You are on page 1of 37

HE 601: Advanced Microeconomics I

C l a s s No t e s
© S ye d M . A h s a n

Spring, 2023

Adv Micro-I, Lect 02_ Preferences_Ahsan 3/6/2023


Lecture 2: Individual Preferences
2

Outline of Lecture
 2.1 Introduction: Modelling Choice
 2.2 Preference Relations
 2.3 Utility Functions
 2.4 Choice Rules
 2.5 Preference Relations & Choice Rules
 2.6 Conclusion

Adv Micro-I, Lect 02_ Preferences_Ahsan 3/6/2023


2.1 Introduction: Modelling Choice
3

 We want to understand how individuals make a choice? Do they start


by introspection of their tastes, i.e., thinking over their preferences
over objects of choice, or, do we start by examining directly the
choices people make among the ultimate objects and study how they
had actually behaved?
 The first requires the imposition of certain rationality hypotheses on
preference relations (directly unobservable) and from there goes on to
study how choices among actual objects (observable) are made.
 The first is the more traditional, but abstract, while the second is
grounded on behavioural approach.
 The MWG text focusses on the first approach.
 Development of the second approach entails invoking some
consistency restrictions directly on the choice process (e.g., the weak
axiom of revealed preferences, WARP).
Adv Micro-I, Lect 02_ Preferences_Ahsan 3/6/2023
… Modelling Choice
4

 The latter is in parallel with the axioms of rationality required of the


preference relations.
 It has the advantage that in principle, it can accommodate more
general (yet unspecified) structure of choices than would be
permissible within the preference relation framework.
 Another aspect of note is that the second approach is firmly grounded
on behavioural economics.
 The compatibility of the eventual choices based on the two distinct
approaches is of interest in itself and will be pursued in this lecture.

Adv Micro-I, Lect 02_ Preferences_Ahsan 3/6/2023


2.2 Preference Relations
5

 The object of choice is ultimately elements of a bunch of alternatives in


the set X. Frequently, we shall denote a pair of typical elements, not
ordered pairs, as {x, y}, i.e., {x, y} X.
 We denote a preference relation by the notation, , a binary relation
on the set of alternatives, X. Thus, for example, x y is normally read
as ‘x is at least as preferred as y’ (or ‘at least as good as’).
 Preference vs Indifference: Given , we can derive two other relations
 (i) Strict Preference Relation, , is defined by
x y x y, but not y x,
i.e., ‘x is preferred to y’. Here the symbol reads ‘if and only if’.
 (ii) Indifference Relation, , is defined by
x y x y and y x,

Adv Micro-I, Lect 02_ Preferences_Ahsan 3/6/2023


… Preference Relations
6

 The latter is read as ‘x is indifferent to y’.


 Rationality: All discussion of preference relations entail the
imposition of axioms of rationality.
 Definition 1: A preference relation is said to be rational if it
possesses the following two properties.
 (i) Completeness: For all {x, y} X, we have that either
x y or y x (or both), and
 (ii) Transitivity: For all {x, y, z} X, we have
if x y and y z, then x z.
 Completeness implies that the individual is able to compare any two
bundles. The text claims that this need not be a mean task! Often it
take a lot of hard work to make up one’s mind (say, between two
homes, two job offers). Recall the joke re Buridan’s ass (c. 1300)!
Adv Micro-I, Lect 02_ Preferences_Ahsan 3/6/2023
Figure 2.1 Cartoon on Buridan’s Ass
7
Political cartoon c. 1900,
showing the United States
Congress as Buridan's ass
(in the two hay piles
version), hesitating
between a Panama route or
a Nicaragua route for
an Atlantic–Pacific canal.

Aristotle (325 BC) scolds:


‘a man, being just as
hungry as thirsty, and
placed in between food and
drink, must necessarily
remain where he is and
starve to death.’
Adv Micro-I, Lect 02_ Preferences_Ahsan 3/6/2023
… Preference Relations
8
 Varian and most other texts, also enumerate a third property of
rationality:
 (iii) Reflexivity For all {x} X, x x.
 MWG appears to imply this without having stated so.
 Transitivity especially in the strict version, is a strong assumption. and is
necessary for any discussion of preference maximization; if preferences
were not transitive, there might be sets of bundles which had no best
elements.
 In particular, it rules out cyclical pattern of preferences. MWG suggests
that without transitivity, substantial parts of economic they may cease to
hold.
 The Condorcet paradox (1785): Cyclical voting behaviour when choosing
among three alternatives, one is led to a paradox in a majority voting
situation with no majority winner. (You need a minimum of 3 voters and
three candidates). The 2-party US system generally averts this.
Adv Micro-I, Lect 02_ Preferences_Ahsan 3/6/2023
… The Condorcet paradox (1785)
9
 Given a cyclical pattern of voting by 3 individuals (1, 2, & 3) over
objects of choice {A, B, C}, we get the following pattern
 Person 1: A B and B C, then by transitivity A C
 Person 2: C A and A B, then by transitivity C B
 Person 3: B C and C A, then by transitivity B A
 Two observations follow:
 (a) Whoever may be declared a possible winner, there are 2/3 votes
again that choice. And hence there is no majority-vote winner in this
example. [Easy to verify …]
 (b) If we invoke transitivity, pattern of voting as presented above
cannot occur. [Verify: Looking at the last preference relation in each
line above, we have (perspn-1) A C, (perspn-2) C B therefore A
B; however, person-3 votes B A, contradicting the voice of 1 and 2.
Transitivity rules out the pattern of vote by person-3.
Adv Micro-I, Lect 02_ Preferences_Ahsan 3/6/2023
… Preference Relations
10

 Both completeness and transitivity of preference relations have


implications for strict preference ( ) and indifference ( ), this is
summed up below.
 Proposition 1.B1: If is rational as defined above, then
(i) is irreflexive (i.e., x x never holds) and transitive (i.e.,
if x y and y z, then x z);
(ii) is reflexive (i.e.., x x for all x), transitive (i.e., if x y and
y z, then x z), and symmetric (i.e., if x y then if y x);
(iii) if x y and y z, then x z.
 Hint of Proof: (i) if x y and y z, then it must also be true that
x y and y z, and thus x z. Since x y, let x= (y+),  >0,
i.e., y z implies (x- ) z or x (z + ). Therefore, x z.
Hence transitivity. [Note that here  will have to have the same
dimensionality of {x,y} etc.] QED
Adv Micro-I, Lect 02_ Preferences_Ahsan 3/6/2023
… Preference Relations
11

 (ii) To prove transitivity of : Direct argument would proceed as


follows. Since x y and y z, we can simply replace y by x in the last
relation to obtain x z. Hence transitivity.
 The symmetry part follows immediately from reflexivity: If x y, we
have both x y and y x. In the second relation y x cannot hold
since by replacing y by x, we would get x x, violating reflexivity. QED
 (iii): Here again the result is immediate once you use the argument
that since x y, let x= (y+),  >0. Then we may re-write y z as
(x - ) z, or x (z + ), i.e., x z. Hence transitivity.
 The takeaway from the above proposition is that both strict preference
and indifference are also transitive, while the last element states
that we can have a transitivity type property by combining a strict
preference with at least a weak preference.

Adv Micro-I, Lect 02_ Preferences_Ahsan 3/6/2023


… Preference Relations
12
 While the MWG Text defines the two properties of preferences, namely
(i) completeness and (ii) transitivity as ‘rationality’, at least in chapter
1, and then they move on to the idea of choice rules and utility.
 We have already cited a third, (iii) ‘reflexivity’.
 Many other texts typically discuss other properties that make
preference relations more useful to economists.
 We turn to these, but first we need to refresh ourselves of some
background mathematical ideas. See Box 2.1 (slides # 16-25).
 (iv) CONTINUITY. For all y X, the sets {x : x y) and {x : y x) are
closed sets. It follows that {x : x y) and {x : y x) are open sets.
 See Figure 2.2 next slide.
 In other words, if (x') is a sequence of consumption bundles that are all
at least as good as a bundle y, and if this sequence converges to some
bundle x*, then x* is at least as good as y.
Adv Micro-I, Lect 02_ Preferences_Ahsan 3/6/2023
… Continuity (Fig 2.2)
13
 An important consequence of continuity is that if y is strictly preferred to z
and if x is a bundle that is close enough to y, then x must be strictly
preferred to z.
 Alternatively, this is restating the assumption that the set of strictly
preferred bundles is an open set.
 A common example of a set strictly x2
preferred to another is of curse the
indifference curve, though we are yet B: U(x) > 10
to formally define it.
 Define the set of points labelled A
such that U(x) =10, then all points
in X, where U(x) >10 must be an
A: U(x) = 10
open set.
 A: x R2, st U(x) = 10 0 x1
 B: x R2, st U(x) > 10. Fig 2.2
Adv Micro-I, Lect 02_ Preferences_Ahsan 3/6/2023
… Additional Properties
14
 Other properties include the following:
 (v) Weak Monotonicity: If x  y then x y.
 (vi) Strong Monotonicity : If x  y and x  y, then x y.
 Weak monotonicity implies costless disposal, which is not always easy
(e.g., waste), while its strong version is interpreted as ‘goods’ are goods,
not ‘bads’.
 Another assumption that is weaker than either kind of monotonicity is
non-satiation:
 (vii) Local Non-satiation: Given any x in X and any  > 0, then there is
some bundle y in X with |x - y| <  such that y x.
 The above suggests that consumers find even small changes in their
consumption bundle agreeable. It so happens that strong monotonicity
implies local non-satiation but not vice versa.
 Further, local non-satiation rules out "thick" indifference curves.
Adv Micro-I, Lect 02_ Preferences_Ahsan 3/6/2023
… Additional Properties
15
 Two other properties are invoked to guarantee nice behavior of
consumer demand functions.
 (viii) Convexity: Given x, y, and z in X such that x y and y z, then it
follows that {x + (1 - )y} z for all 0    1.
 (ix) Strict Convexity: Given x  y and z in X, if x y and y z, then
{x + (1 - )y} z for all 0 <  < 1.
 Convexity generally implies that individual prefer a balanced bundle to
extremes.

Adv Micro-I, Lect 02_ Preferences_Ahsan 3/6/2023


Box 2.1: Some Mathematical Definitions
16

 (1) An Open Ball: Given a vector x in Rn and a positive real number ,


we define an open ball of radius  at x as B(x) = {y Rn : | y - x | < ).
 A closed ball would allow | y - x |  .
 (2) The epsilon ()-Neighbourhood N(x,), on a set X: This is an open
set that contains some open ball B(x) as defined above. However,
N(x,) does not have to be a ball.
 (3) Open Set: This is a set of points A X is an Open Set if for all x A,
there is an open ball around x (relative to X), all of whose elements (in
X) are also elements of A. (All x A are thus interior points.)
 (4) Interior point: We also say that x lies in the interior of A.
 (5) Complement of a Set (A): This set in Rn consists of all the points in
Rn that are not in A; it is denoted by Rn\A. Thus, given any two sets A
& B X, B would denote the complement of A (B: X\A), if B includes
all points in X that do not belong to A.
Adv Micro-I, Lect 02_ Preferences_Ahsan 3/6/2023
… Box 2.1/Fig 2.2: Open and Closed Sets
17

 (6) Closed Sets: A set is called ‘closed’ if Rn\A is an open set. In effect, all
boundary points are included in the set.
 In Fig 2.3 (from MWG, p.944), the left panel depicts an open set A, whose
boundary is excluded from A, where the corresponding closed set, A, is
also shown in the right panel.

Adv Micro-I, Lect 02_ Preferences_Ahsan 3/6/2023


… Box 2.1: Math Definitions
18

 (The set A is closed because A\B is an open set and for x A\B, the
B(x) ball lies entirely in A\B.
 (7) Bounded Set: (Varian) The set A Rn is ‘bounded’ if for any  R,
all |x|<  lies in the set A. (Note strict inequality)
 Equivalently, all distances in R2, would be finite.
 An alternative definition (similar to MWG) goes as follows:
 Let A Rn. The set A is said to be ‘bounded if there exists an x Rn and a
positive real number  >0 such that A B(x). A is said to be
‘unbounded’ otherwise. I.e., here the boundedness is defined with the
idea of an open ball. Note that A need not be a proper subset of B.
 See illustration next slide (Fig 2.4)
 (8) Compact Sets: If a nonempty set in Rn is both closed and bounded,
it is called ‘compact’. Fig 2.5 offers an illustration on the real line.

Adv Micro-I, Lect 02_ Preferences_Ahsan 3/6/2023


Fig 2.4 … Box 2.1 .. Boundedness
19

 Think of a function: (x2 + y2) 25.


 We know that (x2 + y2) = 25 is drawn as the circle of radius |5|.
 Set A: {x2 + y2) < 25} would be an example y
of an open set in R2, i.e., not including the
(o,5)
actual boundary of the circle.
 But set B: {x2 + y2)  25} would be closed
set. Both A & B are bounded, though.
x
 The complement of B, R \B, call it
2
(-5,0) (0,0) (5,0)
set C would another open set. The
latter set of course consists of the points
shown by arrows away from the circle:
C: {x2 + y2) > 25}. (0,-5)
 This clearly is an unbounded set
(of an indeterminate radius).
Adv Micro-I, Lect 02_ Preferences_Ahsan 3/6/2023
Fig 2.5 ….Box 2.1 … Compact Sets
20

 Illustration: the interval A = (−∞, −2] is not compact because it is not


bounded. The interval C = (2, 4) is not compact because it is not closed
(but bounded). The interval B = [0, 1] is compact because it is both
closed and bounded (Heine-Borel theorem).

 Another example appears below (in R2-space).

Adv Micro-I, Lect 02_ Preferences_Ahsan 3/6/2023


Fig 2.6 ….Box 2.1 … Compact Sets
21

 Here the mustard-yellow shaded set is compact; the boundedness is


illustrated by the fact that every point in K, less than  in absolute
value, fully resides in some open ball (of radius  > 0) around x, say x1,
x2, .., x5.

Adv Micro-I, Lect 02_ Preferences_Ahsan 3/6/2023


… Box 2.1: Math Definitions
22
 (9) Limits, Sequences and their Convergence: An infinite sequence in
Rn, ( xi) = (x1, x2,. . .) is just an infinite set of points, one point for each
positive integer. A sequence (xi) is said to ‘converge’ to a point x* if for
every  > 0, there is an integer m such that, for all i > m, xi is in B(x*).
 Informally one often says that xi gets arbitrarily close to x*.
 Def: x* is the limit of the sequence (xi ) and write lim i   xi = x*.
 Discuss the familiar rectangular hyperbola, e.g., y = (k/x), k>0, and
its limits (see Fig 2.7).
 Def: If a sequence converges to a point (or to its limit), we call it a
‘convergent sequence’.
 (10) Closed set (an alternate definition): The set A is closed if every
convergent sequence in A converges to a point in A.
 (11) Compact set (an alternate definition): The set A is compact, then
every sequence in A has a convergent subsequence.
Adv Micro-I, Lect 02_ Preferences_Ahsan 3/6/2023
Fig 2.7 … Box 2.1: A Rectangular Hyperbola
23

 The Illustration of a rectangular y


hyperbola: y = (k/x), k>0.
 Noe that here
lim x   y = 0, while
lim x  0 y = .
 The name of the function is derived
from the fact that at any point on it, all
rectangles below are of equal magnitude
simply because xy =k, a constant.
 This function has many well-known 0 x
interpretations in microeconomics, e.g., if it were a demand curve, p(x), the
marginal revenue (MR) would be exactly zero; (this is immediate by
differentiating p(x)x =k).
 Secondly, the price elasticity of demand (-)[(p/x)(dx/dp)] = 1 (verify).
Adv Micro-I, Lect 02_ Preferences_Ahsan 3/6/2023
… Box 2.1/Fig 2.8: Math Definitions
24

 (12) Continuity & Continuous Functions: A function f(x) is continuous


at x* if for every sequence (xi) that converges to x*, we have the
sequence (f(xi)) converging to f(x*).
 Def: A function that is ‘continuous’ at every point in its domain is
called a ‘continuous function’.
 Intuitively, a function ceases to be continuous if it displays a jump in
its value at some point x. See Fig 2.9 next slide.

Adv Micro-I, Lect 02_ Preferences_Ahsan 3/6/2023


Fig 2.9 ….Box 2.1 … Continuous Functions
25

 Clearly the function on the left, panel (a), is continuous, while the one
on the right, panel (b), is not. The jump in the latter is shown by a
dashed vertical line at some point on the unit interval.

Adv Micro-I, Lect 02_ Preferences_Ahsan 3/6/2023


2.3 The Utility Function
26

 Preferences, as you have already seen in prior microeconomics


courses, are typically represented by utility functions, u(x), which is
obtained by relating a numerical value to each x in X. Specifically we
define u(x) as follows.
 Definition: A function u: X R, u(x) is called a utility function
depicting preference relations such that for all {x, y} X whenever
x y u(x)  u(y).
 Ordinality of u(x), i.e.., non-uniqueness of u: Note that u is not unique
for ; any increasing function f: RR, v(x) = f(u(x)), i.e., for f (.) > 0,
also represent the same preferences.
 Examples: f = bu, or f = a +bu, b>0, or f = au(x)b, a>0, b>1.
 Cardinal utility functions: In risk environments, expected utility
functions only admit affine transformations, i.e., of the f(.) = a + bu(.)
variety.
Adv Micro-I, Lect 02_ Preferences_Ahsan 3/6/2023
… Utility Function
27
 Varian: “A function is an affine function if it can be expressed in the
form f(x) = a + bx. Affine functions are sometimes called linear
functions, but strictly speaking, this is only correct when a = 0.”
 Proposition 1.B2: A preference relation can only be represented by a
utility function only if it is rational.
 Proof: The arguments proceeds by taking a utility function u(x)
representing , and then showing that such preferences are complete
and transitive.
 Completeness: Since u(x) is a real-valued function on X, for all {x,y}
X, either u(x)  u(y) or u(y)  u(x). Now, the latter is only possible if,
correspondingly, we have x y or y x, ergo, preferences were
complete.
 Give some examples with numbers.

Adv Micro-I, Lect 02_ Preferences_Ahsan 3/6/2023


… Utility Function
28
 Transitivity: Suppose x y and y z. Since u(.) represents , we must
have u(x)  u(y) and u(y)  u(z). But the latter are real numbers, we must
have u(x)  u(z) and, again since u(.) represents , hence x z. QED
 Now if we ask the converse question: Can any rational preference relation
be represented by some utility function u(x)?
 The general answer is no! Though when X is finite, this is always possible.
 Indifference Curves: The set of all consumption bundles that are
indifferent to each other is called an indifference curve. One can think of
indifference curves as being level sets of the utility function.
 The set of all bundles on or above an indifference curve, {x in X : x y), is
called an upper contour set; (see Fig 2.2, slide 13, for further discussion &
an illustration).
 Convexity implies that individuals prefer averages to extreme bundles; as
shall see later, it can also be seen as a generalization of the Hicksian
concept of diminishing marginal rate of substitution.
Adv Micro-I, Lect 02_ Preferences_Ahsan 3/6/2023
2.4 Choice Rules
29

 So far, we have taken preferences as the primitive concept (where it all


begins …) and examined some properties of the utility function that
describe such preferences.
 Now we conceive of actual choices as the primitive concept, and if such
choices follow some rational preference relations.
 First, some definitions.
 Choice Structures (ℬ ; C(.)): Two elements make up the choice
structure:
 (a) Budget Sets (B): The set ℬ consists of a family of non-empty
subsets of X, such that every element B X is a set in X. Thus B ℬ is
interpreted as a budget set. The latter describes an exhaustible account
of feasible choice experiments in a given context.
 (b) Choice Rule (C): A choice rule is one that assigns a chosen set C(B)
B, for every B ℬ.
Adv Micro-I, Lect 02_ Preferences_Ahsan 3/6/2023
… Choice Rules
30

 If C(B) consists of a single element, then that is what would be chosen


among all elements in B. If multiple points are included in C(B), then
these would represent the set ‘acceptable allocations’ in B.
 Examples (MWG): (i) Let X =(x, y, z) and ℬ ={B1, B2} = {{x,y}, {x,y,z}}.
Then (ℬ ; C1(.)) may be a choice structure, such that C1{x,y} = {x} and
also C2{x,y,z} = {x}. Here {x} is the choice regardless of the budget.
 Alternatively, it may be that C1{x,y} = {x}, but C2{x,y,z} = {x,y},
meaning that whenever the budget is {x,y,z}, the individual may
choose either x or y.
 Thus, C1 and C2 are two distinct choice rules. When we are in the realm
of choice structures, it may be natural to impose some reasonableness
criteria on behaviour. What can that be?
 For example, if we see one choosing x when faced with C1{x,y}, we may
be surprised to see her choose y when facing C2{x,y,z}. Why?
Adv Micro-I, Lect 02_ Preferences_Ahsan 3/6/2023
… Choice Rules
31

 WARP (weak axiom of revealed preference, due to Paul Samuelson): A


choice structure (ℬ ; C(.)) satisfies WARP if the following holds: If for
some B ℬ, and for {x,y} B, x C(B), then for any B ℬ with {x,y}
B and y C(B), we must also have x C(B). Definition in MWG.
 In other words, the weak axiom says that if for some budget, when
both {x,y} are available and x is chosen, then there can be no other
budget allowing both {x,y} where y is chosen, but x is not.
 Def: Revealed preference relation ( *, ‘revealed at least as preferred’).
With this notation, one can express WARP relatively more simply:
Given a choice structure, (ℬ ; C(.)),
x * y  for some B ℬ, {x, y} B and x C(B).
 However, the relation * need not be either complete or transitive,
though for x and y to be comparable, either x C(B), y C(B), or both.

Adv Micro-I, Lect 02_ Preferences_Ahsan 3/6/2023


… Choice Rules
32

 Now we can state WARP as: ‘if x is revealed as good as y, y cannot be


revealed preferred to x’.
 Remarks: (i) Note that choice structure C1{x,y} = {x}, but C2{x,y,z} =
{x,y}, violates WARP. The second part says y * x, but the first part
says x *y, and hence a contradiction.
 (ii) In the next lecture, we discuss additional restrictions on revelated
preference relation as relevant in the demand context.

Adv Micro-I, Lect 02_ Preferences_Ahsan 3/6/2023


2.5 Preference Relations & Choice Rules
33

 Q1: Do choices based on preference relations when faced with


budget sets ℬ necessarily generate a choice structure that satisfies the
WARP?
 Q2: Given a consumer’s choice structure, (ℬ ; C(.)), satisfying WARP,
are there rational preference relations consistent with the former
choice?
 Answer to Q1: YES. The argument to establish this proceeds in two
steps: (a) First, define the preference maximizing choice, C*(B, ), for
a non-empty budget B X, which is to choose elements of the set
C*(B, ) = {x B , x y for all y B}. [Assume C*(.) to be non-empty]
 We then say that rational preference relations generate choice
structure defined by {ℬ; C*(., )}.
 Step 2 appears below.

Adv Micro-I, Lect 02_ Preferences_Ahsan 3/6/2023


… Preferences & Choice
34
 (b) Proposition 1.D.1 (MWG): Let denote some rational preference
relations, then the choice structure { , ℬ; C*(., )} satisfy WARP.
 Proof: For some B ℬ, we have {x,y} B and x C*(B, ) such that
x y. To see if the former satisfy WARP, let B ℬ with {x,y} B, and
y C*(B, ). The latter implies y z, for all z B, but we know x y,
and by transitivity, x z for all z B. Hence x C*(B, ) too. That is
what we needed to argue. If x is chosen in B, and x is still available in
B, it cannot be excluded while choosing among the altered budget set.
 Answer to Q2: … Maybe. The argument is far too technical for us at
this time. Suffice it to say that the weak axiom may not consistently
describe all rational preference relations, .
 In the context of demand theory, we shall see that further restrictions
on the weak axiom may well provide necessary and sufficient
conditions for choice behaviour to be rationalized by preference
relations.
Adv Micro-I, Lect 02_ Preferences_Ahsan 3/6/2023
2.6 Conclusion
35

 With this lecture, we have started the discussion of consumer choice


by first describing the concept of rational preference relations ( ).
 The Varian text defines additional properties of preference relations
beyond the requirements of ‘rationality’ as are relevant to describing
consumer choice more completely.
 In this context, it was necessary to review some concepts of analytical
mathematics (e.g., boundedness and compactness of sets etc), all of
which were utilised in the discussion of the preference relations.
 The idea of a utility function, namely real-valued functions
representing preference relations, was introduced next.
 We then examined actual choice environments given a budget set, and
wondered what would be a meaningful restriction on choices to make
these internally consistent, which led to the idea of the weak axiom of
revealed preference (due to Samuelson), WARP.
Adv Micro-I, Lect 02_ Preferences_Ahsan 3/6/2023
… Conclusion
36

 We then explored whether rational preferences may always be


described by choice structures consistent with WARP and vice versa.
 We saw the outline of the argument involved in establishing the first
correspondence.
 The reverse possibility, that every choice structure satisfying WARP
would be describable by rational preference relations may not always
hold. Due to the technicalities involved, we skipped the details.
 However later in the discussion of the demand theory, we shall return
to the topic and explore further restrictions on the weak axiom that
make it possible for the mutuality to exist.

Adv Micro-I, Lect 02_ Preferences_Ahsan 3/6/2023


References
37

 Mas-Colell, Whinston & Green (MWG, 1995), Ch. 1


 Varian (1992), Ch 7

Adv Micro-I, Lect 02_ Preferences_Ahsan 3/6/2023

You might also like