Preference and Choice

Advanced Microeconomics
Universidad del Pacco

2016-II

Microeconomics (UP)

2016-II

1 / 37

Basic Principles

Rationality (Irrationality/Bounded Rationality)

Optimality
Equilibrium - Arbitrage
E ciency

Microeconomics (UP)

2016-II

2 / 37

Basic Approaches

Simple, focused problems (Game Theory, Contract Theory,

Mechanism Design)
Partial Equilibrium (Price Theory, One market at a time)
General Equilibrium (Welfare Economics, Macroeconomics, Finance)
Dynamics
Uncertainty

Microeconomics (UP)

2016-II

3 / 37

Some Real World Applications

Policy eects in a single market

Policy eects in the macroeconomy: ination, unemployment
Market design
Supply chain, Infrastructure
Institutional design
Labor contract design
Organization Theory
Management
Finance

Microeconomics (UP)

2016-II

4 / 37

Why mathematics?

Formality - Rigour
Internally consistent predictions
Allow meaningful empirical work:
nd economic explanations for phenomena
derive interpretable predictions from contrastable theories

Microeconomics (UP)

2016-II

5 / 37

Consumers Optimality

We model how consumers behave dening:

The set of options that are possible objects of choice
Among those, a subset that describes what are the feasible objects of
choice.
A theory that gives us the way a consumer chooses among the
feasible items.

Microeconomics (UP)

2016-II

6 / 37

Consumers Optimality: Utility Maximization

Basic model of individual choice: A decision-maker (DM) must choose one
alternative x from a set X .
Chooses to maximize a utility function u.
u species how much utility DM gets from each alternative
u:X !R
Example: DM chooses whether to eat an apple or a banana.
X = fapple,bananag.
Utility function might say u (apple) = 7, u(banana) = 12.

max u = 12
x 2X

arg max u = banana

x 2X

Microeconomics (UP)

2016-II

7 / 37

Consumers Preferences and Choices

u is a description of preferences: how a consumer would rank any two

elements that belong to the set of possible objects of choice.
We will try to see what these hypothetical rankings imply about actual
choices: Are choices rational in the sense of being consistent with those
rankings?
Finally, we ask whether rationality can be tested by an outside observer
(the economist) who does not know what is inside the consumers head:
Can we say whether or not a consumer is rational without knowing her
hypothetical rankings?

Microeconomics (UP)

2016-II

8 / 37

Preference and Choice

A preference relation describes the decision makers taste among any two
hypothetical pairs.
Given this preference, the decision makers will make choices whenever she
is oered some set of alternatives to choose from.
This denes a choice procedure : a way to select elements out of a
particular set.
For example, in intermediate micreconomics we talk about a consumer
maximizing her utility given a set of aordable consumption bundles.
Choosing something that maximizes utility given the budget constraint is a
choice procedure by optimization.
The big picture questions then are the following:
When is some choice procedure the by-product of a preference relation?
Can we use observable choices to deduce what is the preference relation
guiding those choices and whether that preference relation satises some
rationality criteria?
Microeconomics (UP)

2016-II

9 / 37

What do Utility Levels Mean?

Revealed-Preference Interpretation:
Utility represents an individuals choices.
Individual choices are primitive data that economists can observe.
Choices are taken to reveal individuals preferences.
Utility is a convenient mathematical construction for modeling choices and
preferences.
u (apple) = 7, u(banana) = 12 =) individual prefers bananas to apples.
u (apple) = 2, u(banana) = 15 =) individual prefers bananas to apples.

Microeconomics (UP)

2016-II

10 / 37

Choice

How can an individuals choices reveal her preferences?

A choice structure (or choice dataset) (B , C ) consists of:
1. A set B of choice sets B X .
2. A choice rule C that maps each B 2 B to a non-empty set of chosen
alternatives C (B ) B.
Interpretation: C (B ) is the set of alternatives the DM chooses from B.

Microeconomics (UP)

2016-II

11 / 37

Preference

Goal: relate observable choice data to preferences over x.

A preference relation % is a binary relation on x:
x %y
means x is weakly preferred to y.
Given a preference relation %, dene:
Strict preference ( ) : x
Indierence ( ): x

Microeconomics (UP)

y , x % y but not y % x.

y , x % y and y % x.

2016-II

12 / 37

RationaI Preferences

To make any progress, we need to impose some restrictions on preferences.

The most important assumption: rationality
Denition: A preference relation % is rational if it satises:
1. Completeness: for all x, y , x % y or y % x.
2. Transitivity: for all x, y , z, if x % y and y % z, then x % z.
If % is rational, then
and
are also transitive (proof?).
Hard to say much about the behavior of an irrationaI DM.

Microeconomics (UP)

2016-II

13 / 37

Maximizing a Preference ReIation

Optimal choices according to %:

C (B, %) = fx 2 B : x % y for all y 2 Bg
% rationalizes choice data (B , C ) if C (B ) = C (B, %) for all B 2 B

Microeconomics (UP)

2016-II

14 / 37

Fundamental Question of Revealed Preference Theory

When does choice data reveal that an individual is choosing according to

rational preferences?
Denition
Given choice data (B , C ), the revealed preference relation % is dened by
x % y , there is some B 2 B with x, y 2 B and x 2 C (B )
x is weakly revealed preferred to y if x is ever chosen when y is available.
x is strictly revealed preferred to y if there is some B 2 B with x, y 2 B,
x 2 C (B ), and y 62 C (B ).

Microeconomics (UP)

2016-II

15 / 37

Weak Axiom of Revealed Preference

Key condition on choice data for % to be rational and generate observed

data: weak axiom of revealed preference (WARP).
Denition
Choice data (B , C ) satises WARP if whenever there exists B 2 B with
x, y 2 B and x 2 C (B ), then for all B 0 2 B with
x, y 2 B 0 , it is not the case that both y 2 C (B 0 ) and x 62 C (B 0 ).
lf x is weakly revealed preferred to y , then y cannot be strictly revealed
preferred to x
Kreps: Choice coherence.

Microeconomics (UP)

2016-II

16 / 37

WARP: Example

X = fx, y , z g, B = ffx, y g, fx, y , z gg

Choice rule C1 : C1 (fx, y g) = fx g, C1 (fx, y , z g) = fx g.
Satises WARP x is weakly revealed preferred to y and z, nothing is
strictly revealed preferred to x.
Choice rule C2 : C2 (fx, y g) = fx g, C2 (fx, y , z g) = fx, y g.
Violates WARP: y is weakly revealed preferred to x, x is strictly revealed
preferred to y .

Microeconomics (UP)

2016-II

17 / 37

A FundamentaI Theorem of Revealed Preference

Theorem
If choice data (B , C ) satises WARP and includes all subsets of x of up
to 3 elements, then % is rational and rationalizes the data;
that is, C (B, % ) = C (B ). Furthermore, this is the only preference
relation that rationalizes the data.
Conversely, if the choice data violates WARP, then it cannot be
rationalized by any rational preference relation.
This theorem tells us how individuals choices reveal her preferences: as
long as choices satisfy WARP, we can interpret choices as resulting from
maximizing a rational preference relation.

Microeconomics (UP)

2016-II

18 / 37

Preference and Utility

Now that know how to infer preferences from choice, the next step is
representing preferences with a utility function.
Denition
A utility function u : X ! R represents preference relation % if, for all
x, y ,
x % y , u (x ) u (y )
banana %apple is represented by both
u (apple) = 7, u(banana) = 12 and
u (apple) = 2, u(banana) = 15.
lf u represents %, so does any strictly increasing transformation of u.
Representing a given preference relation is an ordinal property.
The numerical values of utility are cardinal properties.

Microeconomics (UP)

2016-II

19 / 37

What Preferences have a Utility Representation?

Theorem
Only rational preferences relations can be represented by a utility function.
(if u exists, then preferences must be rational)
Conversely, if x is nite, any rational preference relation can be
represented by a utility function (because C satises nite nonemptiness
and WARP/choice coherence).
Proof

Microeconomics (UP)

2016-II

20 / 37

What Goes Wrong with lnnitely Many AIternatives?

Lexicographic preferences:
X = [0, 1] [0, 1]
(x1 , x2 ) % (y1 , y2 ) if either
x1 > y1 or x1 = y1 and x2 y2
Maximize rst component. In case of tie, maximize second component.
Theorem
Lexicographic preferences cannot be represented by a utility function.

Microeconomics (UP)

2016-II

21 / 37

Continuous Preferences

What if we rule out discontinuous preferences?

Denition
For x R n , preference relation % is continuous if:
whenever x m ! x, y m ! y , and x m % y m for all m, we have x % y .
Theorem
For x Rn , any continuous, rational preference relation can be
represented by a utility function.

Microeconomics (UP)

2016-II

22 / 37

Review of Revealed Preference Theory

lf choice data satises WARP, we can interpret as resulting from

maximizing a rational preference relation.
lf the set of alternatives is nite or preferences are continuous, we can
represent these preferences with a utility function.
Utility function is just a convenient mathematical representation of
individuals ordinal preferences.

Microeconomics (UP)

2016-II

23 / 37

Properties of Preferences and Utility Functions

Doing useful analysis entails making assumptions.

Try to do this carefully: make clearest, simplest, least restrictive
assumptions.
Understand what assumptions about utility correspond to in terms of
preferences, since utility is just a way of representing preferences.
We now cover some of the most important assumptions on preferences.
(And, implicitly, on choices.)

Microeconomics (UP)

2016-II

24 / 37

Some Notation

Assume x R n .
Example: Consumer Problem: given xed budget, choose how much of n
goods to consume
Notation: for vectors x = (x1 , . . . , xn ) and y = (y1 , . . . , yn ),
x y means xk
yk for all k = 1, . . . , n
x > y means xk
yk for all k and xk > yk for some k
x
y means xk > yk for all k
For 2 [0, 1],
x + (1

Microeconomics (UP)

)y = (x1 + (1

)y1 , . . . , xn + (1

)yn )

2016-II

25 / 37

Monotonicity: Preferences

More goods are desirable than less

Denition
% is monotone if x y implies x % y .
% is strictly monotone if x > y implies x

Microeconomics (UP)

y.

2016-II

26 / 37

Monotonicity: Utility

lf preferences are monotone, what does that mean for the utility function?
Theorem
Suppose utility function u represents preferences %. Then:
u is non-decreasing , % is monotone
u is strictly increasing , % is strictly monotone

Microeconomics (UP)

2016-II

27 / 37

LocaI-Nonsatiation

No bliss points. (Not even local ones.)

Let Be (x ) = fy : kx y k < eg.
Denition
% is locally non-satiated if for any x and e > 0, there exists y 2 Be (x )
with y x.
lf u represents %, then % is locally non-satiated i u has no local
maximum.

Microeconomics (UP)

2016-II

28 / 37

Convexity

Diversity is good.
Denition
% is convex if x % y and x 0 % y imply x + (1 )x 0 % y for all
2 (0, 1)
% is strictly convex if x % y , x 0 % y , and x 6= x 0 imply x + (1 )x 0
for all 2 (0, 1)
Does this make sense?
ls 12 beer+ 12 wine a good thing?
We now discuss several properties of convex preferences.

Microeconomics (UP)

2016-II

29 / 37

Contour Sets

For x 2 X , the upper contour set of x (everything superior or equivalent

to x) is
U (x ) = fy : y % x g.
Theorem
% is convex i U (x ) is a convex set for every x 2 X .
Thats why convex preferences are called convex: for every x, the set of all
alternatives preferred to x is convex.
Kreps: NBT (no better than) sets

Microeconomics (UP)

2016-II

30 / 37

Set of Maximizers

Theorem
if % is convex, then for any convex choice set B, the set C (B, %) is
convex.
if % is strictly convex, then for any convex choice set B, the set C (B, %)
is single-valued (or empty).

Microeconomics (UP)

2016-II

31 / 37

Convexity: Utility Functions

The characteristic of utility functions that represent convex preferences is
quasi-concavity.
Denition
A function u : X ! R is quasi-concave if, for every x, y with u (x ) u (y )
and every 2 (0, 1),
u (x + (1

)y )

u (y ).

A function u : X ! R is strictly quasi-concave if, for every x, y with

u (x ) u (y ) and x 6= y and every 2 (0, 1),
u (x + (1

)y ) > u (y ).

u is quasi-concave i, for every r 2 R, the upper contour set

fx 2 x : u (x ) r g is convex.
Microeconomics (UP)

2016-II

32 / 37

Convexity: Utility Functions

Convex preferences are represented by quasi-concave utility functions.

Theorem
Suppose the utility function u represents preferences %. Then:
u is quasi-concave , % is convex
u is strictly quasi-concave , % is strictly convex
Convex preferences get that name because they make upper contour sets
convex.
Quasi-concave utility functions get that name because quasi-concavity is a
weaker property than concavity.

Microeconomics (UP)

2016-II

33 / 37

Separability

Often very useful to restrict ways in which a consumers preferences over

one kind of good can depend on consumption of other goods.
lf allowed arbitrary interdependencies, we would need to observe
consumers entire consumption bundle to infer anything.
Properties of preferences that separation among dierent kinds of goods
are called separability properties.

Microeconomics (UP)

2016-II

34 / 37

Weak Separability: Preferences

Preferences over one kind of goods dont depend on what other goods
are consumed.
Let J1 , . . . Jm be a list of m mutually exclusive subsets of the set of goods.
Let JkC be the complement of Jk .
Given a vector x, let xJk be the vector of those goods in Jk .
Denition
% is weakly separable in J1 , . . . , Jm if, for every k 2 1, . . . , m, every
C
xJk , xJ0 k 2 RjJk j , and every xJ C , xJ0 C 2 RjJk j ,
k

(xJk 0 xJ C ) % (xJ0 k , xJk c ) , (xJk 0 xJ0 C ) % (xJ0 k 0 xJ0 C )

k

Ex. x = {food, clothing, housing}, m = 1, J1 = {food}. ) preferences

separable in food, not separable in clothing or housing.

Microeconomics (UP)

2016-II

35 / 37

Weak Separability: Utility

Theorem
Suppose utility function u represents preferences %, then % is weakly
separable in J1 , . . . , Jm i has utility representation of form
u (x ) = v U1 (xJ1 ), . . . , Um (xJm ), x(J1 [...[Jm )C

Food utility function u1 , total utility is function of (food utility, clothing,

housing).

Microeconomics (UP)

2016-II

36 / 37

Other Kinds of Separability

Strongly separable preferences imply existence of additively separable

utility:
m

u (x ) =

Uk (xJ

).

k =1

No wealth eects in good 1 imply existence of quasi-linear utility:

u (x ) = x1 + v (x2 , . . . , xn ).
Good 1 is called a numerire (or money).

Microeconomics (UP)

2016-II

37 / 37