Scribd
Upload
11 views2 pages

Example

The document discusses the proof of Theorem 1, emphasizing its applicability to strict preferences and addressing concerns raised by referees regarding the correctness of Proposition 1 and the theorem's broader applicability. It suggests clarifications and additional examples to illustrate preferences beyond lexicographic ones, including preferences related to hobbies. The document also includes proposed revisions to improve clarity and address specific points raised by referees.

Uploaded by

Bowen Shi
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
11 views2 pages

Example

The document discusses the proof of Theorem 1, emphasizing its applicability to strict preferences and addressing concerns raised by referees regarding the correctness of Proposition 1 and the theorem's broader applicability. It suggests clarifications and additional examples to illustrate preferences beyond lexicographic ones, including preferences related to hobbies. The document also includes proposed revisions to improve clarity and address specific points raised by referees.

Uploaded by

Bowen Shi
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 2

For Proof of Theorem 1, our proof applies for strict preferences, but not for

general preferences (about countability of some sets). We can add some words
at the beginning of Proof. Some words like:
”First note that the preference induces naturally a strict preference on X/ ∼,
the quotient space of equivalence classes in X. Let [x] = {z ∈ X : z ∼ x} be
the equivalence class of x ∈ X. Then for any x and y in X, either x ∼ y which
means [x] = [y], or x ≻ y which means [x] ≻ [y], or y ≻ x which means [y] ≻
[x]. The representability of (X, succcurlyeq) follows from the representability
of (X/ ∼, ≻). So without loss of generality we assume the preference is strict.”
Referee 1 wrote:
I have doubts regarding the correctness of proposition 1. The first sentence
of the proof states ”As for the if part, suppose not. Then there exists some
maximal element M of S that is different from X”. But this is not obvious to
me. It seems like Zorns lemma (or some other version of the axiom choice)
would be required to show this. Moreover, I think some extra condition is needed
in order to be able to show this (as you point out in your later theorem 1). This
is because the representable class is not generally closed under unions (not even
countable unions) as you show in your example 2 on lexicographic preferences
on R × Q.
There is no problem in the proof of proposition 1. The existence of some
maximal element M is the condition of ”if” part. If M = X, we are done. If
not, see our proof.

Another main concern is the applicability of theorem 1. To me it seems like


this is a theorem tailored to study representability of lexicographic preferences,
and indeed both of the applications considered are of a lexicographic nature. It
is not clear to me whether this result has any applications beyond these.
The example in Section 3.1 is not about lexicographic preferences, though
they look alike—as explained in Footnote 2. To make it more obvious, we can
add to the end of Section 3.1 or merged into Section 3.1:
Exercise 4. We can also add special sites to accommodate preferences incor-
porating special hobbies, like sports car or paragliding flight. For simplicity we
add only one site to the above example: Let A > 0 and X ′ = X ∪ (R+ × {0}) =
R+ × {0, 1, . . . , n}. We view the n + 1 place as cities. Off city 0 the preference is
the same as the previous example. When city 0 is involved, if w ≥ A > w′ ≥ 0,
then for all a > 0, any y ≥ 0, and any i ∈ {1, . . . , n},
(w + a, 0) ≻ (w, 0) ≻ (y, i) ≻ (w′ , 0) ≻ (w′ − a, 0).
Imagine that city 0 is a remote place and the agent is a sports car enthusiast.
He prefer city 0 to any other place if the income exceeds A so that he can afford
to maintain the expensive hobby. Otherwise any other place is preferred to city
0 with whatever the respective income levels.
If u(x, i) is a utility function representing the preference on X, it is straight-
forward to see that the preference on X ′ can be represented by
π π
v(x, j) = arctan{u(x, i)}1{i∕=0} +( +1−eA−x )1{i=0,x≥A} +(− −1+ex−A )1{i=0,x<A} .
2 2

1
Exercise 5.(A preference with pleasure in others’ misfortune.) Still
X = R+ × {1, 2, . . .}. There is a special positive number dj > 0 corresponding
to each j = 1, 2, . . .. For each j there is a least favorite element (dj , j). Denote
the misfortune set M = {(di , i) : i = 1, 2, . . .}. For all other elements (x, i) and
(y, j),

(x, i) ≻ (y, j) ⇐⇒ |x − di | < |x − dj |];


(x, i) ∼ (y, j) ⇐⇒ |x − di | = |x − dj |]
(x, i) ≻ (dj , j)
(dk , k) ≻ (dj , j) ⇐⇒ k > j.
1
X = ∪ı∈N Si where Si = R+ × {i}. Define ui : Si → R by ui (x, i) = |x−d i|
if
x ∕= di , and 0 if x = di . Then (di , i) ∈ Si is the only pseudo-continuous point
of the preference with respect to ui . According to Theorem 1, the preference is
representable on ∪i Si = X by some utility function. In deed, we can take it to
be as follows
1 1
u(x, i) = 1R \{d } (x) − 1{di } (x).
|x − di | + i i

In Example 2, add to its second sentence The lexicographic preference


on R × Q is not representable by any utility function.(cf. Aliprantis and
Border (2006, Example 1.4))
Aliprantis,C.D. and Border, K.C., Infinite Dimensional Analysis:
a Hitchhiker’s Guide, (3rd edition) 2006, Springer Berlin.
Referee 2:“It will be better if it is explicitly stated why both M1,k and M2,k
are countable. I can see this to hold because if any one of them were uncountable
then it would be a violation of countability of rationals.”
Exactly. The following can be taken as a footnote:
According the definition of M1,k , for any z ∈ M1,k , there exists az > 0
such that for all (z − az , z) ∩ uk (Sk ) = ∅. Take a rational rz ∈ (z − az , z).
Thus we have a one-to-one mapping z 󰀁→ rz which is strictly increasing. Since
{rz : z ∈ M1,k } is countable as a subset of Q the set of all rational numbers,
M1,k is also countable. It is similar to see the countability of M1,k .

Immediately after the proof of Theorem 1: Theorem 1 tells us that when


.... It’d be better to replace “that when” with“under what conditions”.

You might also like

pFad - Phonifier reborn

Pfad - The Proxy pFad of © 2024 Garber Painting. All rights reserved.

Note: This service is not intended for secure transactions such as banking, social media, email, or purchasing. Use at your own risk. We assume no liability whatsoever for broken pages.


Alternative Proxies:

Alternative Proxy

pFad Proxy

pFad v3 Proxy

pFad v4 Proxy