Microeconomics MScE: In Class Homework Ch2

Bettina Klaus
J&R Exercise 2.7
Derive the consumers inverse demand functions, p
1
(x
1
, x
2
) and p
2
(x
1
, x
2
), when the utility func-
tion is of the Cobb-Douglas form, u(x
1
, x
2
) = Ax

1
x
1
2
for 0 < < 1.
Let u(x) = Ax

1
x
1
2
; (0, 1). By Hotelling Wold,
p
i
(x) =
u(x)
x
i

2
j=1
x
j
u(x)
x
j
.
We have
u(x)
x
1
= Ax
1
1
x
1
2
,
u(x)
x
2
= (1 )Ax

1
x

2
,
u(x)
x
1
x
1
+
u(x)
x
2
x
2
= Ax

1
x
1
2
+ (1 )Ax

1
x
1
2
= Ax

1
x
1
2
.
So,
p
1
(x) =
Ax
1
1
x
1
2
Ax

1
x
1
2
=

x
1
and
p
2
(x) =
(1 )Ax

1
x

2
Ax

1
x
1
2
=
(1 )
x
2
.
1
J&R Exercise 2.10
Hicks (1956) offered the following example to demonstrate how WARP can fail to result in tran-
sitive revealed preferences when there are more than two good. The consumer chooses bundle
x
i
at prices p
i
, i = 0, 1, 2, where:
p
0
=

1
1
2

x
0
=

5
19
9

p
1
=

1
1
1

x
1
=

12
12
12

p
2
=

1
2
1

x
2
=

27
11
1

(a) Show that these data satisfy WARP. Do it by considering all possible pairwise comparisons
of the bundles and showing that in each case, one bundle in the pair is revealed preferred
to the other.
(b) Find the intransitivity in the revealed preferences.
WARP: p x p x p x > p x. Note that
p
0
x
0
= 42, p
1
x
1
= 36, and p
2
x
2
= 50.
(a) Compare x
0
, x
1
: Note that p
0
x
1
= 48 and p
1
x
0
= 33. Thus,
p
0
x
0
= 42 48 = p
0
x
1
(WARP satised),
p
1
x
1
= 36 33 = p
1
x
0
and p
0
x
1
= 48 > 42 = p
0
x
0
(WARP satised).
The second statement implies that x
1
is revealed preferred to x
0
, i.e., x
1

R
x
0
.
Compare x
1
, x
2
: Note that p
1
x
2
= 39 and p
2
x
1
= 48. Thus,
p
1
x
1
= 36 39 = p
1
x
2
(WARP satised),
p
2
x
2
= 50 48 = p
2
x
1
and p
1
x
2
= 39 > 36 = p
1
x
1
(WARP satised).
The second statement implies that x
2
is revealed preferred to x
1
, i.e., x
2

R
x
1
.
Compare x
2
, x
0
: Note that p
2
x
0
= 52 and p
0
x
2
= 40. Thus,
p
2
x
2
= 50 52 = p
2
x
0
(WARP satised),
p
0
x
0
= 42 40 = p
0
x
2
and p
2
x
0
= 52 > 50 = p
2
x
2
(WARP satised).
The second statement implies that x
0
is revealed preferred to x
2
, i.e., x
0

R
x
2
.
(b) We now have that x
0

R
x
2
and x
2

R
x
1
. Hence, transitivity would imply that x
0

R
x
1
.
However, when comparing bundles x
1
and x
0
we found x
1

R
x
0
. Note that x
0

R
x
1
and
x
1

R
x
0
cannot be true at the same time. Therefore, transitivity is not satised.
2
J&R Exercise 2.25
Consider the quadratic VNM utility function U(w) = a + bw + cw
2
.
(a) What restrictions if any must be placed on parameters a, b, and c for this function to display
risk aversion?
(b) Over what domain of wealth can a quadratic VNM utility function be dened?
(c) Given the gamble:
g = ((1/2) (w + h), (1/2) (w h)),
show that CE < E(g) and that P > 0.
(d) Show that this function, satisfying the restrictions in part (a), cannot represent preferences
that display decreasing absolute risk aversion.
Let U(w) = a + bw + cw
2
. First we need to assume that U

(w) = b + 2cw > 0.

(a) risk aversion U

(w) = 2c < 0 c < 0. Thus, b > 2cw > 0.

(b) Since U

(w) = b + 2cw > 0 w <

b
2c
.
(c) Let g = (
1
2
(w + h),
1
2
(w h)). Show that CE < E(g) and P > 0.
By denition of the certainty equivalent, u(CE) = u(g).
u(g) =
1
2
[a + b(w + h) + c(w + h)
2
] +
1
2
[a + b(w h) + c(w h)
2
]
= a +
1
2
b[w + h + w h] +
1
2
c[(w + h)
2
+ (w h)
2
]
= a + bw +
1
2
c[w
2
+ 2wh + h
2
+ w
2
2hw + h
2
]
= a + bw + c(w
2
+ h
2
)
u(CE) = u(g)
= a + bw + c(w
2
+ h
2
)
< a + bw + cw
2
= u(w).
Thus,
u(CE) < u(w)
U

>0
CE < w.
Then,
P = E(g) CE = w CE > 0.
(d) Decreasing absolute risk aversion means that
R

a
(w) =

(w)
U

(w)

< 0.
However, since b + 2cw > 0 and 4c
2
> 0 we have that
R

a
(w) =

2c
b + 2cw

=
4c
2
(b + 2cw)
2
> 0.
3
J&R Exercise 2.27
Show that for > 0, the VNM utility function u(w) = + ln(w) displays decreasing absolute
risk aversion.
Let > 0 and u(w) = + ln w.
R
a
(w) =
u

(w)
u

(w)
=

w
2

w
=
1
w
.
Hence,
R

a
(w) =
1
w
2
< 0
which means that u(w) displays decreasing absolute risk aversion.
J&R Exercise 2.28
Let u(x
1
, x
2
) = ln(x
1
) +2 ln(x
2
). If p
1
= p
2
= 1, will this person be risk loving, risk neutral, or
risk averse when offered gambles over different amounts of income?
Let u(x) = ln x
1
+ 2 ln x
2
. Let p = (1, 1). Note that the given utility function is dened over
bundles of commodities and not over wealth/income levels. In order to determine the risk attitude
of the agent when offered different gambles over amounts of income, we rst have to nd the
indirect utility function associated with p and y.
max
xR
2
+
u(x) s.t. p
1
x
1
+ p
2
x
2
= y
max
xR
2
+
ln x
1
+ 2 ln x
2
s.t. x
1
+ x
2
= y.
Note that at a maximum x
1
= 0 and x
2
= 0.
L(p, ) = ln x
1
+ 2 ln x
2
+ (y x
1
x
2
)
(FOC)

1
x
1
= ,
2
x
2
= , and x
1
+ x
2
= y.
Hence, = 0 and
1
x
1
=
2
x
2
x
2
= 2x
1
.
Thus, x
1
+ 2x
1
= y x
1
=
y
3
and x
2
=
2y
3
.
So, the agents indirect utility given the prices p = (1, 1) equals
v(y) = ln
y
3
+ 2 ln
2y
3
= ln y ln 3 + 2[ln 2 + ln y ln 3]
= 3 ln y + 2 ln 2 3 ln 3.
Note that v

(y) =
3
y
and v

(y) =
3
y
2
< 0. Hence, the agent will be risk averse.
4