HW15 Diffrentiabilty and Misc.
HW15 Diffrentiabilty and Misc.
HW15 Diffrentiabilty and Misc.
Mathematical Methods for Economists1 y ∗ (w, B) = f (x∗ ) and w · x∗ = B, i.e., the budget
constraint is binding.
1. For each constrained optimization problem, (1) define the
Lagrangian function, (2) derive necessary conditions (in- (b) Show that y ∗ (w, B) is quasi-convex in (w, B) (note
cluding FOCs) for the optimal solution, and (3) find the that we have not assumed that f is concave).
optimal solution
5. Consider the production function f (x1 , x2 ) = δ ln(x1 +
(a) max 3x + 2y s.t. x, y ≥ 0, x + y = 1 1) + (1 − δ) ln(x2 + 1), if (p, w1 , w2 ) 0 denotes the vec-
x,y
tor of commodity prices, we can define the profit func-
(b) max −(x − y)2 s.t. −1 ≤ x ≤ 2 tion π(p, w1 , w2 ) = max{py − w1 x1 − w2 x2 |y ≤ f (x1 , x2 )}
x
(c) max x4 s.t. −1 ≤ x ≤ 2 which represents the maximum profits that can be earned
x from any feasible production activity. For what values of
in
2. For each (unconstrained) optimization problem, derive (p, w1 , w2 ) is π(p, w1 , w2 ) > 0? When does π(p, w1 , w2 ) =
first order conditions, find critical points (points to solve 0? [Hint: You may use Kuhn-Tucker]
FOC), and discuss whether the point you find is a (local)
e.
6. The cost of producing qj units at plant j is Cj (qj ) = αj qj +
maximizer or (local) minimizer or neither of them
qj2
, j = 1, 2, . . . , n. The product price is p.
nc
(a) max −(x − 7)2 2βj
x∈R
(b) max x3 − x (a) Initially assume that α = (α1 , . . . , αn ) is zero and
x∈R
ra
solve for aggregate supply as a function of the output
(c) max −ex − e−x
x∈R price.
3. A is selfish with utility function (b) Suppose instead that α1 < α2 < · · · < αn . For any
nt
price p, which plants will the firm operate? Again
ua (xa , ya ) = u(xa , ya ) solve for aggregate supply as a function of the output
se
where x and y are two commodities. B cares about her price.
own consumption and A’s happiness. Her utility function
7. A monopolist sells his output in two markets (with no
ic
is
resale.) The demand price for product 1 is p1 = 10 − q1
ub (xa , ya , xb , yb ) = u(xb , yb ) + αu(xa , ya )
and for product 2 is p2 = b−q2 . The cost of producing the
om
A has an income of Ia and B has an income of Ib . total output q1 + q2 is C(q) = (q1 + q2 )2 . For what values
of b will the profit maximizing firm only sell in market 2.
(a) If u(x, y) = ln x + ln y, assuming given market prices
of px , py , under what conditions will B give part of 8. Consider the following problem: max(x1 ,x2 ) 3x1 + x2 sub-
her income to A?
on
1
(c) Consider the case in which C(x1 , x2 ) = x21 + 2. x∗2 solves maxx2 ∈B(p∗ ,p∗ ·e2 ) u2 (x2 , x∗11 ) and
x22 , D(k) = k 2 , p1 = 1, and p2 = 3. Is there a so- 3. x∗1 + x∗2 ≤ r
lution in which x1 = x2 = k > 0? Is there a solution
Use the characterization in (b) to show that no
in which 0 < x1 < x2 = k?
strictly positive competitive equilibrium allocation
10. Unbinding Constraints in Programming Problem. Con- x∗ 0 in this economy is Pareto efficient (in partic-
sider the programming problem maxx∈RL f (x) s.t. x ∈ X, ular, the first welfare theorem fails in this economy).
where X is a closed and convex set in RL . Suppose that 2 2
you added another constraint g(x) ≥ 0 to the problem, 12. Consider the following problem: max(x1 ,x2 ) −x1 −x1 x2 −x2
derived an optimal solution x∗ , and found that this addi- subject to x1 −2x2 ≤ −1 and 2x1 +x2 ≤ 2. Use the Kuhn-
∗
tional constraint is not really binding at x . Tucker conditions to find the solution(s) of the problem.
(a) Can you conclude that x∗ is also a solution for 13. Consider the following
problem:
2 2
maxx∈RL f (x) s.t. x ∈ X? If not, find a counterex- max (x1 ,x2 ) −(x 1 − 4) − (x 2 − 4) subject to
ample (a graphical example would suffice). x 1 + 2x 2 ≤ 4 and x 1 + 3x 2 ≤ 9. Use the Kuhn-Tucker
conditions to find the solution(s) of the problem.
(b) Suppose that f is strictly quasi-concave and g is con-
tinuous. Is x∗ a solution for maxx∈RL f (x) s.t. x ∈ X 14. A consumer has income I > 0, and faces a price vector
in this case? p ∈ R3++ for the three commodities she consumes. All
(c) Consider the consumer utility maximization problem commodities mush be consumed in nonnegative amounts.
with truncated consumption set (maxxi ∈Xi ui (xi ) s.t. Moreover, she must consume at least two units of com-
p · xi ≤ p · ei , xi ≤ r̄). Let x̄i (p) be a solution for this modity 2, and cannot consume more than one unit of com-
problem. Apply the above result to show that x̄i (p) is modity 1. Assuming I = 4 and p = (1, 1, 1), calculate the
also a solution for the standard utility maximization optimal consumption bundle if the utility function is given
problem if x̄i (p) r̄. by u(x 1 x2 , x3 ) = x1 x2 x3 . What if I = 6 and p = (1, 2, 3)?
,
11. Consider a pure exchange economy with two consumers, 15. A firm manufactures two outputs y1 and y2 using two in-
who are endowed with strictly positive initial endow- puts x1 and x2 . The production function f : R2+ → R2+
ments ei ∈ RL ++ , i = 1, 2. Consumer 1’s utility func- 1/2 1/2 1/3
is given by (y1 , y2 ) = f (x1 , x2 ) = (x1 , x1 x2 ). Let pi
tion u1 : RL + → R is differentiable, concave and satisfies denote the unit price of yi , and wj that of xj . Describe the
Du1 (x1 ) 0 for any x1 ∈ RL + . Assume that consumer 2 firm’s optimization problem and derive the equations that
cares about consumer 1’s consumption of good 1 as well define the critical points of the Lagrangian L. Calculate
as his or her own consumption. More specifically, con- the solution when p1 = p2 = 1 and w1 = w2 = 2.
sumer 2’s utility function u2 : RL+1 + → R is given by
u2 (x2 , x11 ) = v2 (x2 ) − x11 , where v2 is differentiable, con- 16. Construct a subset of the xy-plane R2 that is
cave and satisfies Dv2 (x2 ) 0 for any x2 ∈ RL +.
Pareto efficiency is defined in the standard way. A fea- (a) closed under vector addition and subtraction, but not
sible allocation (x1 , x2 ) ∈ R2L + (x 1 + x 2 ≤ r = e 1 + e 2 ) scalar multiplication.
is Pareto efficient if and only if there is no other fea- (b) closed under scalar multiplication but not under vec-
sible allocation (x01 , x02 ) such that u1 (x01 ) ≥ u1 (x1 ) and tor addition.
u2 (x02 , x011 ) ≥ u2 (x2 , x11 ) with at least one inequality be-
ing strict. Answer the following questions. 17. Addition (⊕) and scalar multiplication (·) are required to
satisfy these eight rules:
(a) Consider the following programming problem:
1. x ⊕ y = y ⊕ x
(P) max u1 (x1 )
(x1 ,x2 )∈R2L
+ 2. x ⊕ (y ⊕ z) = (x ⊕ y) ⊕ z
s.t. v2 (x2 ) − x11 ≥ ū2 3. There is a unique “zero vector” such that x ⊕ 0 = x
and x1 + x2 ≤ r for all x.
4. For each x there is a unique vector −x such that
Show that a feasible allocation (x∗1 , x∗2 ) ∈ R2L
+ is x ⊕ (−x) = 0.
Pareto efficient if and only it is a solution of prob-
lem (P) with ū2 = v2 (x∗2 ) − x∗11 . 5. 1 · x = x
(b) Write down the Kuhn-Tucker conditions to charac- 6. (c1 c2 ) · x = c1 · (c2 · x)
terize the set of all solutions in R2L
+ of (P). Discuss 7. c · (x ⊕ y) = c · x ⊕ c · y
why they characterize the entire set of (interior) so-
lutions. 8. (c1 + c2 ) · x = c1 · x ⊕ c2 · x
(c) Define competitive equilibrium in this pure exchange (a) Suppose addition in R2 adds an extra 1 to each com-
economy as follows: (x∗ , p∗ ) ∈ R2L L
+ × R+ is a com- ponent, so that (3, 1) ⊕ (5, 0) equals (9, 2) instead of
petitive equilibrium if (8, 1). With scalar multiplication unchanged, which
1. x∗1 solves maxx1 ∈B(p∗ ,p∗ ·e1 ) u1 (x1 ), rules are broken?
2
(b) Show that the set of all positive real numbers, with 26. Find a basis for the subspace of R3 spanned by the three
x ⊕ y and c · x redefined to equal the usual xy and vectors
xc , is a vector space. What is the “zero vector”? 1 −1 0
−1 , 0 , 1
(c) Suppose (x1 , x2 ) ⊕ (y1 , y2 ) is defined to be (x1 +
0 1 −1
y2 , x2 + y1 ). With the usual c · x = (cx1 , cx2 ), which
of the eight conditions are not satisfied? 27. Find a basis for the subspace of R4 spanned by the four
(d) The functions f (x) = x2 and g(x) = 5x are “vectors” vectors
in the vector space F of all real functions i.e. set of 1 2 1 2
all f : R → R. Addition is defined in the usual way 1 −1 1 1
, , ,
((f ⊕ g)(x) = f (x) + g(x), ∀x ∈ R). Which rule 2 −5 −4 1
is broken if multiplying f (x) by c gives the function 4 2 0 6
f (cx).
28. For which values of k does
(e) If the sum of the “vectors” f (x) and g(x) in F is
defined to be (f ⊕ g)(x) = f (g(x)) then what is the kx + y = 1
“zero vector”? Keep the usual scalar multiplication
x + ky = 1
c · f (x) = cf (x), and find two rules that are broken.
have no solution, one solution, or infinitely many solu-
18. Which of the following subsets of R3 are actually sub-
tions?
spaces?
29. Starting with a first plane u+2v−w = 6, find the equation
(a) The plane of vectors b = (b1 , b2 , b3 ) with first compo-
for
nent b1 = 0.
(b) The vectors b = (b1 , b2 , b3 ) with b2 b3 = 0. (a) the parallel plane through the origin.
(c) All linear combinations of two given vectors (1, 1, 0) (b) a second plane that also contains the points (6, 0, 0)
and (2, 0, 1). and (2, 2, 0).
(d) The plane of vectors b = (b1 , b2 , b3 ) that satisfy b3 − (c) a third plane that meets the first and second in the
b2 + 3b1 = 0. point (4, 1, 0).