MiA T5 ConsumerII

Microeconomic Analysis (806L1)
Topic 5: Consumer Theory Part 2
Matthew Embrey
University of Sussex
Autumn 2023
Matthew Embrey (University of Sussex) Microeconomic Analysis (806L1) Autumn 2023 1 / 18

Consumer Theory
Visual scheme

Consumer Theory
Visual scheme

Expenditure Mimimisation
The Dual Problem – Problem Statement
The consumer’s expenditure minimisation problem:

The desired
level of utility u ⇒ utility constraint set
P = (x1 , x2 ) ∈ R2+ |u (x1 , x2 ) ≥ u
Given prices p1 , p2 , choose cheapest bundle from utility constraint set
Objective is to minimise expenditure, p1 x1 + p2 x2


The desired
P = (x1 , x2 ) ∈ R2+ |u (x1 , x2 ) ≥ u
Specific example with General case:
u (h1 , h2 ) = h1 h2 :
min p1 h1 + p2 h2 min p1 h1 + p2 h2
h1 ,h2 h1 ,h2
subject to subject to
h1 h2 ≥ u u (h1 , h2 ) ≥ u
h1 ≥ 0 h1 ≥ 0
h2 ≥ 0 h2 ≥ 0


The desired
P = (x1 , x2 ) ∈ R2+ |u (x1 , x2 ) ≥ u
Specific example with General case:
u (h1 , h2 ) = h1 h2 :
min p1 h1 + p2 h2 min p1 h1 + p2 h2
h1 ,h2 h1 ,h2
subject to subject to
h1 h2 ≥ u u (h1 , h2 ) ≥ u
h1 ≥ 0 h1 ≥ 0
h2 ≥ 0 h2 ≥ 0
Form the Lagrangian: Form the Lagrangian:
L = p1 h1 + p2 h2 + µ(u − h1 h2 ) L = p1 h1 + p2 h2 + µ(u − u (h1 , h2 ))

Expenditure Minimisation
First Order Conditions
Specific example (cont.): General case (cont.)

First order conditions (FOC): First order conditions (FOC):
∂L ∂L
L1 = = p1 − µh2 = 0 L1 = = p1 − µu1 (h1 , h2 ) = 0
∂h1 ∂h1
∂L ∂L
L2 = = p2 − µh1 = 0 L2 = = p2 − µu2 (h1 , h2 ) = 0
∂h2 ∂h2
∂L ∂L
Lµ = = u − h1 h2 = 0 Lµ = = u − u (h1 , h2 ) = 0
∂µ ∂µ

First order conditions (FOC): First order conditions (FOC):
∂L ∂L
L1 = = p1 − µh2 = 0 L1 = = p1 − µu1 (h1 , h2 ) = 0
∂h1 ∂h1
∂L ∂L
L2 = = p2 − µh1 = 0 L2 = = p2 − µu2 (h1 , h2 ) = 0
∂h2 ∂h2
∂L ∂L
Lµ = = u − h1 h2 = 0 Lµ = = u − u (h1 , h2 ) = 0
∂µ ∂µ
Eliminate µ: Eliminate µ:
p1 p2 p1 p2
=µ= =µ=
h2 h1 u1 (h1 , h2 ) u2 (h1 , h2 )
⇒ ⇒
h2 p1 u1 (h1 , h2 ) p1
= =
h1 p2 u2 (h1 , h2 ) p2

Hicksian Demands
This yields the familiar tangency condition

u
For the specific example, re-write utility constraint as h2 = h1 and
substitute into tangency condition
u p1
=
h12 p2
⇒
p2 u 1
h1∗ = ( )2
p1
Similarily,
p1 u 1
h2∗ = ( )2
p2
These are the Hicksian or compensated demands (functions of price

and utility).
Change in Price
Marshallian v. Hicksian Demands

Change in Price
Marshallian v. Hicksian Demands

Expenditure function
The expenditure function e(u, p⃗) informs us about the minimum level
of income required to achieve utility u in face of prices p⃗.
Useful in public-policy analyses - we can monetize trade-offs (instead
of dealing with ‘utils’.)
From our simple example:
Solving expenditure minimisation yields the Hicksian demands:
1 1
h1 = ( pp21ū ) 2 and h2 = ( pp12ū ) 2
1
e = p⃗ · ⃗h = 2(p1 p2 ū) 2
Properties of the expenditure function e(u, p⃗):
Homogeneous of degree 1 in prices (i.e. double all prices ⇒ double
expenditure needed to maintain utility level without changing ⃗h)
Shephard’s lemma: hi = ∂e(u,⃗
∂pi
p)
Concave in p⃗

Shephard’s lemma
∂e
Shephard’s lemma: ∂pi = hi
Minh1 ,h2 e = p1 h1 + p2 h2 subject to u ≥ ū
Form the Lagrangian: L = p1 h1 + p2 h2 + µ(ū − u)
Recall the following FOC: µ = pu11 = pu22
∂L ∂h1 ∂h1 ∂h2 ∂h2

= h1 + (p1 − µu1 ) + (p2 − µu2 )
∂p1 ∂p1 ∂p1 ∂p1 ∂p1
∂h1 p1 ∂h1 ∂h2 p2 ∂h2
= h1 + (p1 − u1 ) + (p2 − u2 )
∂p1 u1 ∂p1 ∂p1 u2 ∂p1
= h1
∂L ∂e
Thus ∂p 1
= ∂p 1
= h1
This result can also be derived directly by applying the envelope
theorem to e(⃗p , u).

Concavity of the expenditure function
Intuitively, as price of a good rises, consumer adjusts bundle (substitutes away) so

that expenditure does not rise proportionally.

Concavity of the expenditure function
Intuitively, as price of a good rises, consumer adjusts bundle (substitutes away) so

that expenditure does not rise proportionally.
Formally, for any p⃗1 , p⃗2 and p⃗ = α⃗p 1 + (1 − α)⃗
p 2 for α ∈ [0, 1],
1 2
p , u) ≥ αe(⃗
e(⃗ p , u) + (1 − α)e(⃗ p , u). Proof:
Let ⃗h1 , ⃗h2 , ⃗h be the Hicksian demands at prices p ⃗1 , p⃗2 , p
⃗ respectively.
p 1 , u) = p
e(⃗ ⃗1 · ⃗h1 ≤ p
⃗1 · ⃗h (since ⃗h1 is the expenditure minimising bundle at p ⃗1 ).
Similarly, e(⃗ 2
p , u) = p 2
⃗ ·h ≤p⃗ 2 2 ⃗
⃗ · h.
αe(⃗p 1 , u) + (1 − α)e(⃗ p 2 , u) ≤ (α⃗ p 2 ) · ⃗h = p
p 1 + (1 − α)⃗ ⃗ · ⃗h = e(⃗
p , u)

Duality in Consumer Theory
Duality graphically
Maximizing utility subject to a budget constraint leads to the same x ∗

as minimising expenditure subject to a utility level.
The first case leads to Marshallian demand x ∗ (m, p⃗) while the second
leads to Hicksian demand x ∗ (ū, p⃗).

Duality graphically


Duality graphically


Duality graphically


Duality graphically


Duality graphically


Duality graphically


Duality graphically


Indirect utility and expenditure as inverses
Utility maximisation: v (m, p⃗) = max⃗x u(⃗x ) subject to p⃗ · ⃗x ≤ m; let

v ∗ be indirect utility (function of income and price) achieved at the
optimum.
Expenditure minimisation: e(v ∗ , p⃗) = min⃗x p⃗ · ⃗x subject to u(⃗x ) ≥ v ∗ .
Same solution ⃗x ∗ in both cases although function of different
parameters.
1
Lagrange multipliers are inverse of each other: λ = µ
Identities:
e(v (⃗p , m), p⃗) ≡ m; the minimum expenditure required to reach utility
v (m, p⃗) is m.
v (e(u, p⃗), p⃗) ≡ u; the maximum achievable utility from income e(u, p⃗)
is u.

Consumer Theory
Visual scheme

Consumer Theory
Slutsky equation
There is a close relationship between the Marshallian and Hicksian

demands.

Consumer Theory
Slutsky equation
There is a close relationship between the Marshallian and Hicksian

demands.
Duality implies that for any chosen utility level ū:
⃗x (⃗ p , ū)) = ⃗h(⃗
p , e(⃗ p , ū)
But the two demand functions respond differently to a price change.
Differentiate the above equality for good 1 w.r.t p1 :
∂h1 ∂x1 ∂x1 ∂e
= +
∂p1 ∂p1 ∂m ∂p1
∂x1 ∂h1 ∂x1 ∂e
= −
∂p1 ∂p1 ∂m ∂p1
∂x1 ∂h1 ∂x1 ∂e
= − h1 (using Shephard’s lemma, = h1 )
∂p1 ∂p1 ∂m ∂p1
This last equation is the Slutsky equation linking the price responses
of Marshallian and Hicksian demands.
Consumer Theory
Substitution and Income effects
∂x1 ∂h1 ∂x1

= − h1
∂p1 ∂p1 ∂m
The first term on the RHS is the substitution effect.
The substitution effect is the compensated price response.
It is always negative (it represents movement along an indifference
curve which slope downwards).
The Hicksian demand curve is composed only of the substitution effect.
The second term is the income effect.
The income effect is scaled by the effective change in income due to
∂e
the price change (recall that h1 = ∂p 1
)
If the consumer is buying a lot of good 1 i.e. h1 is large, a change in p1
has a large income effect.
The sign of the income effect is positive for a normal good and
negative for an inferior good.

Consumer Theory
Marshallian and Hicksian demand curves
Hicksian steeper than Marshallian for a normal good. For instance, an

increase in p1 reduces purchasing power. Consumer has to be
compensated to stay on the same indifference curve and so there is a
smaller drop in q1 .

Consumer Theory
Giffen goods
∂x1 ∂h1
Recall the Slutsky equation: ∂p1 = ∂p1 − h1 ∂x
∂m .
1
If a good is strongly inferior ( ∂x

< 0), then it could be Giffen
∂m
1
(demand goes up as price rises)

Rice & wheat noodles (staple foods) in China as Giffen goods?
Rice in Hunan while wheat noodles in Gansu province.
Vouchers for price subsidy on the staples distributed to poor urban
households in a randomised trial.
About 70% of calories derived from the staple food.
Enough vouchers i.e. a price reduction on the staple with no quantity
constraint.
Evidence of Giffen good for poor (just above subsistence) households -
reduction in purchase of staple and a shift towards expensive meat.
No such shift seen for rich households (staple share of their budget was
very low to being with).
Very poor (subsistence level) households increased their consumption.

Consumer Theory
Welfare change - Compensating Variation
How do we measure the welfare impact of

a price change?
Say customer originally at point a with
utility u0 .
Now price of x1 rises from p10 to p11 ?
Optimum shifts from a to b (lower utility).
Compensating variation (CV) is the
amount of monetary compensation that
would restore the customer to u0 (point c)
p 1 , u0 ) − e(⃗
CV = e(⃗ p 0 , u0 )
∂e
Shephard’s lemma: ∂p 1
= h1
R p11 R p11
CV = p0 de = p0 h1 (⃗ p , u0 )dp1
1 1
CV is the area under the Hicksian demand

curve: ACp11 p10 .

Consumer Theory
Welfare change and Marshallian demand
Consider the Marshallian demand curve

x(⃗
p , m).
Consumer initially at A (x10 , p10 ).
p1 rises to p11 ; consumer shifts to B.
Hicksian demand curves through A and B
are h1 (⃗
p , u0 ) and h1 (⃗
p , u1 ).
CV is area ACp11 p10 .
Equivalent variation (EV) is how much the
consumer is willing to pay to return to old
prices at the new level of utility.
R p1
EV = p01 h1 (⃗p , u1 )dp1 ; area DBp11 p10 .
1
CV or EV can be used to measure welfare

changes.
If income effects are small, area under the
Marshallian demand curve is a good
approximation.

MiA T5 ConsumerII

Uploaded by

Copyright:

Available Formats

MiA T5 ConsumerII

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

MiA T5 ConsumerII

Uploaded by

Copyright:

Available Formats

Microeconomic Analysis (806L1)

Topic 5: Consumer Theory Part 2

Matthew Embrey (University of Sussex) Microeconomic Analysis (806L1) Autumn 2023 1 / 18

Matthew Embrey (University of Sussex) Microeconomic Analysis (806L1) Autumn 2023 2 / 18

Matthew Embrey (University of Sussex) Microeconomic Analysis (806L1) Autumn 2023 2 / 18

The consumer’s expenditure minimisation problem:

Matthew Embrey (University of Sussex) Microeconomic Analysis (806L1) Autumn 2023 3 / 18

The consumer’s expenditure minimisation problem:

Matthew Embrey (University of Sussex) Microeconomic Analysis (806L1) Autumn 2023 3 / 18

The consumer’s expenditure minimisation problem:

Form the Lagrangian: Form the Lagrangian:

L = p1 h1 + p2 h2 + µ(u − h1 h2 ) L = p1 h1 + p2 h2 + µ(u − u (h1 , h2 ))

Matthew Embrey (University of Sussex) Microeconomic Analysis (806L1) Autumn 2023 3 / 18

Specific example (cont.): General case (cont.)

L = p1 h1 + p2 h2 + µ(u − h1 h2 ) L = p1 h1 + p2 h2 + µ(u − u (h1 , h2 ))

Matthew Embrey (University of Sussex) Microeconomic Analysis (806L1) Autumn 2023 4 / 18

Specific example (cont.): General case (cont.)

L = p1 h1 + p2 h2 + µ(u − h1 h2 ) L = p1 h1 + p2 h2 + µ(u − u (h1 , h2 ))

First order conditions (FOC): First order conditions (FOC):

Matthew Embrey (University of Sussex) Microeconomic Analysis (806L1) Autumn 2023 4 / 18

Specific example (cont.): General case (cont.)

L = p1 h1 + p2 h2 + µ(u − h1 h2 ) L = p1 h1 + p2 h2 + µ(u − u (h1 , h2 ))

First order conditions (FOC): First order conditions (FOC):

Matthew Embrey (University of Sussex) Microeconomic Analysis (806L1) Autumn 2023 4 / 18

This yields the familiar tangency condition

These are the Hicksian or compensated demands (functions of price

Matthew Embrey (University of Sussex) Microeconomic Analysis (806L1) Autumn 2023 6 / 18

Matthew Embrey (University of Sussex) Microeconomic Analysis (806L1) Autumn 2023 6 / 18

Matthew Embrey (University of Sussex) Microeconomic Analysis (806L1) Autumn 2023 7 / 18

∂L ∂h1 ∂h1 ∂h2 ∂h2

Matthew Embrey (University of Sussex) Microeconomic Analysis (806L1) Autumn 2023 8 / 18

Intuitively, as price of a good rises, consumer adjusts bundle (substitutes away) so

Matthew Embrey (University of Sussex) Microeconomic Analysis (806L1) Autumn 2023 9 / 18

Intuitively, as price of a good rises, consumer adjusts bundle (substitutes away) so

Matthew Embrey (University of Sussex) Microeconomic Analysis (806L1) Autumn 2023 9 / 18

Maximizing utility subject to a budget constraint leads to the same x ∗

Matthew Embrey (University of Sussex) Microeconomic Analysis (806L1) Autumn 2023 10 / 18

Maximizing utility subject to a budget constraint leads to the same x ∗

Matthew Embrey (University of Sussex) Microeconomic Analysis (806L1) Autumn 2023 10 / 18

Maximizing utility subject to a budget constraint leads to the same x ∗

Matthew Embrey (University of Sussex) Microeconomic Analysis (806L1) Autumn 2023 10 / 18

Maximizing utility subject to a budget constraint leads to the same x ∗

Matthew Embrey (University of Sussex) Microeconomic Analysis (806L1) Autumn 2023 10 / 18

Maximizing utility subject to a budget constraint leads to the same x ∗

Matthew Embrey (University of Sussex) Microeconomic Analysis (806L1) Autumn 2023 10 / 18

Maximizing utility subject to a budget constraint leads to the same x ∗

Matthew Embrey (University of Sussex) Microeconomic Analysis (806L1) Autumn 2023 10 / 18

Maximizing utility subject to a budget constraint leads to the same x ∗

Matthew Embrey (University of Sussex) Microeconomic Analysis (806L1) Autumn 2023 10 / 18

Maximizing utility subject to a budget constraint leads to the same x ∗

Matthew Embrey (University of Sussex) Microeconomic Analysis (806L1) Autumn 2023 10 / 18

Utility maximisation: v (m, p⃗) = max⃗x u(⃗x ) subject to p⃗ · ⃗x ≤ m; let

Matthew Embrey (University of Sussex) Microeconomic Analysis (806L1) Autumn 2023 11 / 18

Matthew Embrey (University of Sussex) Microeconomic Analysis (806L1) Autumn 2023 12 / 18

There is a close relationship between the Marshallian and Hicksian

Matthew Embrey (University of Sussex) Microeconomic Analysis (806L1) Autumn 2023 13 / 18

There is a close relationship between the Marshallian and Hicksian

∂x1 ∂h1 ∂x1

Matthew Embrey (University of Sussex) Microeconomic Analysis (806L1) Autumn 2023 14 / 18