Name

Midterm Examination: Economics 210A

November 6, 2013
Answer Question 1 and any 3 of the remaining 4 questions. Good luck.

1 ) A consumer has utility function

1
u(x1 , x2 ) = 1 4 .
x1 + x2

A) Find this consumer’s Marshallian demand function for each good.

Answer:
m
x1 (p1 , p2 , m) = 1/2 1/2
p1 + 2p1 p2
2m
x2 (p1 , p2 , m) = 1/2 1/2
2p2 + p1 p2
B) Find this consumer’s indirect utility function.
Answer: You can get your answer by substituting the Marshallian demands
that you found into the definition of indirect utility which is

v(p, m) = u(x1 (p, m), x2 (p, m)).

m
v(p1 , p2 , m) = 1/2 1/2
p1 + 4p1 p2 + 4p2
Another way of writing the same answer is
m
v(p1 , p2 , m) =  2
1/2 1/2
p1 + 2p2

C) Find this consumer’s expenditure function.

Since v(p, e(p, u)) = u, it follows that

e(p1 , p2 , u)
1/2 1/2
=u
p1 + 4p1 p2 + 4p2
and therefore
 
1/2 1/2
e(p1 , p2 , u) = u p1 + 4p1 p2 + 4p2 .

A side remark: To check your answer (and see how these structures are
related), you could notice that the denominator of the indirect utility function is

1
a CES function of the p’s. You might recall from our classroom discussion that
if a utility function is of the form

(c1 xa1 + c2 xa2 )1/a

then the corresponding expenditure function is of the form

(c01 pb1 + c02 pb2 )1/b u.

When this is the case, the elasticity of substitution of the direct utility function
is 1/(1 − a) and the elasticity of substitution of the indirect utility function,
which is 1/(1 − b) must be inverses of each other. That is, 1/(1 − b) = 1 − a.
Solving this equation for b, we have b = −a/(1 − a). In our example, a = −1,
so b = 1/2. Thus the expenditure function is of the form
1/2 1/2
e(p1 , p2 , u) = u(c01 p1 + c02 p2 )2 u.
1/2
We also showed that c0i = cσi = ci , So we would have
 2  
1/2 1/2 1/2 1/2
e(p1 , p2 , u) = p1 + 2p2 u = p1 + 4p1 p2 + 4p2 u.

D) Find this consumer’s Hicksian demand function for each good.

The Hicksian demands are equal to the corresponding partial derivatives of
the expenditure function.
We find
−1/2 1/2
h1 (p1 , p2 , u) = u(1 + 2p1 p2 )
and
1/2 −1/2
h2 (p1 , p2 , u) = u(4 + 2p1 p2 ).

E) Find the income and substitution effects of a change in the price of good 1
on the demand for good 1.
The substitution effect is the partial derivative of h1 (p1 , p2 , m) with respect
to p1 . That is
1/2 −3/2
−up2 p1 .
We may want to express this substitution effect entirely as a function of prices
and income. Thus we would make use of the fact that u = v(p1 , p2 , m). Then
we have the substitution effect is
1/2 −3/2
−mp2 p1
1/2 1/2
p1 + 4p1 p2 + 4p2
The income effect is
∂x(p1 , p2 , m) m
−x1 (p1 , p2 ) = − 2
∂m 1/2 1/2
p1 + 2p1 p2

2
2) A consumer has utility function
1 √
u(x1 , x2 , x3 ) = 1 4 + 2 x3 .
x1 + x2

A) If this consumer has income m and the prices of the three goods are p1 , p2 ,
and p3 , how much of her income will she spend on good 1. (Hint: make use of
part B of your answer to the previous question.)
Answer: If the consumer spends a total of s on goods 3 then she will get s/p3
units of good 3 and will have m − s left to spend on goods 1 and 2. Her utility
will then be
p p m−s
2 s/p3 + v(p1 , p2 , m − s) = 2 s/p3 +  2 .
1/2 1/2
p1 + 2p2

She will choose s to maximize this. At an interior maximum, the derivative

of this expression with respect to s must be zero. Then
 2
1/2 1/2
(sp3 )1/2 − p1 + 2p2 =0

and so  4
1/2 1/2
p1 + 2p2
s= . (1)
p3
and total expenditure on goods 1 and 2 is
 4
1/2 1/2
p1 + 2p2
m−s=m− (2)
p3
B) Write down the consumer’s Marshallian demand function for each good.
(Hint: make use of your answer to parts A of this question and the previous
question.)
Answer: The consumer will choose to allocate the expenditure m − s between
goods 1 and 2 in such a way as to maximize utility subject to p1 x2 +p2 x2 = m−s.
So we can just apply the Marshallian demand functions from Problem 1 with
income m − s These are:

  4 
1/2 1/2 !
m−s p1 + 2p2 1
x1 (p1 , p2 , m) = = m −
 
1/2 1/2 p3
 1/2 1/2
p1 + 2p1 p2 p1 + 2p1 p2

  4 
1/2 1/2 !
2(m − s) p1 + 2p2 2
x2 (p1 , p2 , m) = = m −
 
1/2 1/2 p3
 1/2 1/2
2p2 + p1 p2 2p2 + p1 p2

3
 An excellent answer would also note that there will be an interior solution if
and only if   4 
1/2 1/2
p 1 + 2p2
m −  > 0.
 
p3

If this inequality is reversed, there will be a corner solution in which the con-
sumer consumes only good 3.

3 ) Consider the function

f (x1 , x2 , x3 ) = ln x21 + x22 + x23 .

Explain your answers do each of the following questions:

A) Is the function f homogeneous? If so, of what degree?

Answer: No. f (tx) = ln t2 f (x) = 2 ln t + f (x). A function is homogeneous

if and only if f (tx) = tk f (x) for all t and some k. There is no value of k such
that 2 ln t + f (x) = tk f (x).
B) Is the function f homothetic?

Answer: Yes. f (x) is a strictly increasing function of the homogeneous

function x21 + x22 + x23 .
C) Find the gradient vector of f at the point (x1 , x2 , x3 ) = (1, 3, 5).

The gradient is (2/35, 6/35, 10/35).

D) At the point (1, 3, 5), find the directional derivative of f in the direction
(2/9, −2/9, 1/9).

Answer: For the directional derivative in the direction of a vector x, we need

to takepthe inner product of the gradient with the normalized vector x/|x| where
|x| = x21 + x22 + x23 . In our case, x/|x| = (2/3, −2/3, 1/3)
The directional derivative is
1 2
(2 × 2 − 6 × 2 + 10 × 1) = .
3 × 35 105

2 ) Ada and Bo both have have convex preferences defined on a set X, but
their preferences are not identical.
A) Let x be a commodity bundle in X. Let A be the set of bundles that both
Ada and Bob like better than x. Is A necessarily a convex set? If so, prove it.
If not, show a counterexample.
Answer: If Ada and Bob both have convex preferences, then the sets A and
B are both context sets. The set of bundles that both like better than x is the
set A ∩ B. The intersection of two convex sets is a convex set. (A good answer
would also prove this last claim.)

4
B) Let x be a commodity bundle in X. Let C be the set of bundles such that if
z ∈ C, then z = yA + yB for some yA ∈ X and yB ∈ X, z = yA + yB such that
Ada prefers yA to x and Bo prefers yB to x. Is C necessarily a convex set. If
so, prove it. If not, show a counterexample.
Answer: This is true. Let z and z 0 be two points in C. Then z = yA + yB
for some yA ∈ A and some yB ∈ B and z 0 = yA 0
+ yB0 0
for some yA ∈ A and
some yB ∈ B. Now for any λ between 0 and 1, and any two points z and z 0 in
0

C, let z(λ) = λz + (1 − λ)z 0 . Then

0 0
z(λ) = (λyA + (1 − λ)yA ) + (λyB + (1 − λyB )) .
0
Since A and B are convex sets, it must be that λyA + (1 − λ)yA ∈ A and
0
λyB + (1 − λyB ) ∈ B. Since
0 0
z(λ) = (λyA + (1 − λ)yA ) + (λyB + (1 − λyB )) ,

it must be that z(λ) ∈ C and hence C is a convex set.

5) Will’s preferences are represented by the utility function

1/2 1/2
u(x1 , x2 ) = min{x1 x2 , x2 }.

A) Draw a couple of indifference curves for Will.

Answer: Let good 1 be on the horizontal axis and good 2 on the vertical
axis. To draw a typical indifference curve for Will, take a point (x, x) on the
diagonal. The indifference curve through this point includes a straight horizontal
line extending to the right of true point (x, x) and above and to the left of (x, x)
is the locus of points (x1 , x2 ) satisfying the equation x1 x2 = x.
B) Find Will’s Marshallian demand functions for goods 1 and 2.
Answer:
If p1 ≤ p2 , x1 (p1 , p2 ) = x2 (p1 , p2 ) = m/(p1 + p2 ).
If p1 > p2 , then x1 = m/(2p1 ) and x2 = m/(2p2 ).