ECO 501: Microeconomics I

August 02, 2019

Preference and choice

Individual decision making

Set of alternatives: X

Two approaches: preferences as the primitive vs choice as the

primitive

In preference approach, rationality constraints will be imposed on the

preferences, based on which consequences for choice will be derived

In the choice-based approach, consistency will imposed and the

underlying preferences analysed

Relationship between these two approaches?

Preference Relations

A preference relation is a binary relation on the set of alternatives

For x, y ∈ X, x  y : x is at least as good as y

Derivation of strict preferences, i.e. , and indifference, i.e. ∼, from

the  relation

Rationality of preferences: completeness and transitivity

Completeness: ∀x, y ∈ X, x  y or y  x

Transitivity: ∀x, y, z ∈ X, if x  y and y  z, then x  z

Do completeness and transitivity always hold?

Preference Relations

What does rationality of preference relation  imply for strict

preference  and indifference ∼ ?

 is irreflexive (i.e. x  x never holds), and transitive (i.e.

∀x, y, z ∈ X, if x  y and y  z, then x  z). Why?

∼ is reflexive (i.e. x ∼ x), transitive (i.e. ∀x, y, z ∈ X, if x ∼ y and

y ∼ z, then x ∼ z), and symmetric (i.e. if x ∼ y and y ∼ x). Why?

∀x, y, z ∈ X, if x  y and y  z, then x  z. Why?

∀x, y, z ∈ X, if x  y and y  z, then x  z. Why?

Preference Relations and Utility functions

A utility function u : X → R simply assigns a real number to every

element of the feasible set X

Relation between utility functions and preference relations?

A utility function u : X → R represents the preference relation  if

∀x, y ∈ X, x  y ⇔ u(x) ≥ u(y)

Is a utility function unique?

Preference Relations and Utility functions

A utility function u : X → R simply assigns a real number to every

element of the feasible set X

Relation between utility functions and preference relations?

A utility function u : X → R represents the preference relation  if

∀x, y ∈ X, x  y ⇔ u(x) ≥ u(y)

Is a utility function unique?

If f : R → R is an strictly increasing function, and u is a utility

function representing the preference relation , then the function
v : X → R defined by v(x) = f (u(x)) is also a utility function
representing the preference relation . Why?

Difference between cardinal and ordinal properties?

Preference Relations and Utility functions

A preference relation  can be represented by a utility function only

if it is rational

What do we need to prove to establish the above statement?

Preference Relations and Utility functions

A preference relation  can be represented by a utility function only

if it is rational

What do we need to prove to establish the above statement?

If a utility function represents the preference relation , then 

must be complete and transitive
Preference Relations and Utility functions

A preference relation  can be represented by a utility function only

if it is rational

What do we need to prove to establish the above statement?

If a utility function represents the preference relation , then 

must be complete and transitive

Completeness of ?
Preference Relations and Utility functions

A preference relation  can be represented by a utility function only

if it is rational

What do we need to prove to establish the above statement?

If a utility function represents the preference relation , then 

must be complete and transitive

Completeness of ?

Because u is a real valued function on X, for any x, y ∈ X, either

u(x) ≥ u(y) or u(y) ≥ u(x); but since u represents , we have that
for any x, y ∈ X, either x  y or y  x, i.e.  is complete
Preference Relations and Utility functions

A preference relation  can be represented by a utility function only

if it is rational

What do we need to prove to establish the above statement?

If a utility function represents the preference relation , then 

must be complete and transitive

Completeness of ?

Because u is a real valued function on X, for any x, y ∈ X, either

u(x) ≥ u(y) or u(y) ≥ u(x); but since u represents , we have that
for any x, y ∈ X, either x  y or y  x, i.e.  is complete

Transitivity of ?
Preference Relations and Utility functions

A preference relation  can be represented by a utility function only

if it is rational

What do we need to prove to establish the above statement?

If a utility function represents the preference relation , then 

must be complete and transitive

Completeness of ?

Because u is a real valued function on X, for any x, y ∈ X, either

u(x) ≥ u(y) or u(y) ≥ u(x); but since u represents , we have that
for any x, y ∈ X, either x  y or y  x, i.e.  is complete

Transitivity of ?

Suppose x  y and y  z; this implies u(x) ≥ u(y) and

u(y) ≥ u(z).
By transitivity of real numbers, u(x) ≥ u(z), and hence z  x, i.e.
 is transitive
Preference Relations and Utility functions

We have seen that if a utility function represents a preference

relation, then it must be rational

But can any rational preference relation be represented by a utility

function?
Preference Relations and Utility functions

We have seen that if a utility function represents a preference

relation, then it must be rational

But can any rational preference relation be represented by a utility

function?

What if X is a finite set?

Preference Relations and Utility functions

We have seen that if a utility function represents a preference

relation, then it must be rational

But can any rational preference relation be represented by a utility

function?

What if X is a finite set?

If X is finite and  is a rational preference relation on X, then there

is a utility function U : X → R that represents 

Proof?
Preference Relations and Utility functions

We have seen that if a utility function represents a preference

relation, then it must be rational

But can any rational preference relation be represented by a utility

function?

What if X is a finite set?

If X is finite and  is a rational preference relation on X, then there

is a utility function U : X → R that represents 

Proof?

What if X is countable?
Preference Relations and Utility functions

We have seen that if a utility function represents a preference

relation, then it must be rational

But can any rational preference relation be represented by a utility

function?

What if X is a finite set?

If X is finite and  is a rational preference relation on X, then there

is a utility function U : X → R that represents 

Proof?

What if X is countable?

If  is a rational preference relation relation on a countable set X,

then there is a utility function U : X → R that represents 

Proof?
Choice
Choice structure (B, C(·)) consists of: (i) B, which is a family of
non-empty subsets of X, and (ii) C(·) is a choice rule that assigns a
non-empty set C(B) ⊂ B, for every B ∈ B

For example, X = {x, y, z}, B = {{x, y}, {x, y, z}}

One choice structure is (B, C1 (·)), where C1 ({x, y}) = {x} and
C1 ({x, y, z}) = {x}
Another choice structure is (B, C2 (·)), where C2 ({x, y}) = {x} and
C2 ({x, y, z}) = {x, y}
Choice
Choice structure (B, C(·)) consists of: (i) B, which is a family of
non-empty subsets of X, and (ii) C(·) is a choice rule that assigns a
non-empty set C(B) ⊂ B, for every B ∈ B

For example, X = {x, y, z}, B = {{x, y}, {x, y, z}}

One choice structure is (B, C1 (·)), where C1 ({x, y}) = {x} and
C1 ({x, y, z}) = {x}
Another choice structure is (B, C2 (·)), where C2 ({x, y}) = {x} and
C2 ({x, y, z}) = {x, y}

We will impose a consistency requirement on the choice structure:

the weak axiom of revealed preference (WARP)

The choice structure (B, C(·)) satisfies WARP if the following

property holds:
if for some B ∈ B with x, y ∈ B, we have x ∈ C(B), then for any
B 0 ∈ B with x, y ∈ B 0 and y ∈ C(B 0 ), then we must also have
x ∈ C(B 0 )

If C({x, y}) = {x}, then C({x, y, z}) = ...?

Choice and Revealed preference
Given a choice structure (B, C(·)), the revealed preference relation
∗ is defined as:
x ∗ y ⇔ there is some B ∈ B such that x, y ∈ B and x ∈ C(B),
or x is revealed at least as good as y

Is ∗ necessarily complete or transitive? Why or why not?

Choice and Revealed preference
Given a choice structure (B, C(·)), the revealed preference relation
∗ is defined as:
x ∗ y ⇔ there is some B ∈ B such that x, y ∈ B and x ∈ C(B),
or x is revealed at least as good as y

Is ∗ necessarily complete or transitive? Why or why not?

x is revealed preferred to y, or x ∗ y if ∃B ∈ B such that x, y ∈ B,

x ∈ C(B) and y 6∈ C(B)

Alternative statement of WARP: if x is revealed at least as good as

y, then y cannot be revealed preferred to x

Another alternative statement of WARP: suppose B, B 0 ∈ B, and

x, y ∈ B and x, y ∈ B 0 . If x ∈ C(B) and y ∈ C(B 0 ), then
{x, y} ⊂ C(B) and {x, y} ⊂ C(B 0 )

Is ∗ transitive?
Choice and Revealed preference
Given a choice structure (B, C(·)), the revealed preference relation
∗ is defined as:
x ∗ y ⇔ there is some B ∈ B such that x, y ∈ B and x ∈ C(B),
or x is revealed at least as good as y

Is ∗ necessarily complete or transitive? Why or why not?

x is revealed preferred to y, or x ∗ y if ∃B ∈ B such that x, y ∈ B,

x ∈ C(B) and y 6∈ C(B)

Alternative statement of WARP: if x is revealed at least as good as

y, then y cannot be revealed preferred to x

Another alternative statement of WARP: suppose B, B 0 ∈ B, and

x, y ∈ B and x, y ∈ B 0 . If x ∈ C(B) and y ∈ C(B 0 ), then
{x, y} ⊂ C(B) and {x, y} ⊂ C(B 0 )

Is ∗ transitive?

Is ∗ transitive if B contains all three-element subsets of X?

Preference Relations and Choice Structures

Two main questions:

1 If an individual has a rational preference ordering, then will the

resulting choice structure necessarily satisfy WARP?

2 If the individual’s choice structure satisfies WARP, is there an

underlying rational preference relation that is consistent with the
choices?
Choice and Revealed preference

Suppose an individual has a rational preference  relation on X

The preference maximising behaviour is to choose from the set

C ∗ (B, ) = {x ∈ B : x  y for every y ∈ X}

Assuming C ∗ (B, ) to be non-empty, the rational preference

generates the choice structure (B, C ∗ (·, ))

If X is finite, then C ∗ (B, ) is non-empty. Proof?

Main result: (B, C ∗ (·, )) satisfies WARP. Proof?

Thus, if an individual has a rational preference ordering, then the

resulting choice structure necessarily satisfies WARP
Choice and Revealed preference
Now, consider the second question

Definition: given a choice structure (B, C(·)), the rational preference

relation  rationalises C(·) relative to B if C(B) = C ∗ (B, )
∀B ∈ B, i.e.  generates the choice structure (B, C(·))

Can there be more than one rationalising preference relation for a

given choice structure (B, C(·))?

WARP is a necessary condition for a rationalising preference relation.

Why?

Is WARP sufficient for a rationalising preference relations?

Example: X = {x, y, z}, B = {{x, y}, {y, z}, {x, z}}, C({x, y}) =
{x}, C({y, z}) = {y}, C({x, z}) = {z}
No rational preference ordering can rationalise the above choice
structure. Yet, the choice structure satisfies WARP

What drives the above example?

Choice and Revealed preference
Main result:

If (B, C(·)) is a choice structure such that (i) WARP is satisfies, and (ii)
B includes all subsets of X up to three elements, then there is a unique
rational preference relation  that rationalises C(·) relative to B, i.e.
C ∗ (B, ) = C(B) ∀B ∈ B
Choice and Revealed preference
Main result:

If (B, C(·)) is a choice structure such that (i) WARP is satisfies, and (ii)
B includes all subsets of X up to three elements, then there is a unique
rational preference relation  that rationalises C(·) relative to B, i.e.
C ∗ (B, ) = C(B) ∀B ∈ B

Proof: We will show that ∗ is the rationalising preference relation

The proof is in three steps: (a) ∗ is a rational preference relation, (b)
∗ rationalises C(·) on B, and (c) ∗ is the unique preference relation to
rationalise C(·) on B

(a) Required to show: ∗ is complete and transitive?

(b) Required to show: C(B) = C ∗ (B, ∗ ), ∀B ∈ B

First, establish C(B) ⊂ C ∗ (B, ∗ ) by showing that if x ∈ C(B), then
x ∈ C ∗ (B, ∗ )
Second, establish C ∗ (B, ∗ ) ⊂ C(B) by showing that if x ∈ C ∗ (B, ∗ ),
then x ∈ C(B)

(c) Uniqueness?
Summary

Two approaches: one takes preferences as the primitive while the

other takes choices as the primitive

In the preference approach, rationality axioms (i.e. completeness and

transitivity) are imposed

If there is a utility function that represents the preferences, then the

preferences must be rational; the converse is true when the set of
alternatives is finite, but it is not generally true

In the choice based approach, the weak axiom of revealed preference

is imposed as a consistency check

The revealed preference relation may not be complete or transitive

Rational preference orders always give rise to choice structures that

satisfy the weak axiom

Under certain conditions, any choice structure that satisfies the

weak axiom had a unique rationalising preference order
Reference

Mas-colell, Whinston and Green (1995) Microeconomic Theory, Ch 1