Babson College ECN6310 – Modern Econometrics for Business Prof.
Simon
In-class practice questions for Midterm Exam
1. The table below shows the joint distribution between the change of the unemployment rate in an election
year and the share of the candidate of the incumbent party (the political party of the President in a given
election year) since 1928. You think of this data as a population that you want to describe, rather than a
sample.
Joint Distribution of Unemployment Rate Change and Incumbent Party's Vote
Share in Total Vote Cast for the Two Major-Party Candidates,
1928-2000
(Incumbent- 50%) > 0 (Incumbent- 50%) ≤ 0 Total
△u > 0 (B = 0)
(A = 0) (A = 1)
△u ≤ 0 (B = 1)
0.053 0.211 0.264
0.579 0.157 0.736
Total 0.632 0.368 1.00
(a) Compute and interpret E(A) and E(B).
(b) Calculate E(A|B = 1). Interpret your results in words.
(c) What is the probability that the unemployment rate increases in an election year?
(d) Imagine you want to use the dataset you have created to run a regression to understand the relationship
between unemployment and incumbent party vote share. Write out the regression equation you would estimate
it and explain it in words. Would you expect the intercept of your regression to be interpretable? Explain why or
why not.
Babson College ECN6310 – Modern Econometrics for Business Prof. Simon
In-class practice questions for Midterm Exam
2. You have collected weekly earnings and age data from a sub-sample of 1,744 individuals using the Current
Population Survey in a given year. You find that the mean weekly earnings for your subsample are $434.49 with a
standard error of $7.06. Next, you run a regression. The results are:
= 239.16 + 5.20 × Age, R2 = 0.05
(10.10) (2.78)
where Earn and Age are measured in dollars and years respectively.
(a) Is the coefficient on Age statistically significant? Explain briefly. (5 points)
(b) Someone in your sample is 38 years old. What are predicted weekly earnings? (2 points)
(c) Why should age matter in the determination of earnings? What do the regression results suggest about
increases in earnings and age? Explain whether the results make sense. (8 points)
(d) Thinking about the first econ lab, explain one other variable you would want to add next. Why is the
variable important, and what would it tell you? (5 points)
Babson College ECN6310 – Modern Econometrics for Business Prof. Simon
In-class practice questions for Midterm Exam
3. You want to test if there is international price discrimination for pharmaceuticals. For there to be
international price discrimination, you would see that the coefficient on income (measured by GDP per
capital, GDPN) would be positive and significant.
To predict pharmaceutical prices in 32 countries, you use variables similar to a published paper. You
estimate a regression using the following data:
• Pi = pharmaceutical price level in the ith country divided by that of the US
• GDPNi = per capita gross domestic product in the ith country divided by that of US
• CVNi = per capita volume of consumption of pharmaceuticals in the ith country divided by that of US
• PPi = dummy variable equal to 1 if patents for pharma products are recognized in ith country; 0
otherwise
• DPCi = dummy variable equal to 1 if ith country applied strict price controls; 0 otherwise
• Pci = dummy variable equal to 1 if ith country encouraged price competition, 0 otherwise
Your equation is:
^
Pi= 38.22+1.43GDPNi -0.6CVNi +7.31PPi-15.63DPCi -11.38PCi
(0.21) (0.22) (6.12) (6.93) (7.16)
t= 6.69 -2.66 1.19 -2.25 -1.59
N = 32 R 2 = 0.775
a. Which variables are statistically significant? How do you know? (5 points)
b. You want to answer the question, Does the pharmaceutical industry engage in international price
discrimination? Does the regression suggest that international price discrimination is happening? Why? (5
points)
c. What do the VIF statistics tell you about the regression? What do you think you should do about it, given
that you chose the variables based on the published paper? Why? (10 points)
