
You are modelling birth weight using the following model:
BWi = ↵+ �EDUCi + �ALCi + �HGTi + “i
where EDUC is education of the mother in years, ALC is the average
number of alcoholic drinks the mom drank per week during the pregnancy
and HGT is the height of the mom.
(a) Do you think that the ALC variable is reliable? Why or why not?
What is this problem called? (3)
(b) You have access to data on alcohol taxes. Assume that alcohol is a
normal good. How would you use it in this context? (6)
(c) What conditions must be met for what you describe in (b) to work?
(4)
(d) How can you test that the problem you describe in (a) is actually an
issue here? (5)
(e) You only have data on immigrant mothers (i.e. here defined as ones
that came to Canada less than 10 years ago). Do you think that this will
be a problem? Why or why not? How would you establish whether this is
an issue? (7)
Extra space for Question 3 – your test is not missing a page
5