University Ca’ Foscari Venice AY 2021-22

MICROECOMICS 1A, 2021-22 - Exercises

QEM Master, Joint degree Erasmus Mundus, PhD

Lecturer: Michele Bernasconi

Version: August, 19th, 2021

Exercises are for class corrections (practical sessions and TA sessions). During class cor-
rections students can be selected randomly to propose their solution to starred (*) exercises.

Double starred (**) exercises are for PhD students. They will be solved in dedicated, addi-
tional, practical sessions.

Exercises: ‘Introduction’ and ‘Consumption’

I.1 What is microeconomics about? And why are you interested in microeconomics?

I.2 Discuss the following thoughts by Gerard Debreu on the mathematization and axiom-
atization of economic theory: “In its mathematical form, economic theory is open to an
efficient scrutiny for logical errors... The greater logical solidity has enabled researchers to
build on the work of their predecessors and to accelerate the cumulative process in which
they are participating... But a Grand Unified Theory will remain out of the reach of eco-
nomics, which will keep appealing to a large collection of individual theories. Each one
of them deals with a certain range of phenomena that it attempts to understand and to
explain. When it acquires an axiomatic form, its explicit assumptions delimit its domain of
applicability and make illegitimate overstepping of its boundary flagrant” (Gerard Debreu,
AER 1991)

C.1 (Ex 1.B.1 in MCWG): Suppose that % is rational. Prove that if x  y % z, then
x  z.

C.2* (Ex 1.B.2 in MCWG): Suppose that % is rational. Prove the following:
a)  is both irreflexive (x  x never holds) and transitive;
b) ∼ is reflexive (x ∼ x for all x), transitive and symmetric (if x ∼ y, then y ∼ x).

C.3* Suppose that the choice structure (B, C(.)) with B = {{x, y}, {x, y, z}} and C({x, y}) =
{x}. Prove that if (B, C(.)) satisfies WARP, then it must hold one of the following choice
rules: C({x, y, z}) = {x} or C({x, y, z}) = {z} or C({x, y, z}) = {x, z}.
University Ca’ Foscari Venice AY 2021-22

C.4 (Ex 1.B.4 in MCWG): Consider a rational preference relation %. Show that if u(x) =
u(y) implies x ∼ y and if u(x) > u(y) implies x  y, then u(·) is a utility function
representing %.

C.5* Suppose that % is rational. Show that if X is finite, then there exists an utility
function u(x) representing %.

C.6* (Lexicographic preferences). For all x = (x1 , x2 ) ∈ R2+ and x̄ = (x̄1 , x̄2 ) ∈ R2+ ,
x % x̄ ⇐⇒ “x1 > x̄1 ” or “x1 = x̄1 and x2 ≥ x̄2 ”.
a) For every x̄ ∈ R2+ , determine and draw the upper contour set U (x̄).
b) Show that every x̄ ∈ R2+ , the indifference set I(x̄) is a singleton.
c) Show that this preference relation is strongly monotone and strictly convex, but not
continuous.

C.7* (Linear preferences). For all x = (x1 , x2 ) ∈ R2+ and x̄ = (x̄1 , x̄2 ) ∈ R2+ ,
x % x̄ ⇐⇒ ax1 + bx2 ≥ ax̄1 + bx̄2
with a > 0 and b > 0.
a) For every x̄ ∈ R2+ , determine and draw the indifference set I(x̄) and the upper contour
set U (x̄).
c) Show that this preference relation is continuous, convex, strongly monotone, but not
strictly convex.

C.8* (Leontief preferences). For all x = (x1 , x2 ) ∈ R2+ and x̄ = (x̄1 , x̄2 ) ∈ R2+ ,
x % x̄ ⇐⇒ min{x1 , x2 } ≥ min{x̄1 , x̄2 }
a) For every x̄ ∈ R2+ ,
determine and draw the indifference set I(x̄) and the upper contour
set U (x̄).
c) Show that this preference relation is continuous, convex, monotone, but it is not
strictly convex and it is not strongly monotone.

C.9 (Ex 3.C.5a in MCWG). Prove the following result: a continuous % is homothetic
if and only if it admits a utility function u(x) that is homogeneous of degree one; i.e.
u(αx) = αu(x) for all α > 0.

C.10* Let p = (p1 , p2 )  0 be a price system and let w > 0 be the wealth of the consumer.
Using the definition of the demand of the consumer, determine graphically the demand of
the consumer in the three following cases:
University Ca’ Foscari Venice AY 2021-22

a) Lexicographic preferences;
b) Linear preferences.
c) Leontief preferences.

C.11* For an utility function u(x1 , x2 ) representing rational preferences in X = R2+ provide
the graphical representations of the following solutions of an utility maximization problem
(UMP) :
a) internal solution x(p, w) which is a global maximizer for u(x1 , x2 ) everywhere quasi-
concave;
b) corner solution x(p, w) which is a global maximizer for u(x1 , x2 ) everywhere quasi-
concave;
c) internal solution x(p, w) which is a global maximizer for u(x1 , x2 ) not everywhere
quasiconcave;
d) internal x(p, w) which is a local but not global maximizer for u(x1 , x2 ).

C.12* Consider a twice continuously differentiable utility functions u(x1 , x2 ) representing

a consumer’s preference. Prove the following results:
a) convexity of preference, that is quasi-concavity of u(x1 , x2 ), implies that at any bundle
∂u(x̄)/∂x1
x̄ = (x̄1 , x̄2 ) the marginal rate of substitution M RS12 (x̄) = is decreasing
∂u(x̄)/∂x2
in x1 .
b) convexity of preference guarantees that the second order conditions for the utility
maximization problem (UMP) are met.

C.13* (Cobb-Douglas utility function). For all x = (x1 , x2 ) ∈ R2+ , the utility function
representing Cobb-Douglas preferences takes the general form u(x1 , x2 ) = (x1 )a (x2 )b with
a > 0 and b > 0
a) For every x̄ ∈ R2+ , determine and draw the indifference sets I(x̄) and the upper
contour set U (x̄) for the following three cases: i) a = b; ii) a > b; and iii) b > a.
b) Determine the following properties of u(x1 , x2 ): continuity, differentiability, strictly
increasing, strictly quasiconcavity.
1
c) Consider the following real-valued function f (z) = (z) a+b . Show that the utility
function defined by v(x1 , x2 ) = f (u(x1 , x2 )) represents the same Cobb-Douglas pref-
erences for all x = (x1 , x2 ) ∈ R2+ ;
d) Consider now the real-valued function g(z) = ln(z). Show that the utility function
defined by ũ(x1 , x2 ) = g(u(x1 , x2 )) also represents the same Cobb-Douglas preferences
for all x = (x1 , x2 ) ∈ R2++ .
University Ca’ Foscari Venice AY 2021-22

e) Let p = (p1 , p2 )  0 be a price system and w > 0 be the wealth of the consumer
with the above Cobb-Douglas preferences. Determine the ordinary demand of this
consumer.
f) Provide a graphical representation of the solution of the previous point in the (x1 , x2 )-
space; assume in the diagram that the price of good x1 changes. Show diagrammati-
cally the offer curve. Assume that wealth changes and show the wealth-consumption
path.

C.14* As usual, let x(p1 , p2 , w) = (x1 (p1 , p2 , w), x2 (p1 , p2 , w)) denote the demand of the
consumer. For every commodity l = 1, 2, the demand of commodity l is given by
w
xl (p1 , p2 , w) =
p1 + p2
a) Prove that this demand is homogeneous of degree zero.
b) Prove that this demand satisfies Walras’ Law.
c) State the Weak Axiom of Revealed Preferences (WARP) in the framework of the
demand.
d) Prove that this demand satisfies WARP.

C.15* Let L be the number of commodities. Let (−∞, ∞) × RL−1 + be the consumption set.
The consumer has strictly convex preferences which are represented by a utility function
u(x) = x1 + φ(x2 , x3 , ..., xL ). We assume p  0, and we normalize p1 = 1.

a) Show that the demand for commodities {2, 3, ..., L} must be independent of wealth.
How does demand for commodity 1 react to changes in wealth w?
b) Using the previous result, define the indirect utility function as usual, i.e. v(p, w) := u(x∗ ),
where x∗ belongs to the demand, given p and w. Show that v(p, w) is linear in wealth:
v(p, w) = w + ψ(p) for some function ψ : RL ++ → R.
c) Now let L = 2 and φ(x2 ) = α ln(x2 ). Solve the UMP as a function of (p, w) (Recall
that we allow demand for commodity 1 to be negative).

C.16 (Ex 3.C.6 in MCWG). Suppose that in a two-commodity world, the consumer’s utility
function takes the form u(x) = [α1 xρ1 + α2 xρ2 ]1/ρ . This utility function is known as constant
elasticity of substitution (CES) utility function.

a) Show that when ρ = 1, indifference curves become linear.

b) Show that as ρ → 0, this utility function comes to represent the same preferences as
the (generalized) Cobb-Douglas utility function u(x) = xα1 1 xα2 2 .
University Ca’ Foscari Venice AY 2021-22

c) Show that as ρ → −∞, indifference curves become right angles; that is, this utility
function has in the limit the indifference map of the perfect complement - Leontief
utility function u(x) = min{x1 , x2 }.

C.17 (Ex 2.E.1 in MCWG). Suppose L = 3 and consider the demand function x(p, w)
defined by:
p2 w
x1 (p, w) = ,
p1 + p2 + p3 p1
p3 w
x2 (p, w) = ,
p1 + p2 + p3 p2
βp1 w
x3 (p, w) = ,
p1 + p2 + p3 p3

Does this demand function satisfy homogeneity of degree zero and Walras’ law when β = 1?
What about when β ∈ (0, 1)?

C.18* (Ex 2.E.4 in MCWG): Show that if x(p, w) is homogeneous of degree one with
respect to w and satisfies Walras’ law, then the elasticity of a generic commodity l to
wealth is εlw (p, w) = 1. Interpret. Can you say something about the form of the Engel
functions and curves in this case?

C.19 (Ex 2.E.7 in MCWG): A consumer in a two-good economy has a demand function
x(p, w) that satisfies Walras’ law. His demand function for the first good is x1 (p, w) =
αw/p1 . Derive his demand function for the second good. Is his demand function homoge-
neous of degree 0?

C.20 (Ex 2.D.1 in MCWG). A consumer living for two periods consumes a single consump-
tion good denoted c1 and c2 in period 1 and period 2, respectively. His wealth in period
1 is w1 > 0 and in period 2 is w2 > 0. There are perfect capital markets so that wealth
can be transferred between the two periods at a constant interest rate r. Prices in the two
periods are p1 = p2 = 1.
a) What is the consumer’s Walrasian (lifetime) budget set?
b) Provide a graphical representation of the budget set.

C.21*. A consumer consumes one consumption good x and hours of leisure R. The time
endowment is T while the consumer has no exogenous wealth endowment. The price of the
consumption good is p and the consumer can work at a hourly wage rate of s.
University Ca’ Foscari Venice AY 2021-22

a) What is the consumer’s Walrasian budget set?

b) Assuming Cobb-Douglas preferences u(x, R) = xα R1−α , with α > 0, compute the
consumer’s labour supply which is given by the time endowment minus the optimal
hours of leisure R (solution of the UMP).
c) Provide a graphical representation of the previous solution.
d) Suppose now that the salary s increases. Compute analytically and show diagram-
matically the change in the labour supply. What can you say about the substitution
effect and the income effect generated by this salary change?
e) Suppose now that, in addition to the time endowment T , the consumer has also an
exogenous wealth endowment E. Repeat all steps a)-d) above.

C.22** (Ex 2.E.2 in MCWG): Show that the equations expressing the Cournot aggregation
and the Engel aggregation lead to the following two elasticity formulas:
L
X
bl (p, w)εlk (p, w) + bk (p, w) = 0, all k = 1, ..., L
l=1
and
L
X
bl (p, w)εlw (p, w) = 1,
l=1

where bl (p, w) = pl xl (p, w)/w is the budget share of the consumer’s expenditure on good l
given price p and wealth w.

C.23**. Consider a UMP in L = 3 with prices p ∈ R3++ and wealth w > 0. The agent’s
preferences  over bundles (x1 , x2 , x3 ) in R3+ can be represented by the utility function:
u(x1 , x2 , x3 ) = x1 + x2 · x3
a) Given the definition of convex, strictly convex, and homothetic preferences,. Verify
whether or not preferences in equation (1) are: (a.1) homothetic; (a.2) convex; (a.3)
strictly convex.
b) Let the price of commodity x1 be p1 = 1. Compute the optimal consumption bundle
which maximizes the consumer’s utility.
c) Compute now the ordinary demands of the three goods. Do the ordinary demand
functions satisfy: (c.i) the Walras’ law? (c.ii) the so called property of no-money
illusion?
d) Compute the Slutsky substitution matrix and verify that is symmetric.

C.24**. In a two-commodity world and for utility functions: A) u(x) = ax1 + bx2 , B)
u(x) = Min{ax1 , bx2 }, C) u(x) = xa1 xb2 (with a > 0 and b > 0 in all the three cases). Sovle
for the following:
University Ca’ Foscari Venice AY 2021-22

a) the compensated demand functions;

b) the expenditure function and verify that the derivative of the expenditure function
with respect to price of good l delivers the compensated demand of good l;
c) using duality, compute the ordinary demand function from the compensated demand
functions and the indirect utility function from the expenditure function.

C.25** (Ex 3.G.6 in MWG). A consumer in a three-goods economy (goods denoted x1 ,

x2 , x3 and prices p1 , p2 , p3 ) with wealth w > 0 has demand functions for commodities 1
and 2 given by:
p1 p2 w
x1 (p, w) = 100 − 5 +β +δ ,
p3 p3 p3
p1 p2 w
x2 (p, w) = α + β + γ + δ ,
p3 p3 p3
where Greek letters are nonzero constant.

a) Indicate how to calculate the demand for good 3 (but do not actually do it).
b) Are the demand functions for x1 and x2 appropriately homogeneous?
c) Calculate the restrictions on the numerical values α, β, γ and δ implied by utility
maximization.
d) Given your results in part (c), for a fixed level of x3 draw the consumer’s indifference
curves in the (x1 ,x2 ) plane.
e) What does your answer to (d) imply about the form of the consumer’s utility function
u(x1 , x2 , x3 )?
University Ca’ Foscari Venice AY 2021-22

Exercises on ‘Production’

P.1* The basic properties of the production set Y are the following ones:

• Possibility of inaction
• Closedness
• Impossibility of free production (“no free lunch”)
• Free-disposal
• Irreversibility
• Convexity
• Increasing/decreasing/constant returns to scale.

Let L = 2 be the number of commodities. A firm produces commodity 2 using commodity

1 as an input. The production function is f (z) = αz with α > 0 and z ≥ 0.

a) Determine, both formally and graphically, the production set Y that corresponds to
the production function f .
b) Determine if the production Y verifies the basic properties.

Now answer questions a) and b) for the two alternative production functions below:
√
• f (z) = α z with α > 0 and z ≥ 0.
• f (z) = αz 2 + βz with α > 0, β > 0 and z ≥ 0.

P.2* For a general single output technology, show that the production set Y is convex if
and only if the production function q = f (z) is concave.

P.3* Let L be the finite number of commodities. A firm produces commodity L using
the other L − 1 commodities as inputs. z = (z1 , ..., zL , ..., zL−1 ) ∈ RL−1
+ denotes a generic
bundle of inputs. Show that if the production function f : RL−1 + → R + is concave, then the
transformation function defined by
F (y) := yL − f (z)
is quasi-convex on the convex set A = {y = (−z, yL ) ∈ RL
+ : z ≥ 0 and yL ≥ 0}.

P.4* Let L = 3 be the number of commodities. The firm produces commodity 3 using
commodities 1 and 2 as inputs. The production function is given by f (z1 , z2 ) = (z1 )α (z2 )β
with α > 0, β > 0, z1 ≥ 0 and z2 ≥ 0.

a) Write the production set Y determined by the production function f .

University Ca’ Foscari Venice AY 2021-22

b) Determine if the production Y verifies the basic properties (same as in EX1).

c) Discuss under which conditions of the scalars α and β the technology exhibits de-
creasing, increasing or constant return to scale.

P.5* (Ex 5.C.1 in MCWG). Show that, in general, if the production set Y exhibits non-
decreasing returns to scale, then either π(p) ≤ 0 or π(p) = +∞.

P.6* Let L = 2 be the number of commodities. The firm produces commodity 2 using
commodity 1 as an input. The production function is f (z) = αz with α > 0 and z ≥ 0.
a) Write the profit maximization problem of this firm.
b) Consider the production set Y determined by the production function f . Using the
shape of Y and the iso-profit lines, determine graphically the supply of this firm.
c) Determine the profit function of this firm.

P.7* Let L be the finite number of commodities. Assume that the production set Y of the
firm is represented by a transformation function F such that Y = {y ∈ RL : F (y) ≤ 0}.
a) State the profit maximization problem (PMP) of the firm.
b) Let F be continuous and strictly quasi-convex. Show that if PMP has a solution for
p  0, then it must be unique.

P.8*. Let L = 2 be the number of commodities. The firm produces commodity 2 using
√
commodity z as an input. The production function is given by f (z) = α x with α > 0
and z ≥ 0.
a) Write the transformation function and the profit maximization problem (PMP) of
this firm.
b) Show that if ȳ = (ȳ1 , ȳ2 ) belongs to the supply of the firm, then ȳ1 < 0 and ȳ2 > 0.
c) Consider the open and convex set A = {y = (−z, y2 ) ∈ R2 : z > 0 and y2 > 0}.
Write the first order conditions associated with (PMP) on the set A.
d) Compute the supply and the profit function of this firm.

P.9* (Ex 5.C.9 in MCWG). Derive the profit function π(p) and supply function (or corre-
spondence) y(p) for the single-output technologies with production functions given by: (a)
√
f (z) = z1 + z2 ; (b) f (z) = M in{z1 , z2 }; (c) f (z) = (z1ρ + z2ρ )1/ρ .
p

P.10* Let L be the number of commodities. A firm uses a single-output technology to

produce output q using (L − 1) commodities as inputs. Let z := (z1 , ..., zl , ..., zL−1 ) ∈ RL−1
denote a generic bundle of inputs.
University Ca’ Foscari Venice AY 2021-22

b) Define the cost function c(w, q) of the firm.

b) Show that the cost function c(w, q) is a concave function of the input price vector
w := (w1 , ..., wl , ..., wL−1 ).
c) Show that if the production function f : RL−1
+ → R+ is concave, then the cost function
c(w, q) is a convex function of the output level q.

P.11 A firm has the single-output technology, f (z) = z1a + z2 , with 0 < a < 1, and where
p is the price of output and w1 and w2 are the prices of inputs.
a) Determine the conditional input demands, the profit and the supply function of the
firm;
b) How does you answer to point a) change when a > 1?

P.12*. Let L = 2 be the number of commodities. The firm produces commodity 2 using
commodity 1 as an input. The production function is f (z) = α(1 − exp(−kz)) with k > 0,
α > 0 and z ≥ 0.
a) Write and draw the production set Y determined by the production function f .
b) For every level of output y2 ≥ 0, determine and draw the following set Y (y2 ) := {z ∈
R2 : z ≥ 0 and f (z) ≥ y2 }.
c) Write the cost minimization problem of this firm.
d) Determine the conditional demand of inputs and the cost function of the firm.
Now answer questions a) to d) for the two alternative production functions below:
√
• f (z) = α z with α > 0 and z ≥ 0.
• f (z) = αz 2 + βz with α > 0, β > 0 and z ≥ 0.

P.13* A firm uses a single-output technology to produce output q using two commodities as
2 1
inputs with prices w = (w1 , w2 ). The cost function is given by c(w, q) = 2(q)2 (w1 ) 3 (w2 ) 3 .
a) Show that the cost function is homogeneous of degree one in the inputs prices w.
b) Verify that this cost function is a convex function of the output level q.
c) Compute the supply and the profit function of the firm.

P.14* Let L = 3 be the number of commodities. The firm produces commodity q using
commodities 1 and 2 as inputs. The production function is given by
f (z1 , z2 ) = (z1 )α (z2 )β with α > 0, β > 0, z1 ≥ 0 and z2 ≥ 0

a) Determine the conditional input demands.

University Ca’ Foscari Venice AY 2021-22

b) Determine the cost function of the firm and verify that the derivatives of the cost
function with respect to price of input l = 1, 2 deliver the conditional input demands.
c) Discuss the conditions under which the technology admits a strictly positive, unique
and finite level of output maximizing the firm’s profit. For the case in which it is
admitted, determine the supply and the profit function of the firm.

P.15* (Ex 5.D.1 in MCWG). Let AC(·) and C 0 (·) denote a firm’s average cost function
and marginal cost function, respectively. Show that AC(q) = C 0 (q) at any q satisfying
AC(q) ≤ AC(q) for all q. Does this result depend on differentiability of C()˙ everywhere?

P.16 For the single-output technology, with two inputs, q = (M in{z1 ; 2z2 })λ , with λ > 0:

a) Discuss under which conditions of λ the technology exhibits decreasing, increasing or

constant return to scale.
b) Compute the cost function.
c) Derive the profit maximizing output for the firm for the different values of λ.
d) Suppose now that factor z2 is fixed in the short run. Derive the cost function in the
short run.
e) Provide a diagrammatic representation of the cost curves in the short run and use
the diagram to discuss the profit maximizing output of the firm in the short run,
taking care to consider the conditions under which the firm decides to operate in the
market.
Micro IB - Exercises with (partial) solutions1
QEM Master Program - Ca’ Foscari University of Venice

Lecturer: Pietro Dindo

Version: 200821

Note: I expect all students to be confident with the solution of unmarked questions and
to be able to solve the questions marked with ? with a bit more effort. Questions marked
with ?? are more involved. I encourage to answer these ?? questions after you have covered
all the others.

1 Edgeworth Box

Exercise 1
Consider an economy with two consumption goods, L = {1, 2}, and two consumers
I = {a, b}. Consumer a has endowment ωa = (3, 0) while consumer b has endowment
ω2 = (1, 3). Both consumers have utility U (x) = (x1 )2 (x2 )3 .

1. Find the set of Pareto optimal allocations.

2. Determine the subset of Pareto optimal allocations where both consumers are at least
as well-off as with their endowment (the contract curve).

3. Find the competitive equilibria. Are they Pareto optimal?

Exercise 2
Consider an economy with two consumers, I = {1, 2}, and two commodities, L = {1, 2}.
√
Consumer 1 has utility U1 (x1 ) = x11 + x21 and endowment ω1 = (2, 0). Consumer 2
values only good 2, U2 (x2 ) = x22 , and has endowment ω2 = (0, 2).

1.? Find the set of Pareto optimal allocations (a graphical solution is enough).

2.? Find the competitive equilibria (a graphical solution is enough).

3.?? Repeat the same analysis (both Pareto set and competitive equilibria) with the en-
dowments ω1 = (2, 1) and ω1 = (0, 1).
1
Exercises have been gathered from various sources: exercise sessions of the QEM Micro1B course,
Ca’ Foscari University of Venice; the textbook “Microeconomic Theory” by Mas-Colell, Whinston, and
Green [MWG]; past Microeconomics I exams of the QEM Joint Degree Program; Class Notes written by
Jean-Marc Bonisseau and Elena del Mercato for the QEM1 Microeconomics I course, Université Paris 1.
Note: This exercise is inspired by Figure 15.B.10(a) of the MWG textbook, you are en-
couraged to have a look at it.

Exercise 3
Consider an economy with two consumers, I = {1, 2}, and two commodities, L = {1, 2}.
√
Consumer 1 has utility U1 (x1 ) = x11 + x21 . Consumer 2 values only good 1, U2 (x2 ) = x12 .
The total endowment is ω̄.

1.? Find the set of Pareto optimal allocations (a graphical solution is enough).

2.? Find the competitive equilibria when ω21 = 0 (a graphical solution is enough).

3.? Find the competitive equilibria when ω12 = 0 (a graphical solution is enough).

Note: This exercise is inspired by Figure 15.B.10(a) of the MWG textbook, you are en-
couraged to have a look at it.

Exercise 4
Consider an economy with two consumers, I = {1, 2}, and two commodities, L = {1, 2}.
The endowments are ω1 and ω2 . Assume that both consumers have locally non satiated
preferences. Prove that if the market for good l clears and p  0, then also the market for
good l0 6= l clears.

Exercise 5
Consider an economy with two consumers, I = {1, 2}, and two commodities, L = {1, 2}.
The total endowment is ω̄ = (4, 3). Find the set of Pareto optimal allocations in the
following cases:
√ √ 1 2
1. U1 (x1 ) = x11 x21 and U2 (x2 ) = (x12 ) 3 (x22 ) 3 ;

2. U1 (x1 ) = x11 + x21 and U2 (x2 ) = 2x12 + x22 (a graphical solution is enough);
√ √
3.? U1 (x1 ) = x11 x21 and U2 (x2 ) = x12 + 2x22 .

Exercise 6
Consider an economy with two consumption goods, L = {1, 2}, and two consumers
I = {1, 2}. Consumer 1 has endowment ω1 = (1, 2) while consumer 2 has endowment
ω2 = (2, 1). Consumer 1 has utility U1 (x1 ) = x11 (x21 )2 while 2 has utility U2 (x2 ) = x12 x22 .

1. Find the set of Pareto optimal allocations.

2. Determine the subset of Pareto optimal allocations where both consumers are at least
as well-off as with their endowment (the contract curve).
 3. Find whether the allocation on the Pareto set with x11 = 2 can be supported as a
competitive equilibrium with transfers. If so, for which transfer using only good 1 ?
Is there a transfer of only good 2 that achieves the same allocation?

Exercise 7 (JD-QEM ’20-’21)

We consider an exchange economy with two consumers and two goods.

• Consumer 1 has consumption set R2+ , initial endowment e1 = (1, 1) and utility
1/3
u1 (x11 , x12 ) = x11 (x12 ) /3
2

• Consumer 1 has consumption set R2+ , initial endowment e2 = (1, 3) and utility
1/2 1/2
u2 (x21 , x22 ) = x21 x22

1. Represent in the Edgeworth box the initial endowment and the indifference curves
going through the initial endowment for both agents.

2. Represent in the Edgeworth box the set B of allocations that are better, in the sense
of Pareto, than the initial endowments.

3. Give the definition of a Pareto Optimum for this economy.

4. Determine the set C of Pareto Optima of the economy.

5. Represent (approximately) the set C in the Edgeworth box. Do B and C coincide ?

Explain why/not ?

6. Give the definition of a general equilibrium for this economy.

7. Determine the general equilibrium of this economy (normalize to 1 the price of com-
modity 2).

Exercise 8
Solve Example 15.B.2 from the MWG textbook.

Exercise 9
Solve Exercise 15.B.2 from the MWG textbook.

Exercise 10
Solve Exercise 15.B.9 from the MWG textbook.
2 Robinson Crusoe Economy

Exercise 11
Solve Exercise 15.C.2 from the MWG textbook.

Exercise 12
Consider an economy with two periods, t = 0, 1, and a consumption good per period.
The consumer has utility U (x0 , x1 ) = u(x0 ) + βu(x1 ) with β ∈ (0, 1) and endowment k0
of consumption good in t = 0. The economy has one firm, J = 1, which transforms the
good in t = 0 (input, k1 ), into an output (y) in t = 1. The firm production function is
f (k) = Ak α with A > 0 and α ∈ (0, 1]. Assume that u(x) = log x and α ∈ (0, 1).
1. Find the competitive equilibrium.
2. Find the firm profit in the competitive equilibrium and discuss its dependence on α.
3. Find the Pareto optimal allocation of this economy and compare it with the allocation
found in 1.

Exercise 13
Solve Exercise 12 when
1.? u(x) = x and α ∈ (0, 1);
2. u(x) = log x and α = 1;
3.?? u(x) = x and α = 1.

Exercise 14
Consider an economy with two consumption goods, L = 2, one consumer, I = 1, and one
firm, J = 1. The consumer has endowment ω = (4, 1) and utility U1 (x) = x1 x2 . The firm
has production set Y1 = {y2 ≤ −ay1 , y1 ≤ 0} with a > 0. The consumer is the owner of
the firm.
1. Write down the definition of a competitive equilibrium for this economy.
2. Fix a = 1, find the set of competitive equilibria and the corresponding firm profits.
3. Are the the competitive equilibria found above Pareto optimal?
4. Find the Pareto optimal allocations for a = 1.
5.? Answer to 2. and 3. for each given a > 0.
6.? Find the Pareto optimal allocations for each given a > 0.
3 Competitive equilibrium and welfare theorems

Exercise 15
Consider an economy with L commodities, I consumers, and J firms. The endowments

I
and ownership shares are ωi , (θji )Jj=1 i=1 . Assume that all consumers have locally non
satiated preferences. Prove that, for all choices of a good l0 , if the markets for all goods
l 6= l0 clears and p  0, then also the market for good l0 clears.

Exercise 16
Consider an exchange economy with two consumption goods, L = 2, and two consumers,
I = 2. Consumer 1 has endowment ω1 = (2, 0) and utility U1 (x1 ) = (x11 )α + (x21 )α .
Consumer 1, has endowment ω2 = (1, 3) and utility U2 (x2 ) = x12 x22 . Consider α = 21 first.
1. Find the competitive equilibria and the set of Pareto optimal allocations.

2. Discuss the validity of both Welfare Theorems.

3. The social planner has the possibility to implement some transfer between the initial
endowments of commodity 2. Is there a transfer such that, in the competitive equi-
librium achieved after trading, both agents consume the same amount of good 1. If
so, find it. Can the same competitive equilibrium arise after a transfer of good 1? If
so, find it.

4. Repeat the exercise taking α = 1.

Exercise 17
Consider an economy with two consumption goods, L = 2, two consumers, I = 2, and
one firm, J = 1. Consumer 1 has endowment ω1 = (4, 1) and utility U1 (x1 ) = x11 x21 .
Consumer 2 has endowment ω2 = (1, 1) and utility U (x2 ) = x12 + x22 . The firm has
√
production set Y = {y2 ≤ −y1 , y1 ≤ 0}. Consumer two is the owner of the firm.
1. Write down the definition of a competitive equilibrium in this economy.

2. Find the set of competitive equilibria and the corresponding firm profits.

3. Are the competitive equilibria allocation found above Pareto optimal?

4.?? Find the set of Pareto optimal allocation of this economy.

Exercise 18
Consider three commodities, time for leisure/labor (commodity 1), durable good (com-
modity 2) and consumption good (commodity 3). There is a single consumer with utility
function U (x2 , x3 ) = x2 + x3 , so that the consumption of leisure does not have any impact
on his utility. The consumer has an endowment of leisure equal to 1 and no endowment of
durable and consumption good.

There are two firms. Firm 1 produces only durable goods using labor as input. Firm 1
√
production set is Y1 = {− −y11 + y21 ≤ 0, y11 ≤ 0}. Firm 2 produces only consumption
√
good using labor as input. Firm 2 production set is Y2 = {−2 −y12 + y32 ≤ 0, y12 ≤ 0}.
The consumer is the owner of both firms.

1. Write down the definition of a competitive equilibrium in this economy.

2. Check that
( √      )
p1 5 p3 1 4 1 1 4 4
= , = 1, x = 0, √ , √ , y1 = − , √ , y2 = − , √
p2 2 p2 5 5 5 5 5 5

is the only competitive equilibrium of this economy.

3. Is the competitive equilibria allocation found above Pareto optimal?

4.? Find the set of Pareto optimal allocation of this economy.

Exercise 19
Find the competitive equilibrium and the Pareto set in an economy as above with
U (x2 , x3 ) = x2 x3 and Y2 = {y12 + y32 ≤ 0, y12 ≤ 0} (firm 1 keeps the same technol-
ogy).

Exercise 20
Consider an economy with two consumption goods, L = 2, two consumers, I = 2,
and one firm, J = 1. Consumer 1, the worker, has endowment ω1 = (1, 0) and utility
U1 (x1 ) = x11 x21 . Consumer 2, the capitalist, is the owner of the firm, θ21 = 1, and has
√
utility U2 (x1 ) = x12 x22 . The firm has production set Y = {y1 ≤ 2 −y2 ≤ 0, y1 < 0}.

1. Provide the definition of a competitive equilibrium for this economy.

2. Find the set of competitive equilibria and the set of Pareto optimal allocations.

3. Discuss the validity of both Welfare Theorems.

4. Find the competitive equilibrium with transfer such that both consumers consume
the same amount of output. How can you implement such a transfer starting from
the given endowment/ownership?
Exercise 21
Consider an exchange economy with two consumption goods, L = 2, and two con-
sumers, I = 2. Consumer 1 has endowment ω1 = (2, 2 − δ) with δ ∈ (0, 2) and util-
ity U (x1 ) = x11 + x21 for  > 0. Consumer 2 has endowment ω2 = (2, δ) and utility
U (x2 ) = x12 + x22 .

1. Consider  = δ = 0. Find the set of competitive equilibria and the set of Pareto
optimal allocations. Discuss the validity of both Welfare Theorems in this economy.

2. Consider  = δ = 0. State whether p = (0, 1), p = (1, 1), p = (1, 2), p(2, 1) support
x∗1 = (0, 2), x∗2 = (4, 0) as a quasi-equilibrium transfer. If so, is each {p, x∗ } also a
competitive equilibrium?

3. Consider  > 0 and δ = 0. Find and plot the excess demand z1 (p) = x11 (p) +
x12 (p) − ω̄1 of good 1, the set of competitive equilibria, and the set of Pareto optimal
allocations. Discuss the validity of both Welfare Theorems in this economy.

4. Consider  = 0 and δ ∈ (0, 2). Find and plot the excess demand function of good 1,
the set of competitive equilibria, and the set of Pareto optimal allocations. Discuss
the validity of both Welfare Theorems in this economy.

Exercise 22
Consider an economy with two consumption goods, L = 2, one consumer, I = 1, and
one firm, J = 1. Consumer 1 has endowment ω1 = (2, 1) and utility U (x1 ) = x11 x21 . The
firm has production set Y = {y1 + y2 ≤ 0 if y1 ≤ −a, y2 ≤ 0 if y1 ∈ (−a, 0]} with a ≥ 0.
Consumer one is the owner of the firm. Fix a = 1.

1. Provide the definition of a competitive equilibrium for this economy.

2. Find the set of competitive equilibria.

3. Find the set of Pareto optimal allocations.

4. Discuss the validity of both Welfare Theorems.

Exercise 23
Solve Ex. 23 for each given a ∈ [0, 1].

Exercise 24
Consider an exchange economy with I = {1, 2}, L = {1, 2}, U1 (x1 ) = x21 , U2 (x2 ) =
x12 + x22 , and ω1 = (1, 0), ω2 = (0, 1).
 1.? Find the utility possibility set

2.? Find the Pareto frontier

3.?? Show that for some levels of Ū2 the solutions x∗ of

Max U1 (x1 )
x ∈ R4+
x1 + x2 ≤ ω̄
U2 (x2 ) ≥ Ū2

are not Pareto optimal (and thus (U1 (x∗1 ), U (x∗2 )) does not belong to the Pareto
frontier).

Exercise 25 (Economics-QEM exam ’20-’21 I)

Consider an economy with two commodities, two consumers, and one firm. Consumer
1 has endowment ω1 = (6, 0), owns half of the firm, and has utility U1 (x1 ) = x11 x21 .
Consumer 2 has endowment ω2 = (4, 2), owns the other half of the firm, and has utility
U2 (x2 ) = x12 x22 . The firm uses commodity 1 to produce commodity 2 and has production
function f (z) = z with z ≥ 0 (the amount of commodity 1).

1. Derive supply and profit of the firm.

2. Derive the demand of both consumers.

3. Provide the definition of a competitive equilibrium for this economy.

4. Compute the unique competitive equilibrium of this economy.

5. Is the competitive equilibrium found above Pareto optimal? Reply without comput-
ing the Pareto set.

6. Provide a definition of Pareto optimal allocation.

7. Explain how to find all the Pareto optimal allocation of this economy (write down the
related maximization problem and the first order conditions for an interior solution).
4 Expected Utility

Exercise 26
Solve Exercise 6.B.2 in the MWG textbook.

Exercise 27
Show that a preference relation on lotteries that can be represented in an Expected Utility
form satisfies transitivity.

Exercise 28
Consider the space of simple lotteries L = {C; (p1 , p2 , p3 )}. Show graphically that if a pref-
erence relation on lotteries satisfies the independence axiom, then indifference curves are
parallel. (hint: show that if two indifference curves are not parallel, then the independence
axiom is violated)

Exercise 29
Solve Exercise 6.B.4 in the MWG textbook.

Exercise 30
Solve Exercise 6.B.7 in the MWG textbook. (Note: a preference over lotteries is monotone
if, given two real outcomes C1 > C2 , then the lottery with sure outcome C1 is strictly
preferred to the lottery with sure outcome C2 )

Exercise 31
When faced with the choice between the lottery A = {(3500, 2800, 0); p = (0.3, 0.66, 0.04)}
and B = {3500; p = 1}, a decision maker chooses lottery B. When, instead, he is asked
to choose between the lottery A0 = {(3500, 2800, 0); p = (0.3, 0, 0.7)} and lottery B 0 =
{(3500, 2800, 0); p = (0, 0.34, 0.66)}, he chooses A0 . Say whether the decision maker’s
preferences are consistent with the expected utility form, explaining how you have reached
your conclusion. What if lottery B becomes B 00 = {2800; 1}?

Exercise 32 (JD-QEM exam ’19-’20)

We consider an expected-utility decision-maker facing the following possible professional
occupations: working in the financial industry (A), working in the movie industry (B),
working in the car industry (C). We further assume that the decision-maker can apply to
three different schools.
• After School 1, he is certain to find a job in the car industry.

• After School 2, he is certain to find a job in the financial industry.

 • After School 3, he would find a job in the movie industry with probability 0.1 and in
the financial industry with probability 0.9.
We assume that School 2 is the least preferred option of the decision-maker and that he is
indifferent between School 1 and School 3.
1. Give a representation of the utility function of this decision-maker.

2. We assume that a new school (School 4) opens. After school 4, a student would find
a job in each industry with probability 13 . How would the new school rank compared
to Schools 1, 2, 3?

3. There are too many application in School 4. The ministry of education decides that
rather than applying to School 4 directly, the decision-maker must apply to a lottery
that leads to admission in School 4 with probability α ∈ [0, 1] and to admission in
School 2 with probability 1 − α. For which value of α does the decision-maker prefer
to apply to School 1 rather than to the lottery that may lead to admission to School
4?

Exercise 33 (JD-QEM exam ’20-’21)

We consider an expected-utility decision-maker with utility of the form u(x) = xa with
a < 1.
1. Let X be a lottery whose outcome is uniformly distributed over [0, 1]. Determine
E(u(X)).

2. Let Y be a lottery whose outcome is 0 with probability 1/3 and 1 with probability
2/3. Determine E(u(Y )).

3. Determine the value a∗ (of a) for which the decision-maker is indifferent between X
and Y .

4. If a > a∗ , which lottery, X or Y, is preferred by the decision-maker?

5. Determine, as a function of a, the coefficient of absolute risk-aversion of the decision-

maker.

Exercise 34
Solve Exercise 6.C.1 in MWG assuming u(x) = log(x) and find? for which value of q the
agent does not insure.

Exercise 35
Consider an investor who has to choose between two assets. Asset A has payoff xA =
(12, 6, 9), all with equal probability. Asset B is risk free with payoff xB = (x̄, x̄, x̄).
 1. Assume u(x) = log(x). Find the certainty equivalent of xB .

2. Find the level of x̄ such that the investor is indifferent between asset A and asset B.

3. Assume now that x̄ = E[xA ], find how many units hA of asset A should be given
to the agent to make him indifferent between hA and one unit of the risk free asset.
Provide an intuition.

4. Compute the coefficient and absolute risk aversion.

Exercise 36 (Economics-QEM exam ’20’21 I)

A decision maker has preferences over lotteries represented by an expected utility with
Bernoulli utility u(x). Consider the lottery L = {(0, 4, 9); (1/6, 1/2, 1/3)}

1. Provide the definition of certainty equivalent of L given u.

√
2. Compute the certainty equivalent for L given u(x) = x.

3. Compute the coefficient of absolute risk aversion given u and x.

4. Assume that the agent starts with w = 12 and owns the lottery. Find the minimum
price he is willing to accept to sell the lottery.

5. Compare the minimum price found above and the certainty equivalent and provide
an intuition of their difference based on the coefficient of absolute risk aversion.

Exercise 37
Solve Exercise 6.C.18 in MWG.

Exercise 38
Solve Exercise 6.C.16 in MWG assuming from the beginning
√ the parametrization of point
(d) with p = 0.5. Other than the proposed u(x) = x, consider also u(x) = x and
u(x) = log x. Give and interpretation in terms of the certainty equivalent and the change
of the risk aversion coefficient with wealth.
5 Solutions
Solution of Exercise 1
Consider an economy with two consumption goods, L = {1, 2}, and two consumers I =
{a, b}. Consumer a has endowment ωa = (3, 0) while consumer b has endowment ω2 =
(1, 3). Both consumers have utility U (x) = (x1 )2 (x2 )3 .
1. Find the set of Pareto optimal allocations.

2. Determine the subset of Pareto optimal allocations where both consumers are at least
as well-off as with their endowment (the contract curve).

3. Find the competitive equilibria. Are they Pareto optimal?

1. In order to find the Pareto set one should maximize the utility of one consumer by choos-
ing non-wasteful feasible allocation that do not decrease the utility of the other consumer.
The first order conditions (which are both necessary and sufficient given strict concavity
of U) for such a problem, considering all possible levels of utility, lead to

= 23 xx1b
 2 x2a a b
 3 x1a = M RS12 = M RS12 2b

x + x1b = 4
 1a
x2a + x2b = 3

whose solution is
     
3 3
PS = x1 = t, t , x2 = 4 − t, 3 − t , t ∈ [0, 4] ,
4 4

the Pareto Set. Note that for consumer b on the Pareto set x2b = 34 x1b .

2. To find the contract curve, first we compute both agents’ level of utility in the endow-
ment. It holds

Ua (ωa ) = (3)2 (0)3 = 0 and Ub (ωb ) = (1)2 (3)3 = 27 .

Then, we find the allocation on the Pareto set where both agents are at least as well of as
in their endowment by solving 

 x2a = 34 x1a
x1b = 4 − x1a




x2b = 3 − x2a


 x 1a ∈ [0, 4]
x2 x32a ≥ 0


 1a


x21b x32b ≥ 27
Solving
h the above,
i we find the contract curve as the subset of the Pareto set P S where
3
x1a ∈ 0, 4 − 4 .
5
3. To find the competitive equilibrium, we first find both agents offer curves (demand as
a function of the price ratio), getting
n  o
2 3p1 3 3p1
OCa = , p 1 > 0, p2 > 0 ,
n 5 p1 5 p2  o
2 p1 +3p2 3 p1 +3p2
OCb = 5 p1
, 5 p2
p 1 > 0, p2 > 0 .

(we can exclude the case of pl = 0 because it leads to an infinite demand of good l for both
agents and thus not to an equilibrium). Having the offer curves, we can impose the market
clearing conditions (by Ex. 4 imposing one condition is sufficient for solving the problem).
Equilibrium in the market for the first good leads to
2 3p1 2 p1 + 3p2 p1 1
+ =4⇒ = .
5 p1 5 p1 p2 2
The competitive equilibrium is the price found above and the corresponding optimal allo-
cation:  ∗    
p1 1 ∗ 6 9 ∗ 14 21
CE = = , xa = , , xb = ,
p∗2 2 5 10 5 10
Note that the allocation (x∗a , x∗b ) ∈ P S.

Solution of Exercise 2
Consider an economy with two consumers, I = {1, 2}, and two commodities, L = {1, 2}.
√
Consumer 1 has utility U1 (x1 ) = x11 + x21 and endowment ω1 = (2, 0). Consumer 2
values only good 2, U2 (x2 ) = x22 , and has endowment ω2 = (0, 2).
1.? Find the set of Pareto optimal allocations (a graphical solution is enough).

2.? Find the competitive equilibria (a graphical solution is enough).

3.?? Repeat the same analysis (both Pareto set and competitive equilibria) with the endow-
ments ω1 = (2, 1) and ω1 = (0, 1).
Note: This exercise is inspired by Figure 15.B.10(a) of the MWG textbook, you are en-
couraged to have a look at it.

1. The Pareto set of this economy is

P S = {(2, t), (0, 2 − t), t ∈ [0, 2]}

All other allocations of the Edgeworth box are not in the Pareto set.
At interior allocations agents have different M RS and first order conditions are not sat-
isfied (note that both can improve their welfare by letting agent 1 increase good 1 and
decrease good 2 while agent 2 increases good 2. For the same reason also allocations with
x22 = 0 or x11 = 0 can be improved upon. Allocations with x21 = 0 (but x11 < 2 are not
in the Pareto set because in all these allocations agent 2 has the same utility and agent 1
can be made better off by having x11 = 2.
NOTE: Agent 2 preferences not being strictly monotone implies that O1 = (0, 0) ∈
/ P S.

2. In this economy there are no prices that support a competitive equilibrium with the
given endowment. See also the text commenting Example 15.B.10(a) in the MWG.

3.a Consider ω1 = (2, 1) and ω2 = (0, 1). The endowment is in the Pareto Set and it is
interior for agent 1. Thus, provided
p1 1
p
= M RS12 (ω1 ) = 2 (ω21 ) = 2
p2
we have a competitive equilibrium. The competitive equilibrium is thus
 ∗ 
p1 ∗ ∗
CE = = 2, x1 = (2, 1) , x2 = (0, 1)
p∗2

3.b Consider ω1 = (0, 1) and ω2 = (2, 1). The endowment is not in the Pareto Set. We
expect the C.E. to be in the Pareto Set, thus we look for an allocation in the Pareto Set
that can be reached trading from ω. Assume fist that the C.E. allocation is interior for
agent 1, we look for a solution of


 x11 = 2


 x21 = t > 0 √
= 2 t = pp12 ≥ M RS12
 1 2
 M RS12 =0


x11 + x12 = 2
x21 + x22 = 2




p = p1 2 + tp2


 2


p1 2 + p2 = (2 − t)p2

Solving the above gives

√
 ∗
(p∗ )2 (p∗ )2
   
p1 ∗ ∗ ∗
CE = = p = 20 − 4, x1 = 2, , x2 = 0, 2 −
p∗2 4 4

Finally note that the allocation {(2, 0), (0, 2)} in the Pareto Set, not interior for agent 1,
cannot be a C.E. (see point 2. above).
Solution of Exercise 3
Consider an economy with two consumers, I = {1, 2}, and two commodities, L = {1, 2}.
√
Consumer 1 has utility U1 (x1 ) = x11 +x21 . Consumer 2 values only good 1, U2 (x2 ) = x12 .
The total endowment is ω̄.

1.? Find the set of Pareto optimal allocations (a graphical solution is enough).

2.? Find the competitive equilibria when ω21 = 0 (a graphical solution is enough).
 3.? Find the competitive equilibria when ω12 = 0 (a graphical solution is enough).

Note: This exercise is inspired by Figure 15.B.10(a) of the MWG textbook, you are en-
couraged to have a look at it.

1. The Pareto set of this economy is

P S = {(t, ω¯2 ), (ω¯1 − t, 0), t ∈ [0, ω¯1 ]}

All other allocations of the Edgeworth box are not in the Pareto set for similar reasons as
given in the solution of point 1. Ex. 3.
NOTE: Also here the fact that agent 2 preferences are not strictly monotone implies that
O1 = (0, 0) ∈
/ P S.

2. and 3. To be proposed in class (see the notes at the end of Lecture IV, ’20-21).

Solution of Exercise 4
Consider an economy with two consumers, I = {1, 2}, and two commodities, L = {1, 2}.
The endowments are ω1 and ω2 . Assume that both consumers have locally non satiated
preferences. Prove that if the market for good l clears and p  0, then also the market for
good l0 6= l clears.

First note that if preferences are LNS, then the optimal consumption is chosen on the
budget line:
pxi = pωi i = 1, 2.
Adding up the latter leads to
X X
pl (xli − ωli ) = 0
i=1,2 l=1,2

or, equivalently, X X
pl (xli − ωli ) = 0.
l=1,2 i=1,2
P
The latter together with market clearing in good l, i (xli − ωli ) = 0, leads to
X
pl0 (xl0 i − ωl0 i ) = 0.
i=1,2

If pl0 > 0, then it must be i (xl0 i − ωl0 i ) = 0 so that also the market for good l0 clears.
P

Solution of Exercise 5
Consider an economy with two consumers, I = {1, 2}, and two commodities, L = {1, 2}.
The total endowment is ω̄ = (4, 3). Find the set of Pareto optimal allocations in the
following cases:
 √ √ 1 2
1. U1 (x1 ) = x11 x21 and U2 (x2 ) = (x12 ) 3 (x22 ) 3 ;
2. U1 (x1 ) = x11 + x21 and U2 (x2 ) = 2x12 + x22 (a graphical solution is enough);
√ √
3.? U1 (x1 ) = x11 x21 and U2 (x2 ) = x12 + 2x22 .
1. We proceed as in point 1. of Ex. 1. Both agents preferences are strictly convex and
strongly monotone, thus the FOC are necessary and sufficient. Computing MRS for both
agents gives  x21
 x11 = M RS12 1 2
= M RS12 = 12 xx12
22

x + x12 = 4 ,
 11
x21 + x22 = 3
whose solution is
     
3t 3t
P S = x1 = t, , x2 = 4 − t, 3 − , t ∈ [0, 4] .
8−t 8−t
2. Given that both agents’ preferences are represented by a linear utility, and that
1 2
M RS12 = 1 6= M RS12 = 2, there is not interior solution that satisfies the FOC condi-
tions for Pareto optimality. Evaluating the frontier of the Edgeworth box, we notice that
both when x12 = 0 and when x21 = 0 agent 1 (2) can decrease (increase) consumption of
good 1 and increase (decrease) consumption of good 2 to achieve a higher utility. Thus
these allocations are Pareto optimal. Instead, when x11 = 0 agent 1 cannot decrease fur-
ther consumption of good 1 even if she values, relative to agent 2, good 1 less than good
1 2
2, M RS12 = 1 < M RS12 = 2. In the same way, when x22 = 0 agent 2 cannot decrease
further the consumption of good 2 even if she values, relative to agent 1, good 2 less than
2 1
good 1, M RS21 = 1/2 < M RS21 = 1. Thus
P S = {x1 = (0, t), x2 = (4, 3 − t), t ∈ [0, 3]} ∪ {x1 = (t, 3), x2 = (t, 0), t ∈ [0, 4]} .
3. To find the Pareto set we can proceed as in point 1., i.e. by maximizing agent 1 utility
by choosing allocations in the Edgeworth both where the utility of agent 2 does not fall
below a given level. Linearity of agent 2 utility implies that the Pareto Set might be on
the frontier of the Edgeworth box. In particular, to maximize agent 2 utility we might end
up giving to her only one good. The FOC for an internal solutions are
 x21

 x11
= M RS12 1 2
= M RS12 = 12
x11 + x11 = 4

.

 x21 + x21 = 3
xli > 0 ∀l, i


The latter is solved by x21 = x11 /2 provided x11 ∈ (0, 4) (the bundle of agent 2 follows
by imposing the F.C.). Another set of FOC considers the possibility that agent 2 has no
consumption of either good. Let us start with good 1, we obtain:
 x21
= M RS121
> M RS122
= 12
 x11


x11 + x11 = 4
,

 x 21 + x21 = 3
x12 = 0, x11 = 4, x21 > 0, x22 > 0

with solution {x12 = 0, x11 = 4, x21 = t ∈ [2, 3), x22 = 3 − t}. When, instead, agent 2 does
not consume good 2 we obtain:
 x21

 x11
= M RS121
< M RS12 2
= 12
x11 + x11 = 4

.

 x21 + x21 = 3
x12 > 0, x11 > 0, x21 = 3, x22 = 0


The latter implies, at the same time, x11 > 6 and x11 ≤ 4, thus leading to no solutions.
The Pareto set is thus
     
t t
P S = x1 = t, , x2 = 4 − t, 3 − , t ∈ [0, 4) ∪{x1 = (4, t), x2 = (0, 3 − t), t ∈ [2, 3]} .
2 2

(note that we have added the allocations when either agent has all the endowment)

Solution of Exercise 6
Consider an economy with two consumption goods, L = {1, 2}, and two consumers I =
{1, 2}. Consumer 1 has endowment ω1 = (1, 2) while consumer 2 has endowment ω2 =
(2, 1). Consumer 1 has utility U1 (x1 ) = x11 (x21 )2 while 2 has utility U2 (x2 ) = x12 x22 .
1. Find the set of Pareto optimal allocations

2. Determine the subset of Pareto optimal allocations where both consumers are at least
as well-off as with their endowment (the contract curve)

3. Find whether the allocation on the Pareto set with x11 = 2 can be supported as a
competitive equilibrium with transfers. If so, for which transfer using only good 1 ?
Is there a transfer of only good 2 that achieves the same allocation?
1. In order to find the Pareto set one should maximize the utility of one consumer by choos-
ing non-wasteful feasible allocation that do not decrease the utility of the other consumer.
The first order conditions (which are both necessary and sufficient given strict concavity
of U) for such a problem and considering all possible levels of utility and interior solutions
are  1 x21
= M RS12 1 2
= M RS12 = xx22
 2 x11

 12
x11 + x12 = 3
x + x22 = 3
 21


xli > 0 ∀l, i
whose solution is (including the two “origins” where either consumer has the total endow-
ment)      
6t 6t
P S = x1 = t, , x2 = 3 − t, 3 − , t ∈ [0, 3] ,
3+t 3+t
the Pareto Set.
2. To find the contract curve, first we compute both agents’ level of utility in the endow-
ment. It holds
U1 (ω1 ) = (1)1 (2)2 = 4 and U2 (ω2 ) = (2)(1) = 2 .
Then, we find the allocation on the Pareto set where both agents are at least as well of as
in their endowment by solving
6x11


 x21 = 3+x 11
x = 3 − x11

12



x22 = 3 − x21


 x 11 ∈ [0, 3]
x x2 ≥ 4


 211 321


x12 x22 ≥ 2
Graphically it can be shown that there exist two levels of x11 , xd11 and xu11 with xu11 > xd11
such that the
 dsolution
 of the above (the contract curve) is the subset of the Pareto set P S
u
with x11 ∈ x11 , x11 .

4. The allocation of the Pareto Set with x11 = 2 has

6x11 12 6x11 3
x21 = |x11 =2 = , x12 = x11 − 3|x11 =2 = 1, x22 = 3 − |x11 =2 = .
3 + x11 5 3 + x11 15
At this allocation we have
1 3 2
M RS12 (x1 ) = = M RS12 (x2 ),
5
thus the allocation can not be supported as a competitive equilibrium with transfer with
supporting price ration
p1 3
= .
p2 5
To find the transfer, we solve for T1 such that

px1 = w1 = pω1 + T1 .

Equivalently
px2 = w2 = pω2 + T2 .
with T2 = −T1 . Expressing everything in terms of good 1, i.e. dividing left and write by
p1 , we find the transfer T1 in good 1
T1 p2 p2
= x11 + x21 − ω11 − ω21 .
p1 p1 p1
Using prices and allocation found above we obtain
T1 5 12 5 5
=2+ −1− 2= .
p1 3 5 3 3
The latter is the amount of good 1 that need to be given to consumer 1 to support x11 = 2
as a competitive equilibrium. Note that the transfer can be implemented because ω11 + Tp11 =
8
3
< 3 = ω̄1 and gives thus a feasible new endowment. Note also that working with T2 we
would have found the same result:
T2 5 T1 5
=− ⇒ = .
p1 3 p1 3
The same transfer can be expressed in good 2 by using
T1 T1 p1
= =1
p2 p1 p2

Also this transfer can be implemented leading to the new endowment (3, 2) for agent 1 and
(0, 1) for agent 2.

Solution of Exercise 7
We consider an exchange economy with two consumers and two goods.

• Consumer 1 has consumption set R2+ , initial endowment e1 = (1, 1) and utility
1/3
u1 (x11 , x12 ) = x11 (x12 ) /3
2

• Consumer 1 has consumption set R2+ , initial endowment e2 = (1, 3) and utility
1/2 1/2
u2 (x21 , x22 ) = x21 x22

1. Represent in the Edgeworth box the initial endowment and the indifference curves
going through the initial endowment for both agents.

2. Represent in the Edgeworth box the set B of allocations that are better, in the sense
of Pareto, than the initial endowments.

3. Give the definition of a Pareto Optimum for this economy.

4. Determine the set C of Pareto Optima of the economy.

5. Represent (approximately) the set C in the Edgeworth box. Do B and C coincide ?

Explain why/not ?

6. Give the definition of a general equilibrium for this economy.

7. Determine the general equilibrium of this economy (normalize to 1 the price of com-
modity 2).

TO BE PREPARED
Solution of Exercise 8
Solve Example 15.B.2 from the MWG textbook.

The solution is provided in the MWG textbook.

Solution of Exercise 9
Solve Exercise 15.B.2 from the MWG textbook.

The exercise generalizes Example 15.B.1 in the MWG textbook.

Solution of Exercise 10
Solve Exercise 15.B.9 from the MWG textbook.

Both consumers have Leontief preferences. Deriving the offer curves for both consumers
and crossing them leads to
n √ o
CE = p1 = 0, p2 = p, xα = (x, 0), xβ = (30 − x, 20), p > 0, x ∈ [0, 30 − 20]

when ω1 = (30, 0) and ω2 = (0, 20). The latter is a set of competitive equilibria also
when ω1 = (5, 0) and ω2 = (0, 20), provided 30 is replaced by 5. When ω1 = (5, 0) and
ω2 = (0, 20), in addition, we have
( √ √ √ ! √ √ !)
p 1 9 − 61 9 − 61 9 − 61 1 + 61 31 + 61
CE 1 = = √ , xα = , , xβ = ,
p2 1 + 61 2 2 2 2

and
CE 2 = {p1 = p, p2 = 0, xα = (5, 20 − x), xβ = (0, x), p > 0, x ∈ [0, 15]} .
Solution of Exercise 11
Solve Exercise 15.C.2 from the MWG textbook.

The exercise is a special case of Ex. 12 with A = 1, α = 12 , β = 1. Please have a

look at the solution of Ex. 12.
Solution of Exercise 12
Consider an economy with two periods, t = 0, 1, and a consumption good per period. The
consumer has utility U (x0 , x1 ) = u(x0 ) + βu(x1 ) with β ∈ (0, 1) and endowment k0 of
consumption good in t = 0. The economy has one firm, J = 1, which transforms the
good in t = 0 (input, k1 ), into an output (y) in t = 1. The firm production function is
f (k) = Ak α with A > 0 and α ∈ (0, 1]. Assume that u(x) = log x and α ∈ (0, 1).
1. Find the competitive equilibrium
2. Find the firm profit in the competitive equilibrium and discuss its dependence on α
3. Find the Pareto optimal allocation of this economy and compare it with the allocation
found in 1
1. To find the competitive equilibrium we first solve for the the firms optimal decision as a
function of the price ratio. This gives an optimal demand k1 (p), optimal supply y(p), and
an equilibrium profit π(p). Setting p0 = p and p1 = 1 and solving for
Max y − pk1 , such that
0 ≤ y ≤ Ak1α
k1 ≥ 0
leads to
 1
 1−α  α
 1−α  α
 1−α
αA αA αA
k1 (p) = , y(p) = A , π(p) = A(1 − α) .
p p p
Having the profit of the firm, we can compute agent optimal demand for each good. Agent
1 β
1 has Cobb Douglas utility with coefficient 1+β for good 1 and 1+β for good 2, her demand
is thus
1 pk0 +π(p)
x0 (p) = 1+β p
,
β
x1 (p) = 1+β (pk0 + π(p)) .
Imposing market clearing, e.g. in good 1, leads to
 1−α
∗ 1 + αβ
x0 (p) + k1 (p) = k0 ⇒ p = αA
αβk0
Note that market clearing for good 2 is implied by the latter together with the agent BC
(which includes the firm profit) and the definition of firm profit. The CE is thus
   α   α 
∗ k0 k0 αβ k0 αβ k0 αβ
CE = p , x = ,A , k1 = , y1 = A
1 + αβ 1 + αβ 1 + αβ 1 + αβ
2. Substituting the market clearing price p∗ in the optimal profit function π(p) leads to
 α
∗ αβk0
π(p ) = A(1 − α)
1 + αβ
The latter goes to 0 when α → 1. Note that this is consistent with a linear technology (and
thus zero profit in equilibrium) when α = 1. When α → 0, the profits goes to A (taking
the log and using de L’Hopital rule to derive that limx→0 x log x = 0). This is because, on
the cost side, the use of input goes to zero but its price does not diverge, so that the cost
goes to zero; on the revenue side, the price of output is normalized to one and the output
goes to A, so that the revenue goes to A.

3. To find the Pareto Set, we maximize the agent utility given the feasibility constraint:
Max log x0 + β log x1 , such that
x0 + k1 ≤ k0
0 ≤ y ≤ Ak1α
x1 ≤ y
Imposing all constraints with equality the problem becomes

Max log (k0 − k1 ) + βα log Ak1 , such that

k1 ∈ [0, k0 ]
k0 αβ
whose unique (internal) solution is k1 = 1+αβ
. Imposing the constraints leads to
   α   α 
k0 k0 αβ k0 αβ k0 αβ
PS = x = ,A , k1 = , y1 = A .
1 + αβ 1 + αβ 1 + αβ 1 + αβ

Note that the marginal productivity of capital in the Pareto optimal allocation is equal to
the supporting price in the CE:
 1−α
0 1 + αβ
α−1
f (k1 ) = αAk1 = αA = p∗ .
k0 αβ

Solution of Exercise 13
Solve Exercise 12 when

1.? u(x) = x and α ∈ (0, 1);

2. u(x) = log x and α = 1;

3.?? u(x) = x and α = 1.

1. Normalize p1 = 1 and name p0 = p. The firm supply is the same as computed in point
1. of Ex. 12. Given linear utility, consumer demand is
 1
 any bundle on the budget line if p = β
(x0 (p), x1 (p)) = only good x0 ≥ k0 if p < β1
only good x1 if p > β1


No equilibrium is possible if p < β1 . In fact, if p < β1 , then firm has positive demand of
k1 but the consumer consumes (at least) all her endowment k0 . Thus, in a competitive
equilibrium it must be p ≥ β1 .
Let us consider first the case p = β1 . Firm optimal demand is
 
1 1
k1 p = = (αβA) 1−α .
β

Such demand is compatible with an equilibrium only if lower or equal than k0 , leading to
 
1 1 α α
1
CE = p = , x = (k0 − t, At ) , (k1 , y) = (t, At ), 0 < t = (αβA) 1−α ≤ k0 .
β
 1
Note that when (αβA) 1−α = k0 , the marginal product of capital in k0 is exactly β1 . In this
case, the consumer is giving away all her endowment k0 in exchange of consumption f (k0 )
1
in date t = 1. A similar situation occurs when (αβA) 1−α > k0 . In this case there is no
equilibrium with p = β1 , in fact the demand of input at this price is larger than k0 and
cannot be satisfied. However there is another supporting price. Choosing p = f 0 (k0 ) leads
1
to p > β1 if (and only if) (αβA) 1−α > k0 . At this price the firm demand for input is k0 and
the consumer, facing a too high price for the consumption good in t = 0 chooses x0 = 0.
This competitive equilibrium is as follows
n 1
o
CE 2 = p = αAk0α−1 , x = (0, Ak0α ) , (k1 , y) = (k0 , Ak0α ), (αβA) 1−α > k0 .

The set of equilibria is thus

CE = CE 1 ∪ CE 2
By maximizing consumer welfare on the feasible allocations, it can be shown that P S =
CE.

2. Keeping the same price normalization as above, let us start to consider the firm optimal
choice of input and output. Note that the firm has a linear, constant return to scale,
technology. The firm optimal choice is

 k1 ∈ [0, +∞), y = Ak1 when p = A,
(k1 (p), y(p)) = (+∞, +∞) when p < A,
(0, 0) when p > A.


Note that there can not be an equilibrium with p < A -unbounded demand (supply) of
input (output)- and that when p ≥ A the firm profits are zero. Turing to the consumer,
and imposing zero profits for the firm, we have (as in Ex. 12)
1
x0 (p) = k,
1+β 0
β
x1 (p) = 1+β
pk0 .

Imposing the market clearing conditions we note that when p ≥ A, the consumer has
positive demand also for good x1 and thus needs the firm to produce. Thus, there cannot
be an equilibrium with p > A. When p = A, the consumer has a demand
1 β
x0 (A) = k0 x1 (A) = Ak0
1+β 1+β
and the firm, being indifferent on all bundles on the production set frontier, is willing to
provide it. Note that indeed when
β
k1 = k0 − x0 (A) = k0
1+β
the firm has output
β
f (k1 ) = Ak0 = x1 (A),
1+β
so that both markets clear. The competitive equilibrium is thus
    
1 β β β
CE = p = A, x = k0 , Ak0 , (k1 , y) = k0 , Ak0 .
1+β 1+β 1+β 1+β
Also in this case CE = P S.
3. We turn to the case when the consumer has a linear utility with M RS12 1
= β1 > 1 and
the firm has a linear production function with f 0 (k) = A. Optimal behaviors are extreme,
as derived in point 1. above for the consumer and in point 2. above for the firm. For
all possible values of A, we can rule out p < A because it would lead to an unbounded
demand of the firm. Possible supporting prices are thus p ≥ A, and for all these prices the
firm has zero profits.
Let us first consider the case A = β1 > 1. The only possible supporting price is p = A,
with zero profits for the firm and all allocations on the budget line (which coincides with
part of the production frontier) as possible equilibria. We have
 
1 1
CE = p = A, x = (k0 − t, At), (k1 , y) = (t, At), t ∈ [0, k0 ], A = .
β
Let us turn to A > β1 > 1. Then, when p ≥ A it also holds p > β1 so that the consumer
wants to sell k0 and buy only x1 = pk0 (remember that firm profits are zero). If p = A the
firm is willing to absorb all the supply of k0 and produce Ak0 to satisfy the consumer. If
p > A, the firm chooses not to produce and the goods market do not clear. Thus
 
2 1
CE = p = A, x = 0, Ak0 ), (k1 , y) = (k0 , Ak0 ), A > .
β
The last case to consider has A < β1 . When p = A the firm is indifferent and the consumer,
facing p < β1 , chooses x = (k0 , 0). We have
 
3 1
CE = p = A, x = k0 , 0), (k1 , y) = (0, 0), A < .
β
When β1 ≥ p > A, the situation is as above, with the difference that the firm only choice
is not to produce (and still consistent with the consumer choice). We have
   
4 1 1
CE = p ∈ A, , x = k0 , 0), (k1 , y) = (0, 0), A < .
β β
Finally when p > β1 > A, the firm does not produce but the consumer wants to consume
only the firm output, no equilibrium in this case. Summarizing we have
CE = CE 1 ∪ CE 2 ∪ CE 3 ∪ CE 4 .
and it can be shown that CE = P S.
Solution of Exercise 14
Consider an economy with two consumption goods, L = 2, one consumer, I = 1, and one
firm, J = 1. The consumer has endowment ω = (4, 1) and utility U1 (x) = x1 x2 . The firm
has production set Y1 = {y2 ≤ −ay1 , y1 ≤ 0} with a > 0. The consumer is the owner of
the firm.
1. Write down the definition of a competitive equilibrium for this economy.
2. Fix a = 1, find the set of competitive equilibria and the corresponding firm profits.
3. Are the the competitive equilibria found above Pareto optimal?
4. Find the Pareto optimal allocations for a = 1.
5.? Answer to 2. and 3. for each given a > 0.
6.? Find the Pareto optimal allocations for each given a > 0.
1. A competitive equilibrium for this economy is a supporting price vector p∗ = (p∗1 , p∗2 )
and an allocation (x∗ , y ∗ ) with x∗ ∈ R2+ and {y2∗ + ay1∗ ≤ 0, y1 ≤ 0} such that
- The firm maximizes profits given p, that is, p∗ y ∗ ≥ p∗ y for all y with y2 + ay1 ≤ 0
and y1 ≤ 0;
- The consumer maximizes his utility given p∗ and income p∗ ω + p∗ y ∗ , that is, U (x∗ ) ≥
U (x) for all x ∈ R2+ with p∗ x ≤ p∗ ω + p∗ y ∗ ;
- The two goods market clear, that is, x∗l = ωl + yl∗ for l = 1, 2.
2. Take a = 1. To find an equilibrium we start from the firm (we need the firm profit
into the consumer demand). The firm has a linear technology and its optimal choice as a
function of the price vector p is
p1
 −y1 ∈ [0, +∞), y2 = −y1 when p2 = 1,


(y1 (p), y2 (p)) = (−∞, +∞) when pp21 < 1,

(0, 0) when pp12 > 1.


Note that there can not be an equilibrium with pp12 < 1 -due to unbounded demand (supply)
of input (output)- and that when pp12 ≥ 1 the firm profits are zero. Turing to the consumer,
and imposing zero profits for the firm, we have
1 4p1 +p2
x1 (p1 /p2 ) = 2 p1
,
1 4p1 +p2
x2 (p1 /p2 ) = 2 p2
.
Imposing the market clearing conditions we note that when pp12 ≥ 1, the consumer has
positive demand also for good x1 and thus needs the firm to produce. Thus, there cannot
be an equilibrium with pp12 > 1. When pp12 = 1, the consumer has demand
5 5
x1 (1) = x2 (1) =
2 2
and the firm, being indifferent on all bundles on the production set frontier, is willing to
provide it. Note that indeed when
3
y1 = x1 (1) − 4 = −
2
the firm has output
3
y2 = −y1 = = x2 (1) − 1,
2
so that both markets clear. The competitive equilibrium is thus
 ∗    
p1 ∗ 5 5 ∗ 3 3
CE = = 1, x = , ,y = − , .
p∗2 2 2 2 2

Note Exploiting the I welfare theorem (see below), we could have derived the CE also by
finding the Pareto Set first (here a unique allocation, see point 4.) and a suitable supporting
price vector.
3. Yes, the CE found above is Pareto optimal due to the I welfare theorem, which holds
due to strong monotonicity, and thus LNS, of the consumer preferences as represented by
U (x) = x1 x2 .
4. Fix a = 1. To find the Pareto set we must maximize the consumer utility in the feasible
set. We can solve
Max x1 x2 , such that
x ∈ R2+
y1 ≤ 0
y2 ≤ −y1
x1 = 4 + y 1
x1 = 1 + y 2
Imposing all constraints with equality the problem becomes

Max (4 + y1 )(1 − y1 ), such that

y1 ∈ [−4, 0]

whose unique (internal) solution is y1 = − 23 . Imposing the constraints leads to

    
5 5 3 3
PS = x = , , y1 = − , .
2 2 2 2

5. We proceed similarly to point 2.. First note that with a = 1, the firm provides the
transformation of inputs into output that pleases the consumer with a unitary price vector.
As a changes, we expect the price vector to respond (so that the linear firm is still happy
to supply what the consumer asks) and change as well as pp12 = a. However there will be
low values of a such that the consumer would like to increase his consumption of good 1
beyond ω1 = 4 and decrease the consumption of good 2 below ω2 = 1. Such configuration
cannot be an equilibrium as the firm is bound to use good 1 as input, and cannot supply
it to the consumer.
Having this is mind, the competitive equilibrium is the same “interior” allocation as found
in point 2., that is, with pp12 = a, as long as x1 (a) ≤ 4. Imposing the condition we find

4a + 1 1
x1 (a) = ≤4 ⇒ a≥
2a 4
Using the demand of good 2 and the market clearing condition(s) to find the firm choice
in equilibrium we obtain that if a ≥ 41 , then
 ∗    
a p1 ∗ 4a + 1 4a + 1 ∗ 1 − 4a 4a − 1
CE = = a, x = , ,y = ,
p∗2 2a 2 2a 2
Let’s turn to the case a < 14 . Now the price ratio cannot be equal to a, otherwise the
consumer would demand good 1 in exchange of good 2. A possibility is that pp12 = 14 and
the consumer consumes her endowment (4, 1). The latter can be part of an equilibrium
only if the firm is willing not to operate at this price ratio. Having pp12 = 14 > a = M RT
implies that this is the case. We obtain that if a < 41 , then
 ∗ 
a p1 1 ∗ ∗
CE = = , x = (4, 1) , y = (0, 0) .
p∗2 4
As the preferences of the consumer have not changed, for each a > 0 the CE is in the PS.

6. To find the Pareto set we must maximize the consumer utility in the feasible set. We
can solve
Max x1 x2 , such that
x ∈ R2+
y1 ≤ 0
y2 ≤ −ay1
x1 = 4 + y 1
x1 = 1 + y 2
Imposing all constraints with equality the problem becomes
Max (4 + y1 )(1 − ay1 ), such that
y1 ∈ [−4, 0]
The solution of the latter is internal and equal to
1 − 4a
y1 =
2a
when a > 14 . Otherwise, when ain 0, 41 the solution is at the right-end border, y1 = 0.


Using the feasibility constraints lead to the Pareto Set. If a > 14 , then
    
a 1 + 4a 1 + 4a 1 − 4a 4a − 1
PS = x = , , y1 = , .
2a 2 2a 2
If a ≤ 14 , then
P S a = {x = (4, 1) , y1 = (0, 0)} .
Solution of Exercise 15
Consider an economy with L commodities, I consumers, and J firms. The endowments

I
and ownership shares are ωi , (θji )Jj=1 i=1 . Assume that all consumers have locally non
satiated preferences. Prove that, for all choices of a good l0 , if the markets for all goods
l 6= l0 clears and p  0, then also the market for good l0 clears.

First note that if preferences are LNS, then the optimal consumption is chosen on the
budget line X
pxi = pωi + θji pyj i = 1, . . . , L.
j∈J

Adding up the latter leads to

!
X X XX
pl (xli − ωli ) − θij pl ylj =0
i∈I l∈L j∈J l∈L

or, equivalently, !
X X X
pl (xli − ωli ) − ylj = 0.
l∈L i∈I j∈J

The latter together with market clearing in all goods l 6= l0 , i∈I (xli − ωli ) = j∈J ylj ,
P P
leads to !
X X
pl 0 (xl0 i − ωl0 i ) − yl0 j = 0.
i∈I j∈J
P P
If pl0 > 0, then it must be i∈I (xl0 i − ωl0 i ) = j∈J yl0 j , so that also the market for good
l0 clears.
Solution of Exercise 16
Consider an exchange economy with two consumption goods, L = 2, and two consumers,
I = 2. Consumer 1 has endowment ω1 = (2, 0) and utility U1 (x1 ) = (x11 )α + (x21 )α .
Consumer 1, has endowment ω2 = (1, 3) and utility U2 (x2 ) = x12 x22 . Consider α = 21 first.

1. Find the competitive equilibria and the set of Pareto optimal allocations.

2. Discuss the validity of both Welfare Theorems.

3. The social planner has the possibility to implement some transfer between the initial
endowments of commodity 2. Is there a transfer such that, in the competitive equi-
librium achieved after trading, both agents consume the same amount of good 1. If
so, find it. Can the same competitive equilibrium arise after a transfer of good 1? If
so, find it.
 4. Repeat the exercise taking α = 1.

1. To find the competitive equilibrium of this economy, we first derive the demand of both
agents by solving their utility maximization given their budget constraint. For agent 1 we
obtain:
2 2p
x11 (p) = p1 and x21 (p) =  1 ,
1 + p2 p2 1 + pp12
while for agent 2 we have
1 p1 + 3p2 1 p1 + 3p2
x12 (p) = and x22 (p) = .
2 p1 2 p2
Market clearing in the first good gives the equilibrium price ratio
2 1 p1 + 3p2 p1
p1 + =3 ⇒ = 1.
1 + p2 2 p1 p2

The resulting equilibrium allocation is thus

x∗1 = (1, 1) x∗2 = (2, 2)

(note that also the market for good 2 clears). As for Pareto optimality, strong monotonicity
of both consumers preferences, allows us to pick either agent as the one whose utility is to
be maximized, given different utility levels of the other agent. Choosing agent 1, we have
to solve
M ax U1 (x1 )
x1 ∈ R + +
2 , x2 ∈ R2
x11 + x12 ≤ 3
x21 + x22 ≤ 3
U2 (x2 ) ≥ Ū
for given levels of Ū . The first order conditions at an interior and non wasteful allocation
are √
x
1
M RS1,2 = √x21 11
= xx22
12
2
= M RS1,2
x11 + x12 = 3
x21 + x22 = 3.
Solving the above gives

P S = {x1 = (t, t), x2 = (3 − t, 3 − t), t ∈ [0, 3]} ,

where we have extended the solution to the border cases having x1 = ω̄ or x2 = ω̄.

2. Both consumers have local non satiated preferences (implied by monotonicity), thus the
First Welfare Theorem holds. In fact, the competitive equilibrium allocation found above
belongs to the Pareto Set. Both consumers have also convex and continuous preferences.
Thus, for all allocations x in the Pareto Set, there exists a price vector p = (p1 , p2 ) 6= 0
such that {p, x} is a quasi-equilibrium with transfer. In all such allocations, computing
the MRS, p = (p, p)  0. Thus all quasi-equilibrium with transfer are also equilibria: if
the allocation is interior, each agent has positive wealth; if the allocation is at the border,
agent i with xi = 0 has only 0 in the budget set.

3. A social planner can transfer amounts of good 2 to achieve an allocation where both
consumers consume the same amount of the first good. Such allocation, to be “stable”,
must also be in the Pareto set (otherwise agents would want to trade away from it). The
allocation we are looking for solves


 x11 = x12
x11 = x21

,

 x12 = 3 − x11
x22 = 3 − x21


leading to  
3 3
x01 = x02 = , .
2 2
Such allocation is a competitive equilibrium together with p = (p, p). The transfer that
support such an equilibrium is
 0
 px1 = pω1 + T1 ⇒ 3p = 2p + T1
px0 = pω2 + T2 ⇒ 3p = 4p + T2
 2
T1 + T2 = 0

giving T1 = −T2 = p. For such a transfer to be implemented using good 2 the social
planner needs to transfer
T1 T1
t= = =1
p2 p
from consumer 2 to consumer 1. Given that consumer 2 is endowed with 3 units of the
second good, such transfer is possible and leads to

ω10 = ω1 + (0, 1) = (2, 1) and ω20 = ω2 − (0, 1) = (1, 2).

In terms of good 1 the transfer should be the same (the two goods have the same price in
equilibrium) and it is also implementable, giving

ω10 = ω1 + (1, 0) = (3, 0) and ω20 = ω2 − (1, 0) = (0, 3).

1
4. When α = 1 the first consumer has linear preferences with M RS12 = 1. It can be
checked that the Pareto set and the competitive equilibrium are the same (the first agent
is indifferent with p1 = p2 = p. The discussion on the Welfare theorems and the transfer
are also identical.
Solution of Exercise 17
Consider an economy with two consumption goods, L = 2, two consumers, I = 2, and
one firm, J = 1. Consumer 1 has endowment ω1 = (4, 1) and utility U1 (x1 ) = x11 x21 .
Consumer 2 has endowment ω2 = (1, 1) and utility U (x2 ) = x12 + x22 . The firm has
√
production set Y = {y2 ≤ −y1 , y1 ≤ 0}. Consumer two is the owner of the firm.

1. Write down the definition of a competitive equilibrium in this economy.

2. Find the set of competitive equilibria and the corresponding firm profits.

3. Are the competitive equilibria allocation found above Pareto optimal?

4.?? Find the set of Pareto optimal allocation of this economy.

1. A competitive equilibrium for this economy is a supporting√ price vector p∗ = (p∗1 , p∗2 )
and an allocation (x∗1 , x∗2 , y ∗ ) with x∗ ∈ R4+ and {y2∗ − −y1∗ ≤ 0, y1 ≤ 0} such that
√
- The firm maximizes profits given p, that is, p∗ y ∗ ≥ p∗ y for all y with y2 − −y1 ≤ 0
and y1 ≤ 0;

- Consumer 1 maximizes her utility given p∗ and income p∗ ω1 , that is, U1 (x∗1 ) ≥ U1 (x)
for all x1 ∈ R2+ with p∗ x1 ≤ p∗ ω1 ;

- Consumer 2 maximizes her utility given p∗ and income p∗ ω2 + p∗ y ∗ , that is, U2 (x∗2 ) ≥
U2 (x2 ) for all x2 ∈ R2+ with p∗ x2 ≤ p∗ ω2 + p∗ y ∗ ;

- The two goods market clear, that is, x∗l1 + x∗l2 = ωl1 + ωl2 + yl∗ for l = 1, 2.

2. To find an equilibrium we start from the firm (we need the firm profit into consumer 2
demand). The firm has a concave production function, its optimal choice as a function of
the price vector p is a solution of

Max p2 y2 + p1 y1 , such that

y1 ≤ 0
√
y2 ≤ −y1

Using FOC we derive

p22 p2 p22
y1 (p) = − 2
, y2 (p) = and thus π(p) =
4p1 2p1 4p1

Turing to the consumers, for consumer one (no profits in her income and Cobb Douglas
utility) we have
x11 (p) = 12 4p1p+p
1
2
,
1 4p1 +p2
x21 (p) = 2 p2 .
Consumer 2 has linear utility and thus her demand is given by

x ∈ 0, 94 , 94 − x when pp12 = 1,
  



  p +p +π(p) 
(x12 (p), x22 (p)) =
1 2
p1
, 0 when pp12 < 1,
  
 0, p1 +p2 +π(p) when p1 > 1,

p2 p2

where we have used that π(p)/p1 = π(p)/p2 = 1/4 when p1 /p2 = 1.

Imposing the market clearing conditions gives the supporting price ratio. Let us first check
if pp12 = 1 can be part of a competitive equilibrium. Optimal behavior of the firm and of
consumer 1 gives
5 5 1 1
x11 (1) = , x21 (1) = , y1 (1) = − , y2 (1) = .
2 2 4 2
Using market 1 clearing condition with x12 = x ∈ 0, 49 (which is optimal for this price
 

ratio) gives
5 1 9
+x=4+1− ⇒x= .
2 4 4
As a double check, in the market for good 2 we have
5 9 1 9
+ −x=1+1+ ⇒x= .
2 4 2 4
p1
In principle we could have competitive equilibria with other price ratios. Assuming p2
<1
implies x22 = 0 and leads to the following market clearing condition for good 2
1 4p1 + p2 1 p2
=2+ .
2 p2 2 p1
p1
Solving the II order equation in p2
and considering only the positive solution leads to
p1
=1≥1
p2
p1
The solution cannot be accepted because not smaller than 1. Assuming now p2
> 1 implies
x12 = 0 and leads to the following market clearing condition for good 1
 2
1 4p1 + p2 1 p2
=5− .
2 p1 4 p1
p2
Solving the II order equation in p1
and considering only the positive solution leads to

p1 1
= √ <1
p2 −1 + 13
The solution cannot be accepted because not greater than 1. To summarize the unique
competitive equilibrium is
 ∗      
p1 ∗ 5 5 ∗ 9 1 1
CE = = 1, x1 = , , x2 = ,0 ,y = − ,
p∗2 2 2 4 4 2

3. Both consumers preferences are LNS (due to strong monotonicity) and thus the I first
welfare theorem implies that the allocation of the CE found above is Pareto optimal.
4. To find the Pareto set, we can solve, for a given value of Ū2

Max U1 (x1 ) such that

x ∈ R4+
U2 (x2 ) ≥ Ū2
y1 ≤ 0
√
0 ≤ y2 ≤ −y1
x11 + x12 = 5 + y1
x21 + x22 = 2 + y2

(all these solutions are in the Pareto set due to strong monotonicity of both agents prefer-
ences). Equivalently, due to the concavity of utility functions we could solve

Max λ1 U1 (x1 ) + λ2 U1 (x1 ) such that

x ∈ R4+
y1 ≤ 0
√
0 ≤ y2 ≤ −y1
x11 + x12 = 5 + y1
x21 + x22 = 2 + y2

for all λ ≥ 0, λ 6= 0. we consider first the case of λ  0. The set of FOC of the problems
depends on whether positivity constraints on x are binding. An interior solution, xli > 0
for all l, i, solves
1
= xx21 2

 M RS12 11
= M RS12 =1
M RS12 = 1 = M RT12 = 2√1−y1

 2



x11 + x12 = 5 + y1

x 21 + x22 = 2 + y2
√


 y2 = −y1




xli > 0, y1 < 0, y2 > 0
Solving the system gives
      
1 19 5 1 1 5
P S = x1 = (t, t), x2 = − t, − t , y = − , , t ∈ 0,
4 2 4 2 2

Next we relax the positivity constraint, one at a time. We do so only for consumer 2
because consumer 1, as long as λ1 > 1, gets the lowest level of utility and she is thus never
satisfied when x11 = 0 or x21 = 0. Let us start with x22 = 0. The set of FOCs gives
1
= xx21 2

 M RS12 11
≤ M RS12 =1
x21 √1
 1



 M RS 12 = x11
= M RT 12 = 2 −y1
x11 + x12 = 5 + y1

x21 = 2 + y2
√


y = −y1


 2


x11 > 0, x12 > 0, x21 > 0, y1 < 0, y2 > 0
Solving the system gives
n  q 
4 64+(1−t)12
P S2 = x1 = 2x21 (t)(x21 − 2), x21 (t) = 3
+ 36
,

o
x2 = (t, 0), y = (−(x21 (t) − 2)2 , (x21 (t) − 2)), t ∈ 0, 94

Note that when t = 94 it holds x11 = x22 = 52 and M RS12 1 2

= M RS12 , the same allocation
1 5
in P S when t = 2 . We turn to x12 = 0, implying
1
= xx21 2

 M RS12 11
≥ M RS12 =1
x21
M RS12 = x11 = M RT12 = 2√1−y1

 1



x11 = 5 + y1

x21 + x22 = 2 + y2
√


y2 = −y1





x11 > 0, x12 > 0, x22 > 0, y1 < 0, y2 > 0
It can be checked that the system has no solutions, in particular the feasibility conditions
and positivity constraint lead to
19 10
x11 ≥ and x21 ≥
4 4
1 2
whereas M RS12 ≥ M RS12 = 1 implies x21 ≥ x11 . Two more (sets of) Pareto optimal
allocations are found when λ1 = 0 or when λ2 = 0. Setting t = 0 in P S 1 and in P S 2 gives,
respectively, these allocations.
Solution of Exercise 18
Consider three commodities, time for leisure/labor (commodity 1), durable good (commodity
2) and consumption good (commodity 3). There is a single consumer with utility function
U (x2 , x3 ) = x2 + x3 , so that the consumption of leisure does not have any impact on his
utility. The consumer has an endowment of leisure equal to 1 and no endowment of durable
and consumption good.

There are two firms. Firm 1 produces only durable goods using labor as input. Firm 1
√
production set is Y1 = {− −y11 + y21 ≤ 0, y11 ≤ 0}. Firm 2 produces only consumption
√
good using labor as input. Firm 2 production set is Y2 = {−2 −y12 + y32 ≤ 0, y12 ≤ 0}.
The consumer is the owner of both firms.
 1. Write down the definition of a competitive equilibrium in this economy.

2. Check that
( √      )
p1 5 p3 1 4 1 1 4 4
= , = 1, x = 0, √ , √ , y1 = − , √ , y2 = − , √
p2 2 p2 5 5 5 5 5 5

is the only competitive equilibrium of this economy.

3. Is the competitive equilibria allocation found above Pareto optimal?

4.? Find the set of Pareto optimal allocation of this economy.

∗ ∗ ∗ ∗
1. A competitive equilibrium for this economy is a supporting
√ ∗ price vector p = (p1 , p2 , p3 )
∗ ∗ ∗ ∗ 3 ∗
and an √allocation (x , y1 , y2 ) with x ∈ R+ , {y21 − −y11 ≤ 0, y11 ≤ 0} = Y1 , and
∗ ∗
{y22 − 2 −y12 ≤ 0, y12 ≤ 0} = Y2 such that

- Both firms j = 1, 2 maximize profits given p, that is, p∗ yj∗ ≥ p∗ yj for all yj ∈ Yj ;

- the consumer maximizes her utility given p∗ and income p∗ ω, that is, U (x∗ ) ≥ U (x)
for all x1 ∈ R2+ with p∗ x1 ≤ p∗ ω + p∗ y1∗ + p∗ y2 ;

- The goods market clear, that is, x∗l = ωl + yl1

∗ ∗
+ yl2 for l = 1, 2, 3.

2. To check that an allocation is a competitive equilibrium, we check that it satisfies the

market clearing conditions as well as firms and consumer optimal behavior. The latter is
equivalent to first order conditions, which, given concavity of utility and convexity of the
production sets are both necessary and sufficient. The first order conditions of firm j = 1
are
√1 = p1
2 −y11 √p2
y21 = − −y11
giving  2
p2 p2
y11 (p) = − and y21 (p) = .
2p1 2p1
  √
It can be easily checked that (y11 , y21 ) = − 15 , √15 is optimal when pp21 = 2
5
. Note that
with this allocation and price ratio, firm 1 profits are
 
p1 1
π1 (p) = p2 y2 + p1 y1 = p2 y2 + y1 = p2 √ > 0.
p2 2 5
Turning to firm j = 2, the focs are
√ 1
−y12
= pp13
√
y32 = −2 −y12
giving  2
p3 p3
y12 (p) = − and y32 (p) = 2
p1 p1
  √
It can be easily checked that (y12 , y32 ) = is optimal when pp13 =
− 45 , √45 p1 p2
p2 p3
= 2
5
. Note
that with this allocation and price ratio, firm 3 profits are
 
p1 2
π2 (p) = p3 y3 + p1 y1 = p3 y3 + y1 = p3 √ > 0.
p3 5
We turn to the consumer. Having linear utility in x2 and x3 and being pp32 = 1 the consumer
is indifferent between the two goods. We should check that she is choosing on her budget
line. The latter takes into account the income coming from the endowment and the income
coming from both firm profits, leading to

p2 x2 + p3 x3 = p1 + π1 (p) + π2 (p)

or, equivalently,
p3 p1 π1 (p) p3 π1 (p)
x2 + x3 = + + .
p2 p2 p2 p2 p3
Plugging in the values of profits found above, it can be easily checked that the given
consumption bundle is on the budget line for the given price ratios.
Have checked that we have a competitive equilibrium we should still prove that this is the
unique one. First, we exploit consumer preferences and note
p2 p2 p3 p2
> 1 ⇒ x2 = 0, x3 > 0 ⇒ y2 = 0, y3 > 0 ⇒ = 0, >0⇒ =0
p3 p1 p1 p3
thus leading to a contradiction. No equilibrium exists with such a price ratio. Having
p2
p3
< 1 leads to a similar contradiction. Thus it must be pp23 = 1, and thus also pp31 = pp21 .
The value of pp12 can be found by imposing market clearing in the first market. Using the
optimal demand of inputs by both firms gives
 2  2
1 p2 p3
1 + y11 (p) + y12 (p) = 0 ⇒ + =1
4 p1 p1
√
p1 5
giving only p2
= 2
when p2 = p3 .

3. The consumer preferences are LNS (due to monotonicity) and thus the I first welfare
theorem implies that the allocation of the CE found above is Pareto optimal.
4. To find the Pareto set, we can solve

Max x2 + x3 such that

x ∈ R3+
√
y11 ≤ 0, 0 ≤ y21 ≤ −y11
√
y12 ≤ 0, 0 ≤ y32 ≤ 2 −y12
x1 = 1 + y11 + y12
x2 = y21
x3 = y32

Given that the consumer does not value good 1 we look for a solution with x1 = 0, x2 > 0
and x3 > 0. The latter together with feasibility, imply also y21 > 0 and y32 > 0, leading
√
to y11 < 0 and y12 < 0. Efficiency of the production process leads to y21 = −y11 and
√
y32 = 2 −y12 . Deriving the related first order conditions, we obtain

M U 1 = 0 ≤ µ1 , M U 2 = 1 = µ2 , M U 3 = 1 = µ3


1
−γ1 2√−y + µ1 = 0, −γ1 + µ2 = 0,




 11
1
−γ2 √−y12 + µ1 = 0, −γ2 + µ3 = 0


 0 = 1 + y11 + y12

 x = y21
 2


x3 = y32

where µl is the multiplier associated with the feasibility of good l and γj is the multiplier
associated with the production frontier of firm j. Solving the above gives

 µ1 ≥ 0, µ3 = 1, µ3 = 1
 1 µ µ 1
 2√−y11 = µ12 = µ1 = µ13 = √−y12


0 = 1 + y11 + y12
x = y21


 2


x3 = y32 ,

and, using the second and third equation to solve for y11 (or for y12 ) we find


 µ1 ≥0, µ3 = 1,µ3 = 1

 x = 0, √15 , √45


 
1 √1
y = − ,
 1  5 5



 y2 = − 4 , √4 .

5 5

We have found the Pareto optimal allocation that coincides with the competitive equilib-
rium allocation. In principle, there could be other Pareto optimal allocation where the
consumer consumes only one good, for example only good 3. Feasibility of this allocation
would correspond to firm j = 1 not producing, y21 = 0, leading to y11 = 0 and thus to

M U 1 = 0 ≤ µ1 , M U 2 = 1 ≤ µ2 , M U 3 = 1 = µ3


−γ1 √−y111 =0 + µ1 ≥ 0, −γ1 + µ2 ≤ 0 ⇒ µ1 ≥ +∞





−γ2 √−y112 =1 + µ1 = 0, −γ2 + µ3 = 0 ⇒ µ1 = 1


 y21 = −1

 x = y21 = y11 = 0
 2


x3 = y32

Similarly, we have no solutions also when the consumer consumes only good 2. We can
conclude the the Pareto set contains only the allocation found above with x1 = 0, x2 >
0, x3 > 0.
Solution of Exercise 19
Find the competitive equilibrium and the Pareto set in an economy as above with U (x2 , x3 ) =
x2 x3 and Y2 = {y12 + y32 ≤ 0, y12 ≤ 0} (firm 1 keeps the same technology).

First note that given monotonicity of consumer preferences (and thus LNS) the I wel-
fare theorem applies and we can search for the competitive equilibrium in the Pareto Set.
Having a unique consumer, there is no issue of distributing resources among consumer and
we expect all the allocations of the Pareto Set to be competitive equilibria (together with
supporting prices). To find the Pareto Set we solve

Max x2 x3 such that

x ∈ R3+
√
y11 ≤ 0, 0 ≤ y21 ≤ −y11
y12 ≤ 0, 0 ≤ y32 ≤ −y12
x1 = 1 + y11 + y12
x2 = y21
x3 = y32

Given that the consumer does not value good 1 we look for a solution with x1 = 0, x2 > 0
and x3 > 0. The latter together with feasibility, imply also y21 > 0 and y32 > 0, leading
√
to y11 < 0 and y12 < 0. Efficiency of the production process leads to y21 = −y11 and
y32 = −y12 . Deriving the related first order conditions, we obtain

 M U 1 = 0 ≤ µ 1 , M U 2 = x3 = µ 2 , M U 3 = x2 = µ 3
1




 −γ1 2√−y 11
+ µ1 = 0, −γ1 + µ2 = 0,
−γ2 + µ1 = 0, −γ2 + µ3 = 0


 0 = 1 + y11 + y12
x = y21


 2


x3 = y32

where µl is the multiplier associated with the feasibility of good l and γj is the multiplier
associated with the production frontier of firm j. Solving the above gives the following
Pareto optimal allocation  1 2
 x1 = 0, x2 = √3 , x3 = 3
y11 = − 13 , y21 = √13
y12 = − 23 , y32 = 23


Note that having x2 = 0 (or x3 = 0) gives the consumer the lowest level of utility and
thus cannot be maximal. Having have the Pareto optimal allocation of this economy, we
have a candidate for the allocation of a competitive equilibrium (given for granted that a
competitive equilibrium exists) if we find supporting prices. The latter are given my the
marginal ratio of substitution and by the marginal rate of transformation computed at the
Pareto optimal allocation (x∗ , y ∗ ). We find
p2 x∗ 2 p1
= M RS23 = 3∗ = √ and 2
= M RT13 = 1.
p3 x2 3 p3
Thus, the competitive equilibrium is
 ∗
p∗2
     
p1 2 ∗ 1 2 ∗ 1 1 ∗ 2 2
CE = = 1, ∗ = √ , x = 0, √ , , y1 = − , √ , y2 = − , .
p∗3 p3 3 3 3 3 3 3 3

Solution of Exercise 20
Consider an economy with two consumption goods, L = 2, two consumers, I = 2, and one
firm, J = 1. Consumer 1, the worker, has endowment ω1 = (1, 0) and utility U1 (x1 ) =
x11 x21 . Consumer 2, the capitalist, is the owner of the firm, θ21 = 1, and has utility
√
U2 (x1 ) = x12 x22 . The firm has production set Y = {y1 ≤ 2 −y2 ≤ 0, y1 < 0}
1. Provide the definition of a competitive equilibrium for this economy.

2. Find the set of competitive equilibria and the set of Pareto optimal allocations.

3. Discuss the validity of both Welfare Theorems.

4. Find the competitive equilibrium with transfer such that both consumers consume the
same amount of output. How can you implement such a transfer starting from the
given endowment/ownership?
1. This is standard.
2. To find the set of competitive equilibria we first derive firm optimal supply and demand
of input, as well as consumers optimal demands. Starting from the firm we have
M ax p2 y2 + p1 y1
y1 ≤ 0
√
y2 ≤ 2 −y1

leading to  2
p2 p2 π(p) p2
y1 (p) = − , y2 (p) = 2 , = .
p1 p1 p2 p1
Consumers optimal demand is
   
1 1 p1 1 π(p) 1 π(p)
x1 = , and x2 = , .
2 2 p2 2 p1 2 p2
Market clearing of good 1 leads to
 2  2
1 1 p2 p2 p1 √
+ =1− ⇒ = 3.
2 2 p1 p1 p2
The corresponding equilibrium allocation is
√ !    
∗ 1 3 ∗ 1 1 ∗ 1 2
x1 = , , x2 = , √ y = − ,√
2 2 6 2 3 3 3

(note that also the market for good 2 clears). As for Pareto optimality, strong monotonicity
of both consumers preferences, allows us to pick either agent as the one whose utility is to
be maximized, given different utility levels of the other agent. Choosing agent 1, we have
to solve
M ax U1 (x1 )
x1 ∈ R + +
2 , x2 ∈ R2
x11 + x12 = 1 + y1
x21 + x22 = y2
y1 ≤ 0√
y2 ≤ −y1
U2 (x2 ) ≥ Ū
for given levels of Ū . The first order conditions at an interior and non wasteful allocation
give
1
= xx11 = xx22 2

 M RS1,2 21
12
= M RS1,2
= xx11 1
 1
 M RS1,2 21
= √−y = M RT1,2


1
x11 + x12 = 1 + y1

 x21 +√ x22 = y2


y2 = −y1


Solving the above gives

√ √
      
2 2 1 2 2
P S = x1 = (t, 3t), x2 = − t, √ − 3t , y = − , √ , t ∈ 0,
3 3 3 3 3
where we have extended the solution to the border cases where either consumer consumes
all the output of firm 2.

3. See the answer to point 2 in the exercise above. Note that in this economy the produc-
tion function is concave (and thus the production set convex)
4. The allocation we are looking for solves


 x21 = x√22


 x 21 = 3x11
 y = −1

1 3
y 2 = √2


 3
 x11 + x11 = 1 + y1



x21 + x22 = y2

leading to    
1 1 1 1
x01 = ,√ and x02 = ,√
3 3 3 3
Such allocation is a competitive equilibrium together with
p1 1 2
√
= M RS1,2 = M RS1,2 = M RT1,2 = 3
p2
√
or p = ( 3p, p). The transfer that support such an equilibrium is thus
 0 2
√
 px1 = pω1 + T1 ⇒ √3 p = 3p + T1
px0 = π(p) + T2 ⇒ √23 p = √13 p + T2
 2
T1 + T2 = 0

giving T1 = −T2 = − √13 p. Such a transfer can be implemented using good 1, in this case
the social planner needs to transfer
T1 T1 1
t= = √ =− .
p1 3p 3
from consumer 2 to consumer 1. Given that consumer 1 is endowed with 1 unit of the first
good, such transfer is possible and leads to

ω10 = ω1 + (−1/3, 0) = (2/3, 0) and ω20 = ω2 + (1/3, 0) = (1/3, 0).

Alternatively, the transfer could be implemented by transferring ownership of the firm:

T1 1
θ0 = = −√ .
π(p) 3
However, consumer one is not endowed with any share of the firm and thus such transfer
can not be implemented in this economy.
Solution of Exercise 21
Consider an exchange economy with two consumption goods, L = 2, and two consumers,
I = 2. Consumer 1 has endowment ω1 = (2, 2 − δ) with δ ∈ (0, 2) and utility U (x1 ) =
x11 + x21 for  > 0. Consumer 2 has endowment ω2 = (2, δ) and utility U (x2 ) = x12 + x22 .
 1. Consider  = δ = 0. Find the set of competitive equilibria and the set of Pareto
optimal allocations. Discuss the validity of both Welfare Theorems in this economy.

2. Consider  = δ = 0. State whether p = (0, 1), p = (1, 1), p = (1, 2), p(2, 1) support
x∗1 = (0, 2), x∗2 = (4, 0) as a quasi-equilibrium transfer. If so, is each {p, x∗ } also a
competitive equilibrium?

3. Consider  > 0 and δ = 0. Find and plot the excess demand z1 (p) = x11 (p) +
x12 (p) − ω̄1 of good 1, the set of competitive equilibria, and the set of Pareto optimal
allocations. Discuss the validity of both Welfare Theorems in this economy.

4. Consider  = 0 and δ ∈ (0, 2). Find and plot the excess demand function of good 1,
the set of competitive equilibria, and the set of Pareto optimal allocations. Discuss
the validity of both Welfare Theorems in this economy.

1. The Pareto set is the set of allocations where consumer 1 consumes only good 2. Both
See also the lecture notes of Tuesday November 26.
2. The given allocation x∗ belongs to the Pareto Set, and due to convexity of both agents
preferences the II welfare theorem apply. p = (0, 1) supports a quasi-equilibrium but not
an equilibrium at x∗ . Both p = (1, 1) and p = (1, 2) support an equilibrium (and thus also
a quasi-equilibrium). p = (2, 1) does not support x∗ as quasi-equilibrium. 3. The excess
demand is
when pp12 <  < 1
 2

 p1
2 p
 
 −2,  when p2 = 

 1

z1 (p) = −2 when pp12 ∈ (, 1)
p1
 [ − 4, −2] when =1


p2


 −4 p1
when p2
>1
The equilibrium is achieved for p such that z1 (p) = 0, giving

CE = {p1 /p2 = , x∗1 = (2, 2), x∗2 = (0, 2)} .

The Pareto set is the set of allocations where agent 1 has only good 2 or where agent 2
has only good 1. Both welfare theorems apply.
4. The excess demand is
when pp21 ∈ (0, 1)
 δ
 p1 − 2
p1
z1 (p) = [ − 2 − δ, −2] when p2
=1
p1
−4 when >1

p2

The equilibrium is achieved for p such that z1 (p) = 0, giving

CE = {p1 /p2 = δ/2, x∗1 = (0, 2), x∗2 = (4, 0)} .

The Pareto set is the set of allocations found in 1.

Solution of Exercise 22
Consider an economy with two consumption goods, L = 2, one consumer, I = 1, and one
firm, J = 1. Consumer 1 has endowment ω1 = (2, 1) and utility U (x1 ) = x11 x21 . The
firm has production set Y = {y1 + y2 ≤ 0 if y1 ≤ −a, y2 ≤ 0 if y1 ∈ (−a, 0]} with a ≥ 0.
Consumer one is the owner of the firm. Fix a = 1.

1. Provide the definition of a competitive equilibrium for this economy.

2. Find the set of competitive equilibria.
3. Find the set of Pareto optimal allocations.
4. Discuss the validity of both Welfare Theorems.
The production technology is linear but there are positive set-up costs when a > 0. Positive
set-up costs make the production set non-convex and the II welfare Theorem cannot be
applied. The numerical solution is provided in the solution of Exercise 5.
Solution of Exercise 23
Solve Ex. 23 for each given a ∈ [0, 1].

When a = 0 the firm has a standard linear production function. The firm is active, and
indifferent on the efficient frontier, when pp12 = 1. For such prices, given the endowment
wants to consume 32 , 32 leading to an equilibrium where the


and zero profit, the consumer
firm choose y = − 12 , 12 .

The same equilibrium is present as long as a ≤ 12 . At a = 12 the equilibrium coincides with

the minimum output of the firm.
As long as a increases, there is no equilibrium. In fact, at the allocation (2 − a, 1 + a) the
consumer would have be happy only with a price ratio larger than one, but for such price
ratio the firm would not want to produce. The consumer would be happy to consume her
endowment for a price ratio lower than one, but for such price ratio the firm would want
to have an infinite output. The allocation is however Pareto optimal. There is no other
feasible allocation where the consumer is better off. For a = 1 there is another Pareto
optimal allocation, the endowment in fact
U (2 − a, a) = U (ω)
is solved by a = 1. The first welfare theorem always applies. The second welfare theo-
rem applies only when a = 0. Note that the II welfare theorem provides only sufficient
conditions for a Pareto optimal allocation to be a (quasi)-equilibrium with transfer. As a
result, it is well possible that a Pareto optimal allocation is supported

as a competitive
equilibrium even if the production set is not convex, as when a ∈ 0, 12 .
Solution of Exercise 24
Consider an exchange economy with I = {1, 2}, L = {1, 2}, U1 (x1 ) = x21 , U2 (x2 ) =
x12 + x22 , and ω1 = (1, 0), ω2 = (0, 1).
 1.? Find the utility possibility set

2.? Find the Pareto frontier

3.?? Show that for some levels of Ū2 the solutions x∗ of

Max U1 (x1 )
x ∈ R4+
x1 + x2 ≤ ω̄
U2 (x2 ) ≥ Ū2

are not Pareto optimal (and thus (U1 (x∗1 ), U (x∗2 )) does not belong to the Pareto fron-
tier).

1. To find the UPS find (U1 , U2 ) such that



 xli ≥ 0 ∀l, i
 x11 + x12 ≤ 1


x21 + x22 ≤ 1
 U1 (x1 ) = x21 ≥ U1



U2 (x2 ) = x12 + x22 ≥ U2


The set can be written as 

 U1 ≤ 1
U2 ≤ 2
U1 + U2 ≤ 2


2. The Pareto frontier is the subset of the Pareto set where Ui cannot be increased without
increasing Uj , j 6= i. This is the segment {U2 = 2 − U1 , U1 ∈ (0, 1)}.
3. See the lecture notes of Tuesday November 26.
Solution of Exercise 25
Consider an economy with two commodities, two consumers, and one firm. Consumer
1 has endowment ω1 = (6, 0), owns half of the firm, and has utility U1 (x1 ) = x11 x21 .
Consumer 2 has endowment ω2 = (4, 2), owns the other half of the firm, and has utility
U2 (x2 ) = x12 x22 . The firm uses commodity 1 to produce commodity 2 and has production
function f (z) = z with z ≥ 0 (the amount of commodity 1).

1. Derive supply and profit of the firm.

2. Derive the demand of both consumers.

3. Provide the definition of a competitive equilibrium for this economy.

4. Compute the unique competitive equilibrium of this economy.

5. Is the competitive equilibrium found above Pareto optimal? Reply without computing
the Pareto set.
 6. Provide a definition of Pareto optimal allocation.
7. Explain how to find all the Pareto optimal allocation of this economy (write down the
related maximization problem and the first order conditions for an interior solution).
TO BE PREPARED
Solution of Exercise 26
Solve Exercise 6.B.2 in the MWG textbook.

We have to show that if a preference relation over lotteries that can be represented by
an utility function that has an expected utility form, then its satisfies the independence
axiom. We shall exploit the linearity of the EU form.
Consider two lotteries L = {C; p} and L0 = {C; p0 } with
L < L0 .
The EU form implies that the latter is equivalent to
X X
pn u(Cn ) ≥ p0n u(Cn ).
n n

Take now any other lottery L00 = {C; p00 } and any α ∈ (0, 1). The former inequality implies
also
X X X X
α pn u(Cn ) + (1 − α) p00n u(Cn ) ≥ α p0n u(Cn ) + (1 − α) p00n u(Cn ).
n n n n

By linearity of the EU form, the latter can be re-written as

X X
[αpn + (1 − α)p00n ]u(Cn ) ≥ [αp0n + (1 − α)p00n ]u(Cn ),
n n

which, by definition of compound lottery, is equivalent to

αL + (1 − α)L00 < αL + (1 − α)L00 .

Solution of Exercise 27
Show that a preference relation on lotteries that can be represented in an Expected Utility
form satisfies transitivity.

Consider three lotteries L = {C; p}, L0 = {C; p0 }, and L00 = {C; p00 } with
L < L0 and L0 < L00 .
The latter, together with the fact that the preference relation can be represented in an EU
form, imply, respectively,
X X X X
pn u(Cn ) ≥ p0n u(Cn ) and p0n u(Cn ) ≥ p00n u(Cn ).
n n n n
Thus X X X
pn u(Cn ) ≥ p0n u(Cn ) ≥ p00n u(Cn ),
n n n

so that X X
pn u(Cn ) ≥ p00n u(Cn ).
n n

Given the EU form, the latter implies

L < L00 .

Solution of Exercise 28
Consider the space of simple lotteries L = {C; (p1 , p2 , p3 )}. Show graphically that if a pref-
erence relation on lotteries satisfies the independence axiom, then indifference curves are
parallel. (hint: show that if two indifference curves are not parallel, then the independence
axiom is violated)

The solution is at page 175 − 176, figure 6.B.5. panels a) and c) of the MWG textbook.

Solution of Exercise 29
Solve Exercise 6.B.4 in the MWG textbook.

First not that given the set of outcomes C = {A, B, C, D} with

A No flood and no evacuation

B No flood and evacuation

C Flood and evacuation

D Flood and no evacuation

the expected utility of a generic lottery L = {C, ; p} is

X
U (L) = pi u i
i=A,B,C,D

with

pA Probability of no flood and no evacuation

pB Probability of no flood and evacuation

pC Probability of flood and evacuation

pD Probability of flood and no evacuation

(a) To find the EU form we need to assign values to ui , i = A, B, C, D. We shall use,
the revealed preferences of the decision maker. In particular we have, for p ∈ (0, 1) and
q ∈ (0, 1),

uB = puA + (1 − p)uD and uC = quB + (1 − q)uD and uA > uD .

It follows that uA > uB > uC > uD . Without loss of generality, we can normalize uD = 0
and uA = 1 and get
uB = p and uC = pq.
(b) To evaluate each criterion we have to compute the corresponding probabilities. The
probability of having a flood is given and equal to π = 0.01, so that the probability of not
having a flood is 1 − π = 0.99. Name pN the probability of a necessary evacuation and
pU the probability on an unnecessary evacuation as implied by a criterion. The resulting
probabilities over outcomes are

(pA , pB , pC , pD ) = ((1 − pU )(1 − π), pU (1 − π), pN π, (1 − pN )π)

Applying the above to the criterion 1 and 2 we have, respectively

 
1 90 99 10 99 90 1 10 1
p = , , ,
100 100 100 100 100 100 100 100

and  
2 95 99 5 99 95 1 5 1
p = , , , .
100 100 100 100 100 100 100 100
Note that the second criterion implies a higher probability of necessary evacuation and
also a lower probability of unnecessary evacuation, we thus expected that the second is
always better than the first. Using the EU representation found in (a) and computing the
difference in utility between the two criteria, we find
5 99 5 1
U (L2 ) − U (L1 ) = (1 − p) + pq.
100 100 100 100
The latter, being strictly positive for all p and q in (0, 1) confirms our intuition: criterion
2 is always preferred to criterion 1.

Solution of Exercise 30
Solve Exercise 6.B.7 in the MWG textbook. (Note: a preference over lotteries is monotone
if, given two real outcomes C1 > C2 , then the lottery with sure outcome C1 is strictly pre-
ferred to the lottery with sure outcome C2 )

The decision maker is indifferent between L and the sure amounts xL , so that

L ∼ xL ⇒ xL < L,
and between L0 and the sure amounts xL0 , so that
L ∼ xL ⇒ L0 < xL0 .
(note that above with xL or xL0 we denote the lottery that give the sure amount for sure).
Assume now that the decision maker strictly prefers L to L0 , then by transitivity
xL  xL0 .
Given monotonicity, the latter is implied by
xL > xL0 .
To show the “only if” case we also need that when two real outcomes are equal, then the
decision maker is indifferent between the two lotteries having that amount for sure. The
latter is always the case as the two lotteries coincide.

Solution of Exercise 31
When faced with the choice between the lottery A = {(3500, 2800, 0); p = (0.3, 0.66, 0.04)}
and B = {3500; p = 1}, a decision maker chooses lottery B. When, instead, he is
asked to choose between the lottery A0 = {(3500, 2800, 0); p = (0.3, 0, 0.7)} and lottery
B 0 = {(3500, 2800, 0); p = (0, 0.34, 0.66)}, he chooses A0 . Say whether the decision maker’s
preferences are consistent with the expected utility form, explaining how you have reached
your conclusion. What if lottery B becomes B 00 = {2800; 1}?

Given the EU form of the utility that represents the decision maker preferences, we can
write, for a generic lottery L = {(3500, 2800, 0); p}
U (L) = p3500 u3500 + p2800 u2800 + p0 u0 .
Without loss of generality we can normalize u0 = 0 and u3500 = 1 and get
U (L) = p3500 + p2800 u2800 .
Using this representation of the utility, having U (B) > U (A) implies the following restric-
tion on u2800
70
U (B) = 1 > U (A) = 0.3 + 0.66u2800 ⇒ u2800 < .
66
Repeating the same reasoning for A0 and B 0 , we find
30
U (A0 ) = 0.3 > U (B 0 ) = 0.34u2800 ⇒ u2800 < .
34
Any value of u2800 that satisfies both inequalities, e.g. u2800 = 0.5, gives an expected utility
representation of consumer’s preferences.
If we replace lottery B with B 00 , the first inequality becomes
30
U (B 00 ) = u2800 > U (A) = 0.3 + 0.66u2800 ⇒ u2800 > ,
34
and no values of u2800 such that preferences are consistent with an EU form exist.
Solution of Exercise 32
We consider an expected-utility decision-maker facing the following possible professional
occupations: working in the financial industry (A), working in the movie industry (B),
working in the car industry (C). We further assume that the decision-maker can apply to
three different schools.

• After School 1, he is certain to find a job in the car industry.

• After School 2, he is certain to find a job in the financial industry.

• After School 3, he would find a job in the movie industry with probability 0.1 and in
the financial industry with probability 0.9.

We assume that School 2 is the least preferred option of the decision-maker and that he is
indifferent between School 1 and School 3.

1. Give a representation of the utility function of this decision-maker.

2. We assume that a new school (School 4) opens. After school 4, a student would find
a job in each industry with probability 31 . How would the new school rank compared
to Schools 1, 2, 3?

3. There are too many application in School 4. The ministry of education decides that
rather than applying to School 4 directly, the decision-maker must apply to a lottery
that leads to admission in School 4 with probability α ∈ [0, 1] and to admission in
School 2 with probability 1 − α. For which value of α does the decision-maker prefer
to apply to School 1 rather than to the lottery that may lead to admission to School
4?

1. Given the three outcomes, {A, B, C}, to each school there corresponds a lottery that
can be identified with the following probability vectors

L1 = (0, 0, 1), L2 = (1, 0, 0), L3 = (0.9, 0.1, 0).

The decision maker evaluations are thus

U (L1 ) = uC , U (L2 ) = uA , U (L3 ) = 0.9uA + 0.1uB .

Thus, having U (L1 ) = U (L3 ) > U (L2 ), implies

uC = 0.9uA + 0.1uB > uA .

Without loss of generality we can take uA = 0, uB = 1, uC = 0.1.

2. The new school corresponds to the lottery L4 = (1/3, 1/3, 1/3). Using the values found
above, we find
1 1 1 11 1
U (L4 ) = + = > = U (L1 ) = U (L3 ),
3 3 10 30 3
so that the new school is the most preferred.
3. The new procedure corresponds to the compound lottery

αL4 + (1 − α)L2 .

By linearity of the EU, the utility of the latter is easily found as

11
U (αL4 + (1 − α)L2 ) = αU (L4 ) + (1 − α)U (L2 ) = α .
30
The decision maker prefers school 1 rather than the compound lottery when
1 11 3
U (L1 ) = > α = U (αL4 + (1 − α)L2 ) ⇔ α< .
10 30 11

Solution of Exercise 33
We consider an expected-utility decision-maker with utility of the form u(x) = xa with
a < 1.
1. Let X be a lottery whose outcome is uniformly distributed over [0, 1]. Determine
E(u(X)).

2. Let Y be a lottery whose outcome is 0 with probability 1/3 and 1 with probability 2/3.
Determine E(u(Y )).

3. Determine the value a∗ (of a) for which the decision-maker is indifferent between X
and Y .

4. If a > a∗, which lottery, X or Y, is preferred by the decision-maker?

5. Determine, as a function of a, the coefficient of absolute risk-aversion of the decision-

maker.
TO BE PREPARED

Solution of Exercise 34
Solve Exercise 6.C.1 in MWG assuming u(x) = log(x) and find? for which value of q the
agent does not insure.

Let us consider u(x) = log x first. From the first order conditions one gets
π(1 − q) q(1 − π)
− ≤0
w − αq − D + α w − αq
with equality when α > 0. Let us first check that α = D > 0 is not a solution. Substituting
α = D in the above, one gets
π=q
which contradicts that the insurance premium q is higher than π. To show that the solution
must have α < D rewrite the FOC as
π 1−q 1 1
≤ .
q 1 − π w − αq − D + α w − αq
If the solution is interior, α > 0, then equality of the latter and q > π imply
1 1
> ⇔ α < D.
w − αq − D + α w − αq
If the solution is not interior, α = 0, then also α < D.
To find the conditions on q such that the agent does not insure we must find bounds on
the value of q such that the FOC with inequality is satisfies at α = 0. Assuming positive
final wealth, the FOC with inequality can be re-written as
1−π q−π
α≥D −w .
1−q q(1 − q)
With α = 0, and solving for q, we get
w
q≥π .
w − D(1 − π)
Note that the higher the loss, the higher the insurance price beyond which the agent does
not insure.

Solution of Exercise 35
Consider an investor who has to choose between two assets. Asset A has payoff xA =
(12, 6, 9), all with equal probability. Asset B is risk free with payoff xB = (x̄, x̄, x̄).
1. Assume u(x) = log(x). Find the certainty equivalent of xB .

2. Find the level of x̄ such that the investor is indifferent between asset A and asset B.

3. Assume now that x̄ = E[xA ], find how many units hA of asset A should be given
to the agent to make him indifferent between hA and one unit of the risk free asset.
Provide an intuition.

4. Compute the coefficient and absolute risk aversion.

The certainty equivalent is 8.653. The average is 9.

Solution of Exercise 36
A decision maker has preferences over lotteries represented by an expected utility with
Bernoulli utility u(x). Consider the lottery L = {(0, 4, 9); (1/6, 1/2, 1/3)}
1. Provide the definition of certainty equivalent of L given u.
 √
2. Compute the certainty equivalent for L given u(x) = x.

3. Compute the coefficient of absolute risk aversion given u and x.

4. Assume that the agent starts with w = 12 and owns the lottery. Find the minimum
price he is willing to accept to sell the lottery.

5. Compare the minimum price found above and the certainty equivalent and provide an
intuition of their difference based on the coefficient of absolute risk aversion.

TO BE PREPARED
Solution of Exercise 37
Solve Exercise 6.C.18 in MWG.

TO BE PREPARED

Solution of Exercise 38
Solve Exercise 6.C.16 in MWG assuming from the beginning
√ the parametrization of point
(d) with p = 0.5. Other than the proposed u(x) = x, consider also u(x) = x and
u(x) = log x. Give and interpretation in terms of the certainty equivalent and the change
of the risk aversion coefficient with wealth.

TO BE PREPARED