Economics 500, Fall 2020

Problem Set 5, Due Thursday, October 7

Question 1 There are many similarities between the theory of the firm
and consumer theory. To illustrate this, let Y be a closed, free-disposal
production set and let π(p) and y(p) be the associated profit and supply
functions. Prove:

1.1 π is homogeneous of degree one.

1.2 π is convex.

1.3 If Y is convex, then Y = {y ∈ RL : py ≤ π(p)∀p >> 0}.

1.4 y is homogeneous of degree zero.

1.5 y is convex if Y is convex, and y is singleton-valued if Y is strictly

convex. It often makes good sense to assume you are dealing with
demand functions, but for this question to be meaningful, you need to
allow the possibility that y(p) might be a correspondence.

1.6 Hotelling’s lemma.

1.7 Dy(p) is symmetric, positive semidefinite, and Dy(p)p = 0.

Now let c(w, q) and z(w, q) be the associated cost and input demand
functions. Prove:

1.8 c is homogeneous of degree one in w.

1.9 c is increasing in q

1.10 If for every q the set {z ≥ 0 : f (z) ≥ q} is convex, then Y = {(−z, q) :

wz ≥ c(w, q)∀w >> 0}.

1.11 z(w, q) is convex if {z ≥ 0 : f (z) ≥ q} is convex, and z(w, q) is a

singleton if {z ≥ 0 : f (z) ≥ q} is strictly convex.

1.12 If z(w, q) is a singleton, then c(w, q) is differentiable in w and dc(w, q)/dwℓ =

zℓ (w, q).
2 c(w, q) is symmetric
1.13 If z is differentiable in w, then Dw z(w, q) = Dw
and negative semidefinite, with Dw z(w, q)w = 0.
1.14 if f is homogenous of degree one, then c and z are homogeneous of
degree one in q. This is the case of constant returns.

1.15 If f is concave, then c is a convex function of q.

Question 2 Here, we explore some of the differences between consumer

theory and the theory of the firm.

2.1 Give an example of utility maximization in which the law of demand

fails. Show that the law of supply holds.

2.2 In consumer theory, substitution effects are negative. State the coun-
terpart of this statement for the theory of the firm, and either show
that it holds or present a counterexample.

2.3 The law of supply can be established via (i) Hotelling’s lemma, (ii)
the implicit function theorem, (iii) monotone comparative statistics,
and (iv) revealed preference arguments. Give each argument, and then
identify precisely the assumptions required for each.

2.4 It is intuitive that an increase in the price of an input should reduce

output and profits. Show that the latter is the case, but present an
example in which the former is not the case.

Question 3 Now we relate our model of profit maximization to ideas that

typically appear in undergraduate microeconomics courses.

3.1 We have three sets of terms that describe similar ideas, namely (i)
increasing, constant and decreasing returns, (ii) decreasing, constant
and increasing costs, and (iii) weakly or strictly concave or convex cost
functions. Establish the relationships between these, i.e., for each set
of terms present arguments or counterexamples showing that they are
equivalent, that one suffices for the other but not the converse, and so
on.

3.2 In intermediate micro, the cost functions typically look something like
C(q) = 436q − 8q 2 + q 3 . In graduate school, it is often said to be
a fundamental property of production sets of that they are convex,
which contradicts the claim that C(q) = 436q − 8q 2 + q 3 could be the
cost function.

2
3.2.1 What can you say about the concavity or convexity of the cost
function C(q) = 436q − 8q 2 + q 3 ? Either show that this cost
function is consistent with a convex production set, or verify that
it is inconsistent.
3.2.2 Draw pictures of the total cost, average cost, and marginal cost
for this cost function, and note that they have the shape so loved
by intermediate texts.
3.2.3 Repeat your previous answer, with the cost function C(q) = 512+
436q − 8q 2 + q 3 , so that we have now added a fixed cost. Add to
your pictures variable cost, fixed cost, average variable cost, and
average fixed costs. This should again be familiar.
3.2.4 Now we see why this type of cost function is convenient. Suppose
we have a perfectly competitive market with a linear (inverse)
demand function, given by p = 10500 − Q, where Q is the total
amount produced by the market. The number of firms in the mar-
ket is n, each with the cost function C(q) = 512 + 436q − 8q 2 + q 3 .
Hence, the market quantity Q will equal nq, where q is the output
per firm. Find the first-order condition for profit maximization
for a firm in this market. Solve for the equilibrium price, quantity-
per-firm, number of firms, and quantity sold in the market. Show
that each firm earns zero profits in this equilibrium. This gives
us an example of the standard description of the equilibrium of a
competitive market—entry and exit causes the number of profit-
maximizing firms to adjust until each firm earns zero profit.
3.2.5 Now let’s see if we can do the same thing with a simpler cost
function corresponding to a convex technology. Supppose each
firm has total cost function given by C(q) = 500q + 10q 2 . We
have dropped the cubic term, but the squared term has changed
sign. Explain precisely why it is problematic to work with a
model of competitive markets based on the cost function C(q) =
500q − 10q 2 .
3.2.6 Having resolved to work with the cost function C(q) = 500q +
10q 2 , first draw the total, average and marginal cost functions
and compare them to those corresponding to the earlier, cubic
cost function. Then solve for the the equilibrium price, quantity-
per-firm, number of firms, and quantity sold in the market. What
are the advantages and disadvantages of this model?
3.2.7 Let’s investigate another alternative. The market demand func-
tion is given by p = 1020−Q, where Q is the total quantity of the

3
 good produced in the market. Each firm has a total cost function
given by  2
q + 100 if q > 0
C(q) =
0 if q = 0.
Once again, first draw pictures of the appropriate cost functions.
Then solve for the equilibrium price, quantity-per-firm, number
of firms, and quantity sold in the market. This cost function
looks like it contains a fixed cost, which may be appropriate for
a short-run analysis, but the typical claim is that there are no
fixed costs in the long run. Can you interpret the terms in this
cost function in such a way as to allay his concern?
3.2.8 It would be even simpler to work with the cost function

q + 100 if q > 0
C(q) =
0 if q = 0.

Investigate this case, and explain its advantages and disadvan-

tages.
3.2.9 Suppose you called to present a tutoral to Congress about per-
fectly competitive markets, and have to choose a running exam-
ple. What would it be? How would you explain it?