This preview shows pages 1–2. Sign up to view the full content.
This preview has intentionally blurred sections. Sign up to view the full version.
View Full Document
Unformatted text preview: Page 1 of 18 Microeconomics Prelim June 2005 Answer Keys 1. a . Proof of the property . Let x B 1 B , and let x be chosen at ( B , s ). First, if x = s , then, by (1), u ( s ) > u ( x *( B ))  c ( s ). Because B 1 B , u ( x *( B )) > u ( x *( B 1 )), and hence u ( s ) > u ( x *( B 1 ))  c ( s ) and, again by (1), s is chosen at ( B 1 , s ). Second, let x *( B ) s and let x = x *( B ) be chosen at ( B , s ). Then, by (2), u ( s ) < u ( x *( B ))  c ( s ). Because x = x *( B ) B 1 and because u ( x *( B )) > u ( x *( B 1 )), the uniqueness of the maximizers implies that x *( B 1 ) = x *( B ) = x . Thus, x is also chosen at ( B 1 , s ). Revealed preference interpretation. If x is chosen at ( B , s ), then x is revealed (weakly) preferred to any point in B when the status quo is s . The consistency of choices would then dictate that x be also revealed (weakly) preferred to any point in a subset B 1 of B , as long as x belongs to it and as long as the status quo is still s . This is precisely what the property says. b. Consider the function f : X X : f ( x , s ) = u ( x ) g ( x , s ), where g ( s , s ) = 0 and, for x s , g ( x , s ) = c ( s ) > 0. When X = N + , f is clearly discontinuous at any point in its domain of the form ( s , s ), even when both u and c are continuous functions. We now show that the maximization of f on B is equivalent to the choice rules (1) and (2). First, let x be chosen at ( B , s ). If x = s , then, by (1), f ( s , s ) u ( s ) > u ( x *( B ))  c ( s ) > u ( x )  c ( s ), for all x B , with f ( x , s ) u ( x )  c ( s ), for all x B , x s , i. e., x = s maximizes f on B . (Trivially, f ( s , s ) > f ( s , s ).) If, on the contrary, x s , then, by (2), x = x *( B ) and f ( s , s ) u ( s ) < u ( x *( B ))  c ( s ) f ( x *( B ), s ). Moreover, if x B , then u ( x *( B )) > u ( x ), by the definition of x *( B ), and hence f ( x *( B ), s ) u ( x *( B ))  c ( s ) > u ( x )  c ( s ) f ( x , s ), for all x B , x s . Summarizing, if x is chosen at ( B , s ), then x maximizes f ( x , s ) on B . Conversely, let x solve max x f ( x , s ) subject to x B . If x = s = x *( B ), then (1) is trivially satisfied. If x = s x *( B ), then, because of maximization, f ( s , s ) > f ( x , s ) for all x B , x s , i.e., u ( s ) > u ( x ) c ( s ) for all x B , x s , or u ( s ) + c ( s ) > u ( x ) for all x B , x s . This together with the fact that s x *( B ) implies that u ( s ) + c ( s ) > u ( x *( B )) (otherwise there would be an x in B , x s, with u ( s ) + c ( s ) < u ( x ), a contradiction). Hence, (1) is satisfied. Last, let x s ....
View
Full
Document
This note was uploaded on 02/09/2010 for the course ECON 200D taught by Professor Pontusrendahl during the Winter '06 term at UC Davis.
 Winter '06
 PONTUSRENDAHL
 Microeconomics

Click to edit the document details