More is Better

an investigation of monotonicity assumption in economics

Joohyun Shon
August 2008

Abstract
Monotonicity of preference is assumed in the conventional economic
theory to arrive at important conclusions such as “more commodities are
preferred to less.” However, Gary Becker claims that monotonicity is not
necessary for these conclusions in his book Economic Theory [1]. This
paper investigates Becker’s claim.

1 The Conventional Theory1

Economists assume that economic agents (often referred to as the con-
sumer in the context of utility maximization) attempt to maximize utility
(which means satisfaction in vernacular).
Definition 1.1. Let C denote a set of all possible spending arrangements.
Consumption bundles, c, are elements of this set.
Utility, U , can be representd as a function2 from C to R.
Definition 1.2. % is a preference relation that denotes weak preference.
If c1 is weakly preferred to c2 then c1 % c2 . In other words, U (c1 ) ≥ U (c2 ).
 is a preference relation that denotes strong preference. If c1 is
strongly preferred to c2 then c1  c2 . In other words, U (c1 ) > U (c2 ).
Definition 1.3. ∼ is a preference relation that denotes indifference. If
c1 % c2 and c1 - c2 are both true then the consumer is said to be indif-
ferent. In other words, U (c1 ) = U (c2 )
With these definitions, one can now examine the so-called axioms of
choice.
Definition 1.4. Reflexibility For any bundle ci , ci % ci

1 SeeMas-Colell et al. [2] for more complete exposition of microeconomic theory.

2 There are many utility functions that are not continuous. I will only consider the contin-
uous variety in this paper.

1
 Definition 1.5. Completeness Given two bundles, one of the following
is true:
1. c1 % c2
2. c1 - c2
3. c1 ∼ c2
Completeness implies that the consumer can judge between any two
consumption bundles.
Definition 1.6. Transitivity If c1 % c2 and c2 % c3 then c1 % c3
Transitivity makes the completeness axiom applicable to any number
of bundles. The consumer is said to be rational if his or her preference is
transitive and complete.
Definition 1.7. (Strict) Monotonicity3 Suppose there are n commodi-
ties in both c1 and c2 . Monotonicity means that if c1 contains more of
some or all commodities, but no less of any, than c2 (c1 ≥ c2 ) then c1 % c2 .
Strict monotonicity means that if c1 contains more of each commodity
than c2 (c1 > c2 ) then c1  c2
Now let us consider the consequences that arise from monotonicity of
preference.
Definition 1.8. A function U is non-decreasing if c1 ≥ c2 implies U (c1 ) ≥
U (c2 ). A non-increasing utility function is defined similarly.
A function U is strictly increasing if c1 > c2 implies U (c1 ) > U (c2 ). A
strictly decreasing utility function is defined similarly.
Theorem 1.1. Preferences are monotone if and only if U is non-decreasing
and they are strictly monotone if and only if U is strictly increasing.

Proof. First, we prove that the preference relation % can be represented by

a utility function. Then it becomes obvious that preferences are monotone
if and only if U is non-decreasing.
Utility functions are real valued functions defined on C. Therefore for
any c1 , c2 ∈ C, either U (c1 ) ≥ U (c2 ) or U (c1 ) ≤ U (c2 ) is true (1). If a
utility function represents the % then (1) implies that c1 % c2 or that
c1 - c2 (2). (2) is true because of completeness.
We now extend this result to any c ∈ C. Suppose c1 % c2 and c2 % c3 .
If a utility function represents % then U (c1 ) ≥ U (c2 ) and U (c2 ) ≥ U (c3 ).
Therefore, U (c1 ) % U (c3 ) which implies that c1 % c3 (3). (3) is true
because of transitivity.
It has now been proven that the preference relation % can be rep-
resented by a utility function (and also that the consumer needs to be
rational for the said representation to happen).
The second part of the theorem is proven similarly.

It is now clear that the monotonicity assumption implies that the

consumer derives more “pleasure” from consuming more commodities (or
more precisely, the consumer would not be “saddened” by the prospect
of consuming a greater quantity of commodities). Consider constraints
3 A more general assumption of nonsatiation can be made in lieu of the monotonicity

assumption—it is essentially monotonicity defined in the language of limits. There always ex-
ists some bundle preferred to the bundle being consumed which may or may not be affordable.

2
 the consumer would face when he or she attempts to maximize his or her
utility. Expressed mathematically,P the utility maximization problem is:
max U = U (x1 , . . . , xn ) s.t. I ≥ ni=1 pi xi where pi is the price of the
x1 ,...,xn
ith commodity, xi , and I is the consumer’s income4 . If there were no
constraints, the consumer would continue to consume more commodities
because of monotonicity. And because of monotonicity, the solution to the
utility maximization problem ends up at the boundary of the opportunity
set (I = n
P
i=1 pi xi ). In other words, the consumer would fully spend his
or her income (whilst the consumer would still prefer to consume beyond
his or her income).
Definition 1.9. An indifference curve is the collection of all bundles of
same utility; in other words, U (x1 , . . . , xn ) = Ū . Where Ū is some fixed
utility.
The more technical, and therefore less interesting, consequence of
monotonicity assumption is that the indifference curves cannot intersect.
Theorem 1.2. Indifference curves cannot intersect.

Proof. Consider two indifference curves, I1 and I2

Suppose the said indifference curves intersect and let c1 be the bundle
at the intersection, c2 the bundle on I1 and c3 the bundle on I2 .
Since c2 and c3 are on different indifference curves, U (c2 ) > U (c3 ) or
U (c2 ) < U (c3 ) is true. In turn, either c2  c3 or c2 ≺ c3 is true.
At the same time, c2 is on the same indifference curve with c1 and
c3 on the same curve with c1 . That is, c1 ∼ c2 and c1 ∼ c3 . Since the
preference relation is assumed to be monotone c1 ∼ c2 and c1 ∼ c3 cannot
hold simultaneously—contradiction.

This concludes the very brief discussion of conventional economic the-

ory. Armed with the knowledge of conventional theory, we can now inves-
tigate Becker’s claim.

4 The utility maximization problem can therefore be solved using constrained optimization

methods such as the Lagrangian method.

3
 2 Becker’s Claim
In Economic Theory, Gary Becker claims that monotonicity is implied by
the theory. Indeed, he only lists economic rationality as the necessary as-
sumption for the mathematical treatment of utility maximization. Becker
proves by contradiction why more is preferred to less.5
If less of any good were preferred to more, and if the good had a
non-negative price, a consumer would increase his utility by re-
ducing his demand for that good until he either consumed none
of it or preferred more to less. If a consumer were “satiated”—
indifferent between more and less of all goods—and if work
were “irksome,” he would reduce his hours worked and thus
his income until either his income (or at least his earnings)
vanished or he preferred more of some goods to less; he would
then consume only these goods in positive quantities. It is this
sense that utility theory implies that more is preferred to less
of all goods actually consumed.
Essentially Becker is not considering the consumer’s income as given
but as the consequence of his or her labor. Therefore the “Beckerian”
utility maximization problem can be stated as:
max U = U (x1 , . . . , xn )
x1 ,...,xn
Pn
s.t. i=1 pi xi ≤ I = Y + w · t, t + l = T
where Y is non-earned income, w hourly wage, t hours worked, l
leisure (hours spent not working) and T total time endowment.6

In this model, it is clear that the consumer would always operate

on the boundary of his or her opportunity set. If an interior bundle
is preferred then the consumer
Pn would decrease the hours spent working
thereby causing the line i=1 pi xi = I = Y + wt to move towards the
origin until it touches the desired bundle.
Another point needs to be discussed. We saw that indifference curves
do not intersect because of the monotonicity assumption—does the initial
exclusion of monotonicity allow for the possibility of intersecting indiffer-
ence curves?
This is a very important question. If indifference curves can cross
then the utility function does not accurately represent the consumer’s
preference—the consumer cannot compare different bundles of commodi-
ties.
5 Note even if the consumer decides to save some of his or her income, the solution to his

or her utility maximization problem will still be on the boundary if the savings are defined
to be a “commodity.” Also consider the commodities that the sane would actually prefer less
to more (say, deadly poison)—defining a commodity to be “lack” of that commodity would
ensure the boundary solution.
6 Becker’s proof is merely the interpretation of the Beckerian utility maximization problem’s

first order conditions.

4
Theorem 2.1. Indifference curves cannot intersect (even without the
monotonicity assumption).

Proof. Consider two indifference curves, I1 and I2 .

Suppose the said indifference curves intersect and let c1 be the bundle
at the intersection, c2 the bundle on I1 and c3 the bundle on I2 .
Since c2 and c3 are on different indifference curves, U (c2 ) > U (c3 ) or
U (c2 ) < U (c3 ) is true. In turn, either c2  c3 or c2 ≺ c3 is true.
c1 is on the same indifference curve with c2 and with c3 . c1 ∼ c2 and
c1 ∼ c3 are both true and by transitivity c2 ∼ c3 —a contradiction.

3 Conclusion
In the conventional economic theory, the assumption of monotonicity leads
to three implications: (1) “more is better”; (2) the solution to the utility
maximization problem lies on the boundary of the opportunity set; and
(3) indifference curves cannot intersect.
Becker’s formulation of the utility maximization problem excludes
monotonicity assumption. Even though monotonicity is never assumed,
the three implications of the assumption can still be derived.
The Beckerian theory is stronger mathematically as it arrives at the
same conclusion with fewer assumptions. Also a case can be made that
the Beckerian theory also mimics the real world more closely than the
conventional variety since Becker defines income not as a mere constraint
but as a function of work—only those of us who are incredibly lucky can
count income as given.

4 References
[1] Becker, G. S.: Economic Theory, Transaction Publishers, New Brunswick,
N.J., 2008.
[2] Mas-Colell, A., Whinston, M. D., Green, J. R.: Microeconomic
Theory, Oxford University Press, New York, N.Y., 1995.