Adv.

Micro Theory, ECON 6202-090

Assignment 2 Answers, Fall 2010
Due: Monday, September 27th
Directions: Answer each question as completely as possible. You may work in a group consisting of up
to 3 members for each group please turn in only 1 set of answers and make sure all group member names
are on that set of answers. All group members will receive the same grade.
p
1. Consider the following utility functions: u(x1 ; x2 ) = x1 x2 and v(x1 ; x2 ) = ln(x1 ) + ln(x2 ): Verify
that u and v have the same indierence curves and the same marginal rate of substitution. Explain
why.
Answer:
The marginal rate of substitution for u (x1 ; x2 ) is:
M Ux1

M Ux2

M RS

M RS

1 12 21
x x
2 1 2
1 12 12
x x
2 1 2
1
2 x1

1
2

x22

1 2
2 x1 x2
x2

1
2

x1

When we have v (x1 ; x2 ) the MRS is:

1
x1
1
x2

M Ux1

M Ux2

M RS

1
x1
1
x2

M RS

x2
x1

Note that since the marginal rates of substitution are the same the indierence curves will be the
same. Also note that the utility values will be dierent for the same bundles it is NOT required
that the same bundles have the same utility values. What is required is that all the bundles that
have u (x1 ; x2 ) = z have v (x1 ; x2 ) = s where s typically will not be equal to z. Think of it this
way if I plot the indierence curve that runs through the points (1; 4), (2; 2), (4; 1) basically, the
1=2 1=2
indierence curve for u (x1 ; x2 ) = 2, I get the following picture (based on the fact that 2 = x1 x2
implies x2 = x41 ) drawn in green:

x2

10
9
8
7
6
5
4
3
2
1
0
0

10

x1

If I now plot the indierence curve that runs through the points (1; 4), (2; 2), (4; 1) basically, the
indierence curve for v (x1 ; x2 ) = ln 4, I get the exact same picture (since ln 4 = ln x1 + ln x2 implies
ln x41 = ln x2 which further implies x41 = x2 ) denoted by the dashed red line.
2. Graph an indierence curve, and compute the marginal rate of substitution and the Marshallian demand
functions for the following utility functions:
a Perfect substitutes: u(x1 ; x2 ) = x1 + x2 , where

> 0 and

> 0.

Answer:
The marginal rate of substitution is simply the slope of the indierence curve (or the ratio of the
marginal utilities). In this case:
M Ux1
M Ux2

=
=

So we have:
M RS =
While the linear utility function is dierentiable, if we attempt to use Lagranges method (and assume
an interior solution), we will get something that looks like:
=

p1
p2

Now, this may or may not be true. If it is true, then we have a Marshallian demand correspondence
(and not a function), as any point along the budget constraint will be an optimal solution to the
problem. Hence we would have the set:
x 2 RL
+ : px = y
If we were to have:
p1

>

p2

then the marginal utility per dollar spent on good 1 is higher than that of good 2. So the consumer
would only purchase good 1 and we would be at a corner solution where:
x1

x2

y
p1
0

Finally, if
p2

>

p1

then the consumer would purchase only good 2 and the optimal bundle would be 0; py2 . To summarize:
x1 (p; y)
x2 (p; y)

=
=

0
y
p1
y
p2

if
if

p2

if
if

p2

p2

p2

>
<

p1

>
<

p1

p1

p1

x 2 RL
+ : px = y
b Perfect complements: u(x1 ; x2 ) = minf x1 ; x2 g, where

if

p2

> 0 and

p1
> 0.

Answer:
The indierence curves for this utility function look like:

x2

10
9
8
7
6
5
4
3
2
1
0
0

10

x1

Since this utility function is nondierentiable, we cannot use Lagranges method. However, looking
at the indierence curves shows us that any optimal consumption bundle must be at the kinked point
of the L-shape. This is because at any other point the consumer will be "wasting" money by buying
too much of one good or the other. This means that x2 = x1 at the optimum. Solving for x1 we

have x1 =

x2

. Now we can plug this into the budget constraint and nd x2 :

y

= p1 x1 + p2 x2
x2
+ p2 x2
= p1

y
y
y
p1 + p2

= p1 x2 + p2 x2
= x2

To nd x1 we have:
x1

x2

x1

x1

y
p1 + p2

y
p1 + p2

For the marginal rate of subsitution we can look at the gure: M RS = 1 when x2 > x1 (the
vertical portion of the indierence curve); M RS = 0 when x2 < x1 (the horizontal portion of the
indierence curve), and M RS is not well dened when x2 = x1 (the kink in the indierence curve).
3. We have noted that u(x) is invariant to positive monotonic transformations. One common transformation is the logarithmic transform, ln (x). Take the logarithmic transform of the Cobb-Douglas utility
function; then using that as the utility function, derive the Marshallian demand functions and verify
that they are identical to those derived in class.
Answer:
The utility function used in class was:
u (x1 ; x2 ) = x1 x2
The demand functions we found were:
x1 (p; y)

x2 (p; y)

y
( + ) p1
y
( + ) p2

Taking the natural log leads to:

v (x1 ; x2 ) = ln (u (x1 ; x2 ))
v (x1 ; x2 ) =
ln x1 + ln x2
Setting up the Lagrangian we have:
L (x1 ; x2 ; ) =

ln x1 +

ln x2 + [y

p1 x1

p2 x2 ]

We know there will be an interior solution (since ln (0) is undened), so we can take rst order conditions
and set them equal to zero:
@L
@x1
@L
@x2
@L
@

=
=

x1
x2

= y

p1 = 0
p2 = 0
p1 x1

p2 x2 = 0

Using the rst two partial derivatives we have:

x1 p1
x2 p2
p1

x2 p2

= x1

Now using the budget constraint we have:

y

= p1 x1 + p2 x2
x2 p2
+ p2 x2
= p1
p1
x2 p2
=
+ p2 x2

y
y

=
x2 p2 + p2 x2
= x2 p2 ( + )

y
p2 ( + )

= x2

For x1 :
x1
x1
x1

x2 p2
p1

y
p2 ( + )

=
=

p2

p1
y
( + ) p1

So these demand functions are the same as the ones we derived in class.
4. A consumer of two goods faces positive prices and has a positive income. Her utility function is
u (x1 ; x2 ) = max fax1 ; ax2 g + min fx1 ; x2 g , for 0 < a < 1
Derive the Marshallian demand functions.
Answer:
For this problem the utility of the consumer is determined by the relationship between x1 and x2 :
if x1
if x1
if x1

> x2 : u (x1 ; x2 ) = ax1 + x2

= x2 : u (x1 ; x2 ) = ax1 + x1
< x2 : u (x1 ; x2 ) = ax2 + x1

Note that this creates a piecewise linear function, although the middle "equation" is not really a line
but a point since we need x1 = x2 . Graphing this for a = 31 and u = 12, this leads to the vertex being
at x1 = x2 = 9. For the remaining pieces of the function we have:
if x1

>

x2 : 12

if x1

<

x2 : 36

Plotting this we have:

1
x1 = x2
3
3x1 = x2

x2

40

30

20

10

0
0

10

15

20

25

30

35

40

x1

Now, lets suppose that p1 = p2 = 1 and that income is 18. Then plotting the budget constraint on
the picture we have:

x2

40

30

20

10

0
0

10

15

20

25

30

35

40

x1

So in this instance we will have the optimal solution at the vertex where x1 = x2 (note that we . So
we can nd the Marshallian demand by substituting x1 for x2 into a general budget constraint and
solving for x1 :
y
y
y
p1 + p 2
Since x1 = x2 we also have x2 (p; y) =

= p1 x1 + p2 x2
= p1 x1 + p2 x1
= x1

y
p1 +p2 .

It looks like we are done, but when we had the simple linear function we had a corner solution with
either 0; py2 or py1 ; 0 as the optimal solution. What if we had that p1 = 5 and p2 = 1? The
budget constraint is now (with y = 36):
6

x2

40

30

20

10

0
0

10

15

20

25

30

35

40

x1

Because the budget constraint is steeper than either of the slopes of the pieces of the indierence curve
we will end up at a corner solution. The same will be true if the slope of the budget constraint is
atter than either piece. If it is steeper we end up with all x2 , and if it is atter we end up with all
x1 . So the actual Marshallian demand function is:
if

p2
p1

if a <
if

p2
p1

y
1
: 0;
a
p2
y
p2
1
< : x1 (p; y) = x2 (p; y) =
p1
a
p1 + p2
y
a:
;0
p1

Technically if the price ratio is equal to one of the slopes we have a demand correspondence (there are
many bundles which will lie along both the indierence curve and the budget constraint).
5. Bob consumes ice cream cones (x1 ) and hamburgers (x2 ). His utility function is
1

u(x1 ; x2 ) = (x1 ) 2 (x2 ) 2

Bobs income is $100. The price of each hamburger is $2. The price of ice cream depends on the
quantity that Bob consumes. Specically, he can buy the rst ten ice cream cones at the price of $2
each. For each additional ice cream cone there is a discount, and Bob has to pay only $1 each.
Derive Bobs budget constraint and compute his optimal consumption plan.
Answer:
Bobs budget constraint has a kink in it. The plot below provides his budget constraint:

hamburgers
50

40

30

20

10

0
0

When x1

10

20

30

40

50

60

70

80

90

ice cream

10, Bobs budget constraint is given by:

100 = 2x1 + 2x2
x2 = 50 x1

When x1 > 10, Bobs budget constraint is given by:

(100

20)

(x1

x2

45

10) + 2x2
1
x1
2

The reason is that he has already spent $20 on his 10 ice cream cones.
If we set up the Lagrangian for the original budget constraint we nd that:
1

L (x1 ; x2 ; ) = (x1 ) 2 (x2 ) 2

[y

p1 x1

p2 x2 ]

Now, we can work through the entire problem, but if we go to problem 3 we know that when u (x1 ; x2 ) =
x1 x2 , we have the following Marshallian demand functions:
x1

x2

y
( + ) p1
y
( + ) p2

Using our parameters with the rst budget constraint we have that Bob would consume:
x1

x2

1
2 100
1
1
2 + 2
1
2 100
1
1
2 + 2

2
2

= 25
= 25

However, x1 > 10 , so now Bobs "optimal" consumption bundle of (25; 25) is inside his feasible set
and not on the budget constraint. To see this look at the gure below:

hamburgers
50

40

30

20

10

0
0

10

20

30

40

50

60

70

80

90

ice cream

The diamond shaped green point is at (25; 25). So we then use the parameters from the new budget
constraint, with y = 90 and p1 = 1, to nd:
x1

x2

1
2 90
+ 12
1
2 90
1
1
2 + 2
1
2

1
2

= 45
= 22:5

Looking at the picture below, we see that at (45; 22:5) the indierence curve is tangent to the budget
constraint:

hamburgers
50

40

30

20

10

0
0

10

20

30

40

50

60

70

80

90

ice cream