ECON 520 (2010 Spring) Kansas State University Zijun Luo

Lecture Two: Preference, Utility, and Indifference

Curve

1. Assumptions of Preference
1.1 Completeness

The Axiom of Completeness (referred as “The Axiom of Comparison” in page 70/66): A

person can compare any two baskets of commodities, A and B in such a way that it leads to
one of the three following results: (i) A is preferred to B, (ii) B is preferred to A, or (iii) Both
A and B are the same (They are indifferent).

We have notations for these three cases:

A is preferred to B: A B

B is preferred to A: B A

A and B are the same: A B

This axiom of completeness rules out the possibility that you cannot compare
between two different baskets of commodities. That is to say, it does not permit a person to
say “I just can’t decide,” even though that does sometimes happen.

1.2 Transitivity

The Axiom of Transitivity (page 70/66): Consider any three baskets A, B, and C. If a
consumer prefers A to B, and also prefers B to C, she or he must prefer A to C. Similarly, a
person who is indifferent between A and B, and is also indifferent between B and C, must be
indifferent between A and C.

Violation of the axiom of transitivity, though happen sometimes, is considered as

“irrational” in economics. See paragraph under the box of page 71 in textbook for an
example1 about how someone can be exploited if she/he violates the axiom of transitivity2.

Axiom of Completeness and Axiom of Transitivity are two most basic assumptions
towards people’s preferences. They should always be obeyed when you start an economic

1
This example is not on the 6th edition.
2
Also see “A Discussion of Transitivity” in Lecture one of Lecture Notes in Microeconomic Theory by
Ariel Rubinstein about how and why people would violate the axiom of transitivity occasionally and
unconsciousnessly. You can download the Notes from Prof. Rubinstein’s website for free.

Page 1 of 11
ECON 520 (2010 Spring) Kansas State University Zijun Luo

analysis. The following assumptions, however, although not being formally stated in most
microeconomic textbooks, are still needed when one tries to determine a typical indifference
curve from a “well-behaved” preference.

1.3 The More, The Better

The Axiom of Nonsatiation: Consider two baskets A and B, with different amount of the
same commodities. Suppose A has more than B in every commodity and yet all these
commodities are not economic “bads”, then a consumer will rank A higher than B3.

For economic “goods”, which means these commodities can give you positive
satisfaction, the axiom of nonsatiation simply means “The more, the better.” You may think
of a counter-example that after you have eaten 10 ice-creams, you will not want even a single
one. But don’t forget that in a market you can always trade those additional ice-creams for
money and then purchase other goods. Thus “the more, the better” generally holds4.

1.4 People Prefer Variety

This is straightforward. When I purchase meat in Walmart, I usually buy a certain

combination of pork, beef, chicken, and perhaps some sausages. Formally and mathematically
speaking5, assume any weight t between 0 and 1, if there are two baskets of commodities,

 x1 , y1  and  x2 , y2  such that  x1 , y1   x2 , y2  , we have:

tx  1  t  x , ty  1  t  y   x , y 
1 2 1 2 1 1

You will soon see the application of this in determining a typical indifference curve.

2. Utility and Utility Function

2.1 Preference Ordering and Utility Function

If someone’s behavior satisfied axiom of completeness and axiom of transitivity, we

know that she/he can rank any basket of commodities. Thus we have the following
proposition (page 71/68):

3
A weak version of axiom of nonsatiation states that if for some of the commodities, basket A has
more than basket B while the rests are equal, basket A is preferred to basket B. This weak “axiom of
nonsatiation” is more general.
4
For “economic bads”, the opposite, “the less, the better” holds.
5
See “Well-Behaved Preferences” in Chapter 3 of Intermediate Microeconomics: A Modern Approach
by Hal R. Varian.

Page 2 of 11
ECON 520 (2010 Spring) Kansas State University Zijun Luo

Proposition of Rank Ordering of Preferences: A consumer can consistently rank all baskets
of commodities in order of preference. This ranking is called “the preference function”.

Thus given a certain preference ordering, we will be able to find a utility function (not
a unique one thought) to describe it. Utility function is a mathematical expression that shows
the relationship between utility values and every possible basket of goods. Suppose there are
three commodity baskets A, B, and C, and your preference ordering towards these three
baskets is:

A B C (1)

When we assign a utility function to this preference ordering, it is simply a

transformation from the above expression to:

u  A  u  B   u  C  (2)

Suppose each basket contains the following goods

A: ten apples and ten bananas

B: eight oz of Swiss cheese
C: one intermediate microeconomics textbook
A utility function u    can be defined as “total amount of goods in a given basket”

and we will have

u  A   20

u  B  8

u C   1
satisfied equation (2) which successfully captured preference ordering in equation (1).

2.2 Cardinal and Ordinal

When economists first studied utility in the 19th century, they hoped that individual
preferences could be measured in terms of basic units and could therefore be comparable
between people. Consider the above example with basket A, B, and C. If for Jack,

uJack  A  20

uJack  B   8

u Jack  C   1

Page 3 of 11
ECON 520 (2010 Spring) Kansas State University Zijun Luo

And for Kate6,

uKate  A  10

uKate  B   8

uKate  C   2
For a school of economists who believe in “Cardinal Utility”, they believe the
following statements to be true:

(1) Jack prefers basket A more than Kate does, specifically, Jack prefers basket A twice as
Kate does;
(2) Kate prefers basket C more than Jack does, Kate prefers basket C twice as Jack does;
(3) Jack prefers basket B as same as Kate does;
(4) Jack prefers basket A 20 times as he prefers basket C, he also prefers basket B 8 times as
he prefers basket C;
(5) Kate prefers basket A 5 times as she prefers basket C, she also prefers basket B 4 times as
she prefers basket C;
For another school of economists who believe in “Ordinal Utility”, they believe that
the only thing one can imply from those utility values of Jack and Kate is that the preference
ordering of both of them satisfied equation (1).

There is not yet much solid evidence to support “Cardinal Utility”7, and on the other
hand, for most economic theories, “Ordinal Utility” is enough. Some of the statements you
learned in your principles class, such as the “Equimarginal Rule”, are based on “Cardinal
Utility”. However, in this class, “Ordinal Utility” is our core (if not only) assumption towards
people’s preferences and utility functions.

3. Indifference Curve (Map)

3.1 From Preference to Indifference Curve

Definition of Indifference Curve: Every single indifference curve represents a certain level
of utility. There is infinite number of indifference curves but we usually only draw those
useful ones and call this an “Indifference curve map”.

6
This could be the case when Kate is econ majored, and she considers bananas and apples as perfect
substitutes (simply because they are both fruits).
7
Recent years, there are some researches in neuroscience supports Cardinal Utility. See for example,
Veit Stuphon, “Neuroeconomics: Cardinal Utility in the Orbitofrontal Cortex?”, Current Biology,
16(15), August 2006.

Page 4 of 11
ECON 520 (2010 Spring) Kansas State University Zijun Luo

Characteristics One of Indifference Curve: Negative sloping (page 80/77).

This comes from the axiom of nonsatiation. That is, “the more, the better” for
comparing different baskets of commodities. If we randomly spot a certain point, D, on the
diagram representing two goods, we will see that any combinations of these two goods that lie
in the northeast of this point are better; symmetrically, any combinations lies in the southwest
of this point are less desired. So, if we connect all dots representing the same level of utility as
point A and form a curve, the curve should go from northwest to southeast. I show three
different cases in the following graph.

Better

D
A

Worse

Characteristics Two of Indifference Curve: Indifference curves never intersect/Indifference

curves parallel to each other (page 80/78).

If this characteristic doesn’t hold, it will violate the axiom of transitivity. Suppose
two indifference curves intersect each other at point A. We can find two points, one on each
indifference curve in the way that they have same amount of good X but C has more of good
Y than B. Since A and B are on the same indifference curve, we have A B , and at the same
time, we also have A C . Applying the axiom of transitivity, we can immediately see that
B C should hold. But from the weak version of the axiom of nonsatiation, we also have
C B . Thus a conflict appears. So, indifference curves can never intersect with each other.

Page 5 of 11
ECON 520 (2010 Spring) Kansas State University Zijun Luo

A
C

Characteristics Three of Indifference Curve: There is infinite number of indifference

curves which cover the whole area of the axiom panel (in page 80/78).

Obviously this relates to the axiom of completeness. Since an individual can rank
any of the baskets of goods, and every point on the axiom panel represents one certain basket
of goods, then apparently every point should have one and only one indifference curve gone
through it.

Characteristics Four of Indifference Curve: Indifference curves are convex to the origin
(page 81/79).

This is the result of the fact that people prefer variety. Let’s examine the

mathematical expression in section 1.4. On the graph, point E represents basket  x1 , y1  and

point F represents basket  x2 , y2  . Any point on the dotted line between E and F can be

 
expressed by tx1  1  t  x2 , ty1  1  t  y2 by applying different value of t . Pick a point

between E and F, you always can draw an indifference curve that lies on the northeast of the
dotted line. Followed the more the better, any point on the dotted line that between E and F
has a higher utility than either E or F, and thus points that have the same utility as E or F
(which means they must be on the same indifference curve as E and F) lie on the southwest of
the dotted line. This is called “convex to the origin” in mathematics.

Page 6 of 11
ECON 520 (2010 Spring) Kansas State University Zijun Luo

4. MRS and Diminishing MRS

Marginal Utility, a term that is widely used in economics, is actually a term of
“Cardinal utility” school. The terminology related to “Ordinal utility” that also connected
with MU is called “Marginal Rate of Substitution,” or “MRS” for short.

Definition of Marginal Rate of Substitution: The measurement of how many units of good
Y a consumer would like to give up for getting one additional unit of good X while the
consumer keeps her/his level of satisfaction (utility).

Compared the definition of MRS to the definition of Indifference Curve, we can have
a fifth characteristic for Indifference Curve:

Characteristics Five of Indifference Curve: The absolute value of the slope of the
indifference curve represents Marginal Rate of Substitution.

Mathematically, we can calculate MRS with the help of MU.

MU X
MRS XY 
MU Y

While the slope of indifference curves is always negative, MRS is a positive value.
This is consistent with most textbooks (but oppose to what I did last semester in this same
course).

An important property of MRS is called “Diminishing Marginal Rate of

Substitution”. This captures the fact that while you are getting more and more of a certain
good, say X, you are less likely to give up the other good, Y. For example, if you are very
hungry, you may be willing to pay 4 dollars (good Y) for a hamburger (good X). But when

Page 7 of 11
ECON 520 (2010 Spring) Kansas State University Zijun Luo

you are full, you don’t even want to pay 25 cents for the same hamburger. This property can
be shown graphically:

12
7
5
4

5 6 7 8 X

This graph also shows you one more method to calculate MRS:

Y
MRS XY 
X

For the graph above, X always equals to 1 because I partitioned in such a way that
the consumer is getting only one more unit of X each time. Thus MRS XY can be directly read

from the change in Y (that is, Y ). If the consumer currently has 5 units of X, her/his MRS
to get one more X is 5. If the consumer currently has 6 units of X, her/his MRS to get one
more X is 2. If the consumer currently has 7 units of X, her/his MRS to get one more X is 1.
The property of diminishing MRS is derived from the fact that people prefer variety (hence
indifference curves are convex to the origin).

5. Examples of Utility Function and Indifference Curves

5.1 Perfect Substitution

The utility function for someone towards two perfect substitutes (X and Y) is given
by:

u( X , Y )  aX  bY

That is, for this individual, only the total amount of these two goods (with some
weighted effects) matters. Graphically, the indifference curves look like

Page 8 of 11
ECON 520 (2010 Spring) Kansas State University Zijun Luo

a
b

1 X

The slope of all these indifference curves are the same,  a b . And we don’t observe
diminishing MRS in the case of perfect substitution because MRS is also constant and equals
to a b .

5.2 Perfect complement

The utility function for someone towards two perfect complements is given by:

u ( X , Y )  min{ax, by}

That is to say, for this individual, only the smallest amount of these two goods (with
some weighted effect) matters. A widely quoted example of perfect complement is right foot
shoes and left foot shoes. A hundred right foot shoes and one left foot shoe only makes one
pair of shoes and only works for one normal person. Graphically speaking, the indifference
curves look like:

a
b

1 X

The slope of the indifference curves, as well as MRS, can be calculated according to
the following manner:

Page 9 of 11
ECON 520 (2010 Spring) Kansas State University Zijun Luo

(1) At the turning point of each indifference curve, there is NO slope of MRS;
(2) Above the turning point, the value of slope and MRS are infinity;
(3) On the right of the turning point, the value of slope and MRS are zero.

Mathematically speaking, you cannot calculate the slope of the indifference curve
(thus MRS) at the turning point because at that point the (utility) function is not continuous.

5.3 Cobb-Douglas Utility Function

Charles Cobb was a mathematician at Amherst College, and Paul Douglas was an
economist at the University of Chicago who later became a U.S. senator. In 1928, they
published a paper titled “A Theory of Production” on American Economic Review, proposed
the following function form that can be used to capture the relationship between two different
inputs into the production of outputs:

Output  X aY b

This functional form has later been widely used in nearly every field of economics.
Usually, the exponents a and b satisfied

a0
b0
a b 1

In terms of utility function, a Cobb-Douglas utility function is written as

u  X , Y   X aY b

The indifference curves given by the Cobb-Douglas utility function satisfied all those
four characteristics proposed in section 3 of this lecture. Its wide-used is partly because of its
goodness in fitting the theory, as well as some other mathematical properties that we will
ignore in this class.

5.4 Even More Utility Functions

This following question appears in your Homework One also:

Provide a detailed and specific example in which the indifference curves DO NOT
look like (1) Cobb-Douglas utility function, (2) perfect substitution, and (3) perfect
complement. Please explain and provide a graph in your answer and specify what
goods you are talking about.

Page 10 of 11
ECON 520 (2010 Spring) Kansas State University Zijun Luo

To answer this question, you need to change one or more assumptions towards
people’s preference we have talked about in section 1 of this lecture, hence change one or
more of the characteristics of indifference curves we talked about in section 3. Note that the
axiom of completeness and axiom of transitivity should always be satisfied.

I will gather your answer and send to you (anonymously) as the answer to Homework
One. If I don’t get enough inputs, I will also put some examples from last semester. All
together, you will see some good examples of utility functions other than those we have
already talked over.

Page 11 of 11