EC9D3 Lecture 2

EC9D3 Advanced Microeconomics, Part I: Lecture 2
Francesco Squintani
August, 2020
Budget Set
Up to now we focused on how to represent the consumer’s

preferences.
We shall now consider the sour note of the constraint that is imposed
on such preferences.
Definition (Budget Set)

The consumer’s budget set is:
B(p, m) = {x | (p x) ≤ m, x ∈ X }
Francesco Squintani EC9D3 Advanced Microeconomics, Part I August, 2020 2 / 49

Budget Set (2)
6
x1
L = 2 X = R2+
c
c
c
c
c
c
c
c
c
c
c
B(p, m)
c
c
c
c
c
c
c
c -
x2
Income and Prices
The two exogenous variables that characterize the consumer’s budget

set are:
the level of income m
the vector of prices p = (p1 , . . . , pL ).
Often the budget set is characterized by a level of income represented

by the value of the consumer’s endowment x0 (labour supply):
m = (p x0 )

Utility Maximization
The basic consumer’s problem (with rational, continuous and monotonic

preferences):
max u(x)
{x}
s.t. x ∈ B(p, m)
Result
If p > 0 and u(·) is continuous, then the utility maximization problem has
a solution.
Proof: If p > 0 (i.e. pl > 0, ∀l = 1, . . . , L) the budget set is compact

(closed, bounded) hence by Weierstrass theorem the maximization of a
continuous function on a compact set has a solution.

First Order Condition
Result
If u(·) is continuously differentiable, the solution x ∗ = x(p, m) to the
consumer’s problem is characterized by the following necessary conditions.
There exists a Lagrange multiplier λ such that:
∇u(x ∗ ) ≤ λ p
x ∗ [∇u(x ∗ ) − λ p] = 0
p x∗ ≤ m
λ [p x ∗ − m] = 0.
where
∇u(x ∗ ) = [u1 (x ∗ ), . . . , uL (x ∗ )].

First Order Condition (2)
Meaning that ∀l = 1, . . . , L:
ul (x ∗ ) ≤ λpl
and
xl∗ [ul (x ∗ ) − λpl ] = 0
That is if xl∗ > 0 then ul (x ∗ ) = λpl while if ul (x ∗ ) < λpl then xl∗ = 0.
Moreover
L
" L #
X X
pl xl∗ ≤ m, and λ pl xl∗ − m = 0
l=1 l=1

In other words:
if λ > 0 then
L
X
pl xl∗ = m.
l=1
if
L
X
pl xl∗ < m.
l=1
then λ = 0
If preferences are strongly monotonic (or locally non-satiated) then

L
X
pl xl∗ = m.
l=1

In the case L = 2 and X = R2+ these conditions are:
u1 p1
if x1∗ > 0 and x2∗ > 0 then =
u2 p2
u1 p1
if < then x1∗ = 0 and x2∗ > 0
u2 p2
u1 p1
if > then x1∗ > 0 and x2∗ = 0
u2 p2

Interior Solution L = 2
x2 6
(x1∗ , x2∗ )
u(x1 , x2 ) = ū
p1 x1 + p2 x2 = m
-
x1
0

Corner Solution L = 2
u(x1 , x2 ) = ū
x2 6
p1 x1 + p2 x2 = m (x1∗ , x2∗ )
-
x1
0

Sufficient Conditions
The conditions we stated are merely necessary.
What about sufficient conditions?
Result
If u(·) is quasi-concave and monotone,
∇u(x) 6= 0 for all x ∈ X,
then the Kuhn-Tucker first order conditions are sufficient.

Sufficient Conditions (2)
Result
If u(·) is not quasi-concave then a u(·) locally quasi-concave at x ∗ , where
x ∗ satisfies FOC, will suffice for a local maximum.
Local (strict) quasi-concavity can be verified by checking whether the

determinants of the bordered leading principal minors of order
r = 2, . . . , L
of the Hessian matrix of u(·) at x ∗ have the sign of
(−1)r .

The Hessian is:

∂2u ∂2u
 
∂x12
··· ∂x1 ∂xL

H= .. .. .. 
 . . . 

∂2u ∂2u
∂x1 ∂xL ··· ∂xL2
The bordered leading principal minor of order r of the Hessian is:
[∇u(x ∗ )]T

Hr r
[∇u(x ∗ )]r 0
Hr is the leading principal minor of order r of the Hessian matrix H

and [∇u(x ∗ )]r is the vector of the first r elements of ∇u(x ∗ ).

In the case L = 2 the second order conditions hold whenever the

determinant

∂2u ∂2u ∂u
∂x 2 ∂x1 ∂x 2 ∂x1

1
∂2u ∂2u

∂u > 0.
∂x1 ∂x2 ∂x22 ∂x2
∂u ∂u
∂x
1 ∂x 2
0
That is equivalent to:

d ∂u/∂x2
<0
dx1 ∂u/∂x1
∂u/∂x2 p1
provided that = and p1 x1 + p2 x2 = m.
∂u/∂x1 p2

Marshallian Demands
Definition (Marshallian Demands)

The Marshallian or uncompensated demand functions are the solution to
the utility maximization problem:
 
x1 (p1 , . . . , pL , m)
x = x(p, m) = 
 .. 
. 
xL (p1 , . . . , pL , m)
Notice that strong monotonicity of preferences implies that the budget

constraint will be binding when computed at the value of the Marshallian
demands. (building block of Walras Law)

Indirect Utility Function
Definition
The function obtained by substituting the Marshallian demands in the
consumer’s utility function is the indirect utility function:
V (p, m) = u(x ∗ (p, m))
We derive next the properties of the indirect utility function and of the
Marshallian demands.

Properties of the Indirect Utility Function
∂V ∂V
1 ≥ 0 and ≤ 0 for every i = 1, . . . , L.
∂m ∂pi
2 V (p, m) continuous in (p, m).
It rules out situations in which the consumption feasible set is

non-convex (e.g. indivisibility).

Properties of the Indirect Utility Function (2)
3 V (p, m) homogeneous of degree 0 in (p, m).
Definition
F (x) is homogeneous of degree r iff F (k x) = k r F (x) ∀k ∈ R+
Proof: Multiply both the vector of prices p and the level of income m by
the same positive scalar α ∈ R+ we obtain the budget set:
B(α p, α m) = {x ∈ X | α p x ≤ α m} = B(p, m)
hence the indirect utility (and Marshallian demands) are the same.

4 V (p, m) is quasi-convex in p, that is:
{p | V (p, m) ≤ k}
is a convex set for every k ∈ R.
Proof: let p, m and p 0 , be such that:
V (p, m) ≤ k V (p 0 , m) ≤ k.
and p 00 = tp + (1 − t)p 0 for some 0 < t < 1. We need to show that

also V (p 00 , m) ≤ k. Define:
B = {x | (p x) ≤ m} B 0 = x | (p 0 x) ≤ m B 00 = x | (p 00 x) ≤ m


Claim
It is the case that:
B 00 ⊆ B ∪ B0
Proof: Consider x ∈ B 00 , then
p 00 x = [tp + (1 − t)p 0 ] x
= t (p x) + (1 − t) (p 0 x) ≤ m
which implies either p x ≤ m or/and p 0 x ≤ m, or x ∈ B ∪ B0 .

Now
V (p 00 , m) = max u(x) s.t. x ∈ B 00

{x}
≤ max u(x) s.t. x ∈ B ∪ B0

{x}
= max V (p, m), V (p 0 , m) ≤ k

Since by assumption: V (p, m) ≤ k and V (p 0 , m0 ) ≤ k.

Properties of the Marshallian Demand x(p, m)
1 x(p, m) is continuous in (p, m), (consequence of the convexity of

preferences).
2 xi (p, m) homogeneous of degree 0 in (p, m).
Proof: Once again if we multiply (p, m) by α > 0:
B(α p, α m) = {x ∈ X | α p x ≤ α m} = B(p, m)
the solution to the utility maximization problem is the same.

Constrained Envelope Theorem
Consider the problem:
max f (x)
x
s.t. g (x, a) = 0
The Lagrangian is: L(x, λ, a) = f (x) − λ g (x, a)
The necessary FOC are:
∂g (x ∗ , a)
f 0 (x ∗ ) − λ∗ =0
∂x
g (x ∗ (a), a) = 0

Constrained Envelope Theorem (2)
Substituting x ∗ (a) and λ∗ (a) in the Lagrangian we get:
L(a) = f (x ∗ (a)) − λ∗ (a) g (x ∗ (a), a)
Differentiating, by the necessary FOC, we get:

∗ d x ∗ (a)

dL(a) 0 ∗ ∗ ∂g (x , a)
= f (x ) − λ −
da ∂x da
dλ∗ (a) ∂g (x ∗ , a)
−g (x ∗ (a), a) − λ∗ (a)
da ∂a
∗
∂g (x , a)
= −λ∗ (a)
∂a
In other words: to the first order only the direct effect of a on the
Lagrangian function matters.

Properties of the Marshallian Demand x(p, m) (2)
3 Roy’s identity:
∂V /∂pi
xi (p, m) = −
∂V /∂m
Proof: By the constrained envelope theorem and the observation:
V (p, m) = u(x(p, m)) − λ(p, m) [p x(p, m) − m]
we obtain:
∂V /∂pi = −λ(p, m) xi (p, m) ≤ 0
and
∂V /∂m = λ(p, m) ≥ 0
which is the marginal utility of income.

Notice: the sign of the two inequalities above prove the first property of
the indirect utility function V (p, m).
The proof follows from substituting
∂V /∂m = λ(p, m)
into
∂V /∂pi = −λ(p, m) xi (p, m)
and solving for xi (p, m).

4 Adding up results: From the identity:
p x(p, m) = m ∀p, ∀m
Differentiating with respect to m gives:

L
X ∂xi
pi =1
∂m
i=1
while with respect to pj gives:
L
X ∂xi
xj (p, m) + pi =0
∂pj
i=1

More informatively:
L
X ∂xi
0≥ pi = −xh (p, m)
∂ph
i=1
which means that at least one of the Marshallian demand function has to
be downward sloping in ph .
Consider, now, the effect of a change in income on the level of the

Marshallian demand:
∂xl
∂m

In the two commodities graph the set of tangency points for different
values of m is known as the income expansion path.
In the commodity/income graph the set of optimal choices of the quantity

of the commodity is known as Engel curve.

Income Effect
We shall classify commodities with respect to the effect of changes in

income in:
normal goods:
∂xl
>0
∂m
neutral goods:
∂xl
=0
∂m
inferior goods:
∂xl
<0
∂m

Income Effect (2)
Notice that from the adding up results above for every level of income m
at least one of the L commodities is normal:
L
X ∂xl
pl =1
∂m
l=1
We also classify commodities depending on the curvature of the Engel

curves:
if the Engel curve is convex we are facing a luxury good
If the Engel curve is concave we are facing a necessity.

Income Effect (3)
x(p, m) 6
luxury
necessity
-
m

Expenditure Minimization Problem
The dual problem of the consumer’s utility maximization problem is
the expenditure minimization problem:
min px
{x}
s.t. u(x) ≥ U
Define the solution as the Hicksian (compensated) demand functions:

 
h1 (p1 , . . . , pL , U)
x = h(p, U) = 
 .. 
. 
hL (p1 , . . . , pL , U)
We shall also define the expenditure function as:
e(p, U) = p h(p, U)

Properties of the Expenditure Function
1 e(p, U) is continuous in p and U.
∂e ∂e
2 > 0 and ≥ 0 for every l = 1, . . . , L.
∂U ∂pl
∂e
Proof: > 0. Suppose it does not hold.
∂U
Then there exist U 0 < U 00 such that (denote x 0 and x 00 the,
corresponding, solution to the e.m.p.)
p x 0 ≥ p x 00 > 0
If the latter inequality is strict we have an immediate contradiction of

x 0 solving e.m.p.

Properties of the Expenditure Function (2)
If on the other hand

p x 0 = p x 00 > 0
then by continuity and strict monotonicity of u(·) there exists α ∈ (0, 1)
close enough to 1 such that
u(α x 00 ) > U 0
Moreover
p x 0 > p αx 00
which contradicts x 0 solving e.m.p.

∂e
Proof: ≥0
∂pl
Consider p 0 and p 00 such that pl00 ≥ pl0 but pk00 = pk0 for every k 6= l.
Let x 00 and x 0 be the solutions to the e.m.p. with p 00 and p 0 respectively.
Then by definition of e(p, U)
e(p 00 , U) = p 00 x 00 ≥ p 0 x 00 ≥ p 0 x 0 = e(p 0 , U)
that concludes the proof.

3 e(p, U) is homogeneous of degree 1 in p.
Proof: The feasible set of the e.m.p. does not change when prices
are multiplied by the factor k > 0:
u(x) ≥ U
Hence ∀k > 0, minimizing (k p) x on this set leads to the same

answer.
Let x ∗ be the solution, then:
e(k p, U) = (k p) x ∗ = k e(p, U)
that concludes the proof.

4 e(p, U) is concave in p.
Proof: Let p 00 = t p + (1 − t) p 0 for t ∈ [0, 1].
Let x 00 be the solution to e.m.p. for p 00 .
Then
e(p 00 , U) = p 00 x 00 = t p x 00 + (1 − t) p 0 x 00
≥ t e(p, U) + (1 − t) e(p 0 , U)
by definition of e(p, U) and e(p 0 , U) and u(x 00 ) ≥ U.

Properties of the Hicksian demand functions h(p, U)
1 Shephard’s Lemma.
∂e(p, U)
= hl (p, U)
∂pl
Proof: By constrained envelope theorem.
2 Homogeneity of degree 0 in p.
Proof: By Shephard’s lemma and the following theorem.

Properties of the Hicksian demand functions h(p, U) (2)
Theorem
If a function F (x) is homogeneous of degree r in x then (∂F /∂xl ) is
homogeneous of degree (r − 1) in x for every l = 1, . . . , L.
Proof: Differentiating with respect to xl the identity, F (k x) ≡ k r F (x),

we get:
∂F (k x) ∂F (x)
k = kr
∂xl ∂xl
This is the definition of homogeneity of degree (r − 1):
∂F (k x) ∂F (x)
= k (r −1) .
∂xl ∂xl

Euler Theorem
Theorem (Euler Theorem)

If a function F (x) is homogeneous of degree r in x then:
r F (x) = ∇F (x) x
Proof: Differentiating with respect to k the identity:
F (k x) ≡ k r F (x)
we obtain:
∇F (kx) x = rk (r −1) F (x)
for k = 1 we obtain:
∇F (x) x = r F (x).

3 The matrix of own and cross-partial derivatives with respect to p

(Substitution matrix)
∂h1 ∂h1
···
 
∂p1 ∂pL
S =
 .. .. .. 
. . . 
∂hL ∂hL
∂p1 ··· ∂pL
is negative semi-definite and symmetric.

Proof: Simmetry follows from Shephard’s lemma and Young Theorem:

∂hl ∂ ∂e(p, U) ∂ ∂e(p, U) ∂hi
= = =
∂pi ∂pi ∂pl ∂pl ∂pi ∂pl
Negative semi-definiteness follows from the concavity of e(p, U) and the

observation that S is the Hessian of the function e(p, U).

Identities
Since the expenditure minimization problem is the dual problem of the

utility maximization problem the following identities hold:
V [p, e(p, U)] ≡ U
e[p, V (p, m)] ≡ m
xl [p, e(p, U)] ≡ hl (p, U) ∀l = 1, . . . , L
hl [p, V (p, m)] ≡ xl (p, m) ∀l = 1, . . . , L

Slutsky Decomposition
Start from the identity
hl (p, U) ≡ xl [p, e(p, U)]
if the price pi changes the effect is:

∂hl ∂xl ∂xl ∂e
= +
∂pi ∂pi ∂m ∂pi
Notice that by Shephard’s lemma:
∂e
= hi (p, U) = xi [p, e(p, U)]
∂pi
you obtain the Slutsky decomposition:
∂xl ∂hl ∂xl
= − xi .
∂pi ∂pi ∂m

Slutsky Equation
Own price effect gives Slutsky equation:
∂xl ∂hl ∂xl

= − xl .
∂pl ∂pl ∂m
Substitution effect:
∂hl
∂pl
Income effect:
∂xl
xl
∂m

Slutsky Equation (2)
x1 6
e
e
e
e
e
H e
eH e
e HH 1
e HH eq x
e
e H y
e q HH e
0
e Hq x
e eHH
x 2 e
e e HH
e e HH
e e HH
e e HH
e e HH
e e H -
x2

Slutsky Equation (3)
We know the sign of the substitution effect it is non-positive.
The sign of the income effect depends on whether the good is normal or
inferior.
In particular we conclude that the good is Giffen if

∂xl
>0
∂pl
This is not a realistic feature: inferior good with a big income effect.

EC9D3 Lecture 2

Uploaded by

Copyright:

Available Formats

EC9D3 Lecture 2

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

EC9D3 Lecture 2

Uploaded by

Copyright:

Available Formats

EC9D3 Advanced Microeconomics, Part I: Lecture 2

Up to now we focused on how to represent the consumer’s

Definition (Budget Set)

Francesco Squintani EC9D3 Advanced Microeconomics, Part I August, 2020 2 / 49

The two exogenous variables that characterize the consumer’s budget

the level of income m

the vector of prices p = (p1 , . . . , pL ).

Often the budget set is characterized by a level of income represented

Francesco Squintani EC9D3 Advanced Microeconomics, Part I August, 2020 4 / 49

The basic consumer’s problem (with rational, continuous and monotonic

Proof: If p > 0 (i.e. pl > 0, ∀l = 1, . . . , L) the budget set is compact

Francesco Squintani EC9D3 Advanced Microeconomics, Part I August, 2020 5 / 49

Francesco Squintani EC9D3 Advanced Microeconomics, Part I August, 2020 6 / 49

Francesco Squintani EC9D3 Advanced Microeconomics, Part I August, 2020 7 / 49

If preferences are strongly monotonic (or locally non-satiated) then

Francesco Squintani EC9D3 Advanced Microeconomics, Part I August, 2020 8 / 49

In the case L = 2 and X = R2+ these conditions are:

Francesco Squintani EC9D3 Advanced Microeconomics, Part I August, 2020 9 / 49

Francesco Squintani EC9D3 Advanced Microeconomics, Part I August, 2020 10 / 49

Francesco Squintani EC9D3 Advanced Microeconomics, Part I August, 2020 11 / 49

The conditions we stated are merely necessary.

What about sufficient conditions?

Francesco Squintani EC9D3 Advanced Microeconomics, Part I August, 2020 12 / 49

Local (strict) quasi-concavity can be verified by checking whether the

of the Hessian matrix of u(·) at x ∗ have the sign of

Francesco Squintani EC9D3 Advanced Microeconomics, Part I August, 2020 13 / 49

The Hessian is:

The bordered leading principal minor of order r of the Hessian is:

Hr is the leading principal minor of order r of the Hessian matrix H

Francesco Squintani EC9D3 Advanced Microeconomics, Part I August, 2020 14 / 49

In the case L = 2 the second order conditions hold whenever the

That is equivalent to:  

Francesco Squintani EC9D3 Advanced Microeconomics, Part I August, 2020 15 / 49

Definition (Marshallian Demands)

Notice that strong monotonicity of preferences implies that the budget

Francesco Squintani EC9D3 Advanced Microeconomics, Part I August, 2020 16 / 49

V (p, m) = u(x ∗ (p, m))

Francesco Squintani EC9D3 Advanced Microeconomics, Part I August, 2020 17 / 49

2 V (p, m) continuous in (p, m).

It rules out situations in which the consumption feasible set is

Francesco Squintani EC9D3 Advanced Microeconomics, Part I August, 2020 18 / 49

3 V (p, m) homogeneous of degree 0 in (p, m).

Francesco Squintani EC9D3 Advanced Microeconomics, Part I August, 2020 19 / 49

4 V (p, m) is quasi-convex in p, that is:

is a convex set for every k ∈ R.

Proof: let p, m and p 0 , be such that:

and p 00 = tp + (1 − t)p 0 for some 0 < t < 1. We need to show that

Francesco Squintani EC9D3 Advanced Microeconomics, Part I August, 2020 20 / 49

Proof: Consider x ∈ B 00 , then

which implies either p x ≤ m or/and p 0 x ≤ m, or x ∈ B ∪ B0 .

Francesco Squintani EC9D3 Advanced Microeconomics, Part I August, 2020 21 / 49

V (p 00 , m) = max u(x) s.t. x ∈ B 00

≤ max u(x) s.t. x ∈ B ∪ B0

= max V (p, m), V (p 0 , m) ≤ k

Since by assumption: V (p, m) ≤ k and V (p 0 , m0 ) ≤ k.

Francesco Squintani EC9D3 Advanced Microeconomics, Part I August, 2020 22 / 49

1 x(p, m) is continuous in (p, m), (consequence of the convexity of

2 xi (p, m) homogeneous of degree 0 in (p, m).

Proof: Once again if we multiply (p, m) by α > 0:

That is equivalent to: