Econ 210A PS1 Answer Key

1 Jehle-Reny
1.2 (a) In order to prove this, we must show that every element in  is
in the set . Let x, y ∈ X s.t. x  y. Therefore, x, y ∈. By
definition, x  y and x, y ∈. Therefore ⊂
(b) In order to prove this, we must show every element in ∼ is an
element in . Let x, y ∈ X s.t. x ∼ y. If x ∼ y then, we have
by definition x  yAN Dy  x . Since x  y, then x, y ∈ and
∼⊂
(c) In order to prove this we must show containment both ways. That
is, ( ∪ ∼) ⊂ and ⊂ ( ∪ ∼). Starting with ( ∪ ∼) ⊂:
Let x, y ∈ ( ∪ ∼). Then by definition, x, y are either a member
of  OR ∼. Therefore, by definition either, x  y AND NOT
(y  x) OR x  y AND y  x. In either case, x  y and
x, y ∈. Proving the other containment: let x, y ∈ s.t. X  y.
There are two possibilities between the relationship between y
and x. Either, y  x or NOT(y  x) (these are complementary
scenarios). Since we have x  y AND (y  x OR N OT (y  x)),
x, y ∈ ( ∪ ∼) by definition and ⊂ ( ∪ ∼). Since we have
shown containment both ways, = ( ∪ ∼).
(d) To show the set is empty, we do this by contradiction. Assume
that there exists x, y ∈ ( ∩ ∼). Then, (x  y AND NOT y  x
AND y  x) AND (x  y AND y  x). But, one cannot have
NOT(y  x) and y  x. Therefore, there ∃ no x, y ∈ ( ∩ ∼)

1.3 (a) For any x, y ∈ X where x ∼ y we can have neither x  y nor

y  x. For any x, y ∈ X where x  y, we can have neither x ∼ y
nor y ∼ x.
(b) Suppose x1  x2 . Then N OT x2  x1 . This means that N OT x2 
x1 OR x1 ∼ x2 . The case of x2  x1 can be shown by symme-
try. Suppose x1 ∼ x2 . Then it cannot be true that x1  x2 or
x2  x1 , since a necessary condition for either relation is that
N OT x1 ∼ x2 .

1.4 (a) : Let x1 , x2 , x3 ∈ X and assume that x1  x2 and x2  x3 .

Show that x1  x3 .

1
 This implies that x1  x2 and x2  x3 and by transitivity of ,
x1  x3 . It is also true that x2 N OT  x1 and x3 N OT  x2 (∗) .
Assume that x3  x1 . This is necessary if x3  x1 . This implies
that x3  x2 by transitivity. But this contradicts the earlier
statement.
(b) Suppose x ∼ y and y ∼ z.
x ∼ y implies that x  y and y  x. y ∼ z implies that y  z
and z  y.
By transitivity of , x  y and y  z implies that x  z.
Also, because z  y and y  x, z  x.
Since x  z and z  x, it must be that x ∼ z.
(c) Finally, show that if x1 ∼ x2  x3 , then x1  x3 .
Assume x1 ∼ x2 and x2  x3 .
If x1 ∼ x2 ⇔ x1  x2 and x2  x1 . Since x1  x2 and x2  x3 ,
by transitivity of , then x1  x3 .

1.5 (b) We must again show containment both ways. Let y ∈ (x0 ).
Then, by definition y  x0 . Further, there are two possibilities
either x0  y OR NOT(x0  y) because these are complementary
qualities. Thus, by definition y ∈∼ (x0 ) ∪  (x0 ). Let y ∈∼
(x0 ) ∪  (x0 ). Then y  x AND x  y OR y  x AND NOT
x  y. But we already have that y  x, so y ∈ (x0 ) and we
have containment both directions and the sets are equal.
(c) By contradiction. Suppose y ∈∼ (x0 ) ∩  (x0 ). Then y  x
AND x  y AND y  x AND NOT x  y. Since x  y AND
NOT x  y can’t be both true it is a contradiction.
(g) Contradiction. Suppose x ∈ X,, but x ∈ / y ∈∼ (x0 ) ∪  (x0 ) ∪ ≺
(x0 ) . Then neither x  x0 , x ∼ x0 , and x ≺ x0 . By definition,
neither x  x0 nor x  x0 and this violates the completeness of
the preference relationship.

1.6 Many examples, but something like when consumer goods are indivis-
ible i.e. cars and refrigerators.

1.7 Let  represent convex preferences.
Let x0 , x1 and x2 ∈ X where

x1  x2  x0 . Then x1 , x2 ∈  x0 . For any t ∈ [0, 1], consider the
bundle tx1 + (1 − t) x2 . By convexity of , tx1 + (1 − t)
x2  x0 , which
is true if f tx + (1 − t) x ∈ x . Therefore,  x0 is a convex set
1 2 0

for any x0 ∈ X.

2
 1.8 Case 1: Take two points, xa and xb along the indifference curve. To
show that the preferences are convex, it is seen that for any t ∈ [0, 1],
txa + (1 − t) xb  xa . We have shown Axiom 5 in this case. We
can take this a step further and say that txa + (1 − t) xb  xb and
txa + (1 − t) xb ∈∼ (xa ). Because any two points on the line, txa +
(1 − t) N OT xb  xa ⇒the preferences are not strictly convex. We
have shown how Axiom 5’ does not apply.
Case 2: Take two points, xa and xb such that xa  xb and let t ∈ (0, 1).
Then by construction, xa lies to the northeast to xb . Since t > 0,
that implies txa + (1 − t) xb lies to the northeast of xb . Therefore,
txa + (1 − t) xb  xb and we have shown Axiom 5.

1.9
Strict Monotonicity Axiom 4: ”Preferences increase northeasterly” translates to if xa  xb ,then
xa  xb . The fact that ”indifference sets are parallel right angles
that ’kink’ on the line x1 = x2 ” along with with ”preferences
increase northeasterly” implies that if xa  xb . These facts assert
that  satisfy strict monotonicity.
Strict Convexity Axiom 5: Pick any two point along a ”leg” of an indifference
 curvesuch as
(α, α) and (α, β), where α < β. The point α, 12 (α + β) lies in

3
 between these two and on the same indifference  curve
 as (α, α)
1
and (α, β). Therefore, we cannot have α, 2 (α + β)  (α, α),
illustrating that these preferences are not strictly convex.
Convexity Axiom 5’: Consider any x, y ∈ X ⊂ R2 such that x ∼ y. Given the na-
ture of these preferences, it must be true then that min [x1 , x2 ] =
min [y1 , y2 ] . For any t ∈ [0, 1] consider the point tx + (1 − t) y. If
we can show that min [tx1 + (1 − t) y1 , tx2 + (1 − t) y2 ] ≥ min [x1 , x2 ] =
min [y1 , y2 ] , then we have shown that these preferences are con-
vex.

A.1.2 a Let x ∈ S then we must show x ∈ S ∪ T . S ∪ T contains all

elements in S and T . Therefore, since x ∈ S, x ∈ S ∪ T .
b Same proof
c Let x ∈ S ∩ T . That implies that x is in both S and T . Therefore
x∈T
c Same proof
A.1.5 Let A = [a1 , a2 ] and B = [b1 , b2 ], where a2 < b1 . Since ta2 +(1 − t) b1 ∈
/
A ∪ B for t ∈ (0, 1) , A ∪ B is not convex.
x
A.1.7  convex. For example, (0, 1) and (1, e) ∈ {(x, y) |y = e },
a Thisset is not
but 12 , e+1
2 / {(x, y) |y = ex } .
∈

4
 b This set is convex.
Let (x1 , y1 ) , (x2 , y2 ) ∈ S = {(x, y) |y ≥ ex } . Since y = ex is a con-
tinuous function, it is sufficient to show that (tx1 + (1 − t) x2 , ty1 + (1 − t) y2 ) ∈
S forany particular t ∈ (0, 1). 
Set t = 12 . Our task is to show
1
that 2 (x1 + x2 ) , 12 (y1 + y2 ) ∈ S. 1
2 (y1 + y2 ) ≥ 1
2 (ex1 + ex2 ),
1
since yi ≥ exi for i = 1, 2. Also, 21 (ex1 + ex2 ) ≥ e 2 (x1 +x2 ) =
x1 x2 x1 x2 x1 x2
e 2 · e 2 ⇔ ex1 + ex2 ≥ 2e 2 · e 2 ⇔ ex1 − 2e 2 · e 2 + ex2 ≥ 0 ⇔
(ex1 − ex2 )2 ≥ 0.
   
1 1 9 1
c This set is not convex. For example, 10 , 2 , 1 10 ,2 ∈ S =
  
(x, y) |y ≥ 2x − x2 ; x > 0, y > 0 . However, 1, 12 = 1 1 1

  2 10 , 2 +
1 9 1
2 1 10 ,2 ∈
/ S.
(d) This set is convex.
Consider any (x1 , y1 ) , (x2 , y2 ) ∈ S = {(x, y) |xy > 1, x, y > 0}.
For any t ∈ [0, 1],
(tx1 + (1 − t) x2 ) (ty1 + (1 − t) y2 ) = t2 x1 y1 +t (1 − t) (x1 y2 + x2 y1 )+
(1 − t)2 x2 y2
> t2 + (1 − t)2 + t (1 − t) (x1 y2 + x2 y1 ), since xi yi > 1.
= 1 + 2t2 − 2t + t (1 − t) (x1 y2 + x2 y1 )
= 1 + 2t (t − 1) + t (1 − t) (x1 y2 + x2 y1 )
= 1 + t (1 − t) (x1 y2 + x2 y1 − 2) ≥ 1 if f x1 y2 + x2 y1 ≥ 0
x1 y2 + x2 y1 = x1 y1 yy12 + x2 y2 yy21 − 2 > yy21 + yy12 − 2 ≥ 0
y1 − 2y1 y2 + y2 ≥ 0
(y1 − y2 )2 ≥ 0, which is always true and therefore, (tx1 + (1 − t) x2 , ty1 + (1 − t) y2 ) ∈
S which is convex.
  
1 1
e S is convex ⇐ 2 ln (x1 ) + ln (x2 ) ≤ ln 2 x1 + 12 x2
 
1 1
⇔ 2 ln (x1 x2 ) ≤ ln 2 x1 + 12 x2
 
⇔ (x1 x2 )1/2 ≤ 1
2 x1 + 12 x2
1/2
⇔ x1 − 2 (x1 x2 ) + x2 ≥ 0
1/2 2
 
1/2
⇔ x1 + x2 ≥ 0, which is always true.

A.1.8 R is not complete because there can be no R relation between any two
people who do not know each other.
R is not transitive, the obvious counter example being man R wife,wife
Rwife’s mom, but man not R wife’s mom. That is although a man

5
 may love his wife, and wife may love her mom, the man may not love
her mother in law.

A.1.9 Suppose there exists x such that x ∈ B but f (x) ∈

/ f (A). Since x ∈ B
and B ⊂ A, then x ∈ A. Then, f (x) ∈ f (A) which contradicts our
initial assumption.

A.1.10 Suppose there exists x such that x ∈ B f −1 (x) ∈

/ f −1 (A) . Since x ∈ A
and A ⊂ B, then x ∈ B. Then, f −1 (x) ∈ f −1 (A) which contradicts
our initial assumption.

A.1.16 (a) Let −x1 , −x2 ∈ −S. Then x1 , x2 ∈ S. Convexity of S implies

that for any t ∈ [0, 1], tx1 + (1 − t) x2 ∈ S, which implies that
− [tx1 + (1 − t) x2 ] = t (−x1 ) + (1 − t) (−x2 ) ∈ −S. Therefore,
−S is convex.
(b) Let x1 , x2 ∈ S − T . Then there are s1 , s2 ∈ S and t1 , t2 ∈ T ,
such that x1 = s1 − t1 and x2 = s2 − t2 . Since S and T are
convex, λs1 + (1 − λ) s2 ∈ S and λt1 + (1 − λ) t2 ∈ T for any
λ ∈ [0, 1]. Given this and the fact that
λx1 + (1 − λ) x2 = λ (s1 − t1 ) + (1 − λ) (s2 − t2 )
= [λs1 + (1 − λ) s2 ] − [λt1 + (1 − λ) t2 ], λx1 + (1 − λ) x2 ∈ S − T ,
illustrating that S − T is convex.

A.1.17 (a) Part 1-prove 2 convex sets intersect and form a convex set. See
theorem A1.1 in Jehle and Reny page 414.
Part 2-Show that additional sets formed from the intersection of
convex sets is also a convex set.
Define A12 = A1 ∩ A2 which is convex. A12 ∩ A3 must also
be convex from part 1. This can be done for An = ∩ni=1 Ai and
An ∩ An+1 .
(b) Suppose x ∈ XAi , x = (x1, x2 , ..., xn ) . Suppose that y ∈ XAi ,
y = (y1 , y2 , ..., yn ). The convex combination of x and y is then:
z = (tx1 + (1 − t) y1 , tx2 + (1 − t) y2 , ..., txn + (1 − t) yn ) . Since
txi + (1 − t) yi ∈ Ai ∀i, Z ∈ Xi=1n A and X n A is convex.
i i=1 i
(c) Proof: If x ∈ ΣAi , ∃x1 ∈ A1 , x2 ∈ A2 , ..., xn ∈ An , s.t. Σxi = x
and
y ∈ ΣAi , ∃y1 ∈ A1 , y2 ∈ A2 , ..., yn ∈ An , s.t. Σyi = y.
Therefore, tx1 + (1 − t) y1 ∈ A1 , tx2 + (1 − t) y2 ∈ A2 , ..., txn +
(1 − t) yn ∈ An and Σ (tx1 + (1 − t) y1 ) ∈ ΣAi

6
 t (Σxi ) + (1 − t) (Σyi ) ∈ ΣAi and tx + (1 − t) y ∈ ΣAi and ΣAi is
convex.
(d) See part (c).

A1.18 Proof: Suppose that Ω is not convex. The there exists x1 , x2 ∈ Ω,

x1 6= x2 and t ∈ (0, 1) for which tx1 + (1 − t) x2 ∈ / Ω. Thus for some
j ∈ {1, ..., m}, f (tx1 + (1 − t) x2 ) < 0 ⇒ a [tx1 + (1 − t) x2 ]+bj < 0.
j j

But this contradicts the fact that, since f j (x1 ) = aj x1 + bj ≥ 0 and

f j (x2 ) = aj x1 + bj ≥ 0 ⇔ t aj x1 + bj ≥ 0 and (1 − t) aj x2 + bj ≥
 

0 ⇔ t aj x1 + bj + (1 − t) aj x2 + bj = aj [tx1 + (1 − t) x2 ] + bj ≥ 0.
Hence, Ω must be convex.

2 Rubinstein
Problem 1 Show  satisfies property (1), which is saying that indifference curves
do not cross.
Consider any x, y ∈ X such that I (x) 6= I (y). Suppose that I (x) ∩
I (y) 6= 0, so that there is some z ∈ I (x) ∩ I (y). Then z ∈ I (x)
and z ∈ I (y), implying that z ∼ y and z ∼ x. From transitivity
it follows that z1 ∼ z2 for any z ∈ I (x) , z ∈ I (y), both z1 and
z2 ∈ I (x) ∩ I (y). Thus, I (x) = I (y) ,contradicting the premise that
I (x) 6= I (y). Therefore, I (x) ∩ I (y) = 0, if I (x) 6= I (y). Now
suppose that I (x) = I (y). Then x, y ∈ I (x) ∩ I (y) 6= 0.
Show  satisfies property (2), indifference sets are non-empty.
Proof: For any x ∈ X, x ∈ I (x). So for y = x, x ∈ I (y) .

Problem 2 Solved in class

Problem 4 Induction has three steps. First what are we inducting on? The size
of the set X.

Step 1 Base Case: i = 2. Let X have two elements x and y. Because

of the definition of asymmetry, WLOG, let xP y. Then we have
established an ordering and it is complete because both x and y
are described.
Step 2 Assume for cases up to i = n, that there is a complete ordering.
Step 3 Prove for case i = n + 1 that there is a complete ordering. Let
x1 , x2 , ..., xn+1 ∈ X of size n+1. We know that sets of size n have
a complete ordering. Lets take elements x1 , ...., xn and form a

7
new set. WLOG, we know that we can rank x1 P x2 P...xn . Now
construct a new set with xn+1 and x1 . By our base case, we
know that either x1 P xn+1 or xn+1 P x1 . If, xn+1 P x1 , then we
are done and we have our ordering xn+1 P x1 P x2 P...xn . If not,
we make a new set with x2 and xn+1 which we know we can
rank. If, xn+1 P x2 , then we are done and we have our ordering
x1 P xn+1 P x2 P...xn , we repeat this process n − 2, a finite number,
more times. If xn+1 does not outrank any of them we have our
complete ordering x1 P x2 P...xn P xn+1 , otherwise as specified we
have our complete ordering.