Lahore School of Economics

BSc Econometrics II
Spring 2013
Instructor: Dr. Azam Chaudhry
Problem set No.3

Question 1:

Determine the identification properties of the following equations. In particular, be sure to note the
number of predetermined variables in the system, the number of slope coefficients in the equation and
whether the equation is under-identified, over-identified or exactly identified. (Assume that all equations
are stochastic unless specified otherwise.

Question 2:

In this question we want to analyze the problem of serial correlation in simultaneous models. Equation
14.29 shows the 2SLS version of the consumption function:

where C is consumption and YD is disposable income.

a. Test Equation 14.29 to confirm that we do indeed have a serial correlation problem.
b. Equation 14.29 will encounter both simultaneity bias and bias due to serial correlation with a
lagged endogenous variable. If you could solve only one of these two problems, which would you
choose? Why? (Compare Equation 14.29 with the OLS version of the consumption function
Equation14.30 (pp. 477, Studenmund.)
c. Suppose you wanted to solve both problems? Can you think of a way to adjust for both serial
correlation and simultaneity bias at the same time? Would it make more sense to run GLS first
and then 2SLS, or would you rather run 2SLS first and then GLS? Could they be run
simultaneously?
Question 3:

Suppose that a fad for oats (resulting from the announcement of the health benefits of oat bran) has made
you toy with the idea of becoming a broker in the oat market, Before spending your money, you decide to
build a simple model of supply and demand (identical to those in Section 14.1 and 14.2) of the market for
oats:

Where: QDt = the quantity of oats demanded in period t

QSt = the quantity of oats supplied in time period t
Pt = the price of oats in time period t
Wt = average oat-farmer wages in time period t
YDt = disposable income in time period t
a. You notice that no left-hand-side variable appears on the right side of either of your stochastic
simultaneous equations. Does this mean that OLS estimation will encounter no simultaneity bias?
Why or why not?
b. You expect that when Pt goes up, QDt will fall. Does this mean that if you encounter simultaneity
bias in the demand equation, it will be negative instead of the positive bias we typically associate
eith OLS estimation of simultaneous equations? Explain your answer.
c. Carefully outline how you would apply 2SLS to this system. How many equations (including
reduced forms) would you have to estimate? Specify precisely which variables would be in each
equation.
d. Given the following hypothetical data, estimate OLS and 2SLS versions of oat supply and
demand equations.
e. Compare your OLS and 2SLS estimates. How do they compare with your prior expectations?
Which equation do you prefer? Why?
Question 4:

Consider a simple model to estimate the effect of personal computer (PC) ownership on college grade
point average for graduating seniors at a large public university:

where PC is a binary variable indicating PC ownership.

a. Why might PC ownership be correlated with u?
b. Explain why PC is likely to be related to parents’ annual income. Does this mean parental income
is a good IV for PC? Why or why not?
c. Suppose that, four years ago, the university gave grants to buy computers to roughly one-half of
the incoming students, and the students who received grants were randomly chosen. Carefully
explain how you would use this information to construct an instrumental variable for PC.

Question 5:

The following is a simple model to measure the effect of a school choice program on standardized test
performance [see Rouse (1998) for motivation]:

where score is the score on a statewide test, choice is a binary variable indicating whether a student
attended a choice school in the last year, and faminc is family income. The IV for choice is grant, the
dollar amount granted to students to use for tuition at choice schools. The grant amount differed by family
income level, which is why we control for faminc in the equation.
a. Even with faminc in the equation, why might choice be correlated with u1?
b. If within each income class, the grant amounts were assigned randomly, is grant uncorrelated
with u1?
c. Write the reduced form equation for choice. What is needed for grant to be partially correlated
with choice?
d. Write the reduced form equation for score. Explain why this is useful. (Hint: How do you
interpret the coefficient on grant?)