STAT-S-301 Introduction to Econometrics

PROBLEM SET 3

TA: Elise Petit

December 2016

Instructions
Students are expected to work with the same group as for the previous problem
sets, using the same group name as is published on STATS301_PS2_grades.pdf.
Each group is expected to send their solutions to elipetit@ulb.ac.be (subject:
PS3 STATS301) with two attachments: your group report in pdf format with
the title PS3_group name and your STATA do-le with the same name.

Deadline: Thursday 29/12 at 10 am

Updated oce hours schedule: 15/12 from 11am to 2pm
Q&A session: 10/01 at 6pm (questions on exercises and problem sets,
to be submitted by email by 10pm the day before)
Cartel in rail transport
During the 1880s, a cartel known as the Joint Executive Committee (JEC)
controlled the rail transport of grain from the Midwest to eastern cities in the
United States.

The cartel preceded the Sherman Antitrust Act of 1890, and

it legally operated to increase the price of grain above what would have been
the competitive price. From time to time, cheating by members of the cartel
brought about a temporary collapse of the collusive pricesetting agreement.
Let's use variations in supply associated with the cartel's collapses to estimate
the elasticity of demand for rail transport of grain.

Data description
Dataset_PS3_Cartel contains weekly observations on prices and other factors
from 1880-1886, for a total of 326 weeks. These data were provided by Professor
Rob Porter of Northwestern University and were used in his paper A Study of
Cartel Stability: The Joint Executive Committee, 1880-1886 The Bell Journal
of Economics, Vol. 14, No. 2, Autumn 1983, 301-314.

Variable

Description

week

indicator for week of observation

price

weekly index of price of shipping a ton of grain by rail

indicator =1 if Great Lakes are impassable because of ice

ice
cartel

indicator =1 if railroad cartel is operative

quantity

total tonnage of grain shipped in the week

seas1

indicator = 1 if date is January 1 through January 28

seas2

indicator =1 if date is January 29 through February 25

.
.
.

.
.
.

indicator =1 if date is December 4 through December 31

seas13

Note: the thirteen month binary variables. To match the weekly data, the
calendar has been divided into 13 periods, each approximately 4 weeks long.

Questions
Suppose that the demand curve for rail transport of grain is specied as

ln(Qi ) = 0 + 1 ln(Pi ) + 2 Icei +

12
X

2+m Seasm,i + ui

m=1
where

Qi

is the total tonnage of grain shipped in week i,

shipping a ton of grain by rail,

Icei

Pi

is the price of

is a binary variable that is equal to 1 if the

Seasm,i is a binary variable

Ice is included because grain could

Great Lakes are not navigable because of ice, and

that captures seasonal variation in demand.

also be transported by ship when the Great Lakes were navigable

1
2
3

Estimate the equation of the demand for grain by OLS. Interpret

Explain intuitively if you believe that the variable
strument for

ln(P ).

cartel

c1 .

could be a valid in-

Why did we consider using an instrumental variable?

Estimate the demand equation using instrumental variable regression.

cartel

4 BONUS

Is

a weak instrument ?
Does the evidence suggest that the cartel was charging the prot-

maximizing monopoly price ? Explain. (Hint : What should the monopolist do if the

c1 < 1

?)

Now suppose that the demand curve for rail transport of grain can also be
modelled as an AR(1) process

ln(Qi ) = 0 + 1 ln(Qi1 ) + i

Plot the weekly evolution of

ln(Qi ) from 1880 to 1886.

Comment on whether

or not the series seems stationnary and conrm your intuition using a
statistical test.

6
7
8
9

Estimate the AR(1) process.

How do you interpret

b1

and its associated

p-value?
Plot the weekly evolution of

di ).
ln(Qi ) against its in-the-sample prediction ln(Q

Does the AR(1) model seem reliable?

Comment using the graph and a

regression quality indicator.

Repeat questions 6&7 using an AR(2) modelisation of

ln(Qi ).

Which model

do you prefer?
Based on the information you have from 1880 week 1 to 1886 week 16, what
is your forecast for the total tonnage of grain shipped the 17th week of
1886?