Microeconomic Theory Solved Problems1

Amit Kumar Goyal2

Solved problems
1. Derive the Walrasian demands when the utility function is
β
(a) u(x1 , x2 ) = xα
1 x2
Clearly, if either x1 or x2 is 0, utility is 0. This cannot be optimal since positive utility is possible. Hence
we know that both must be strictly positive at the optimum so we can use a Lagrangian. The Lagrangian
is
β
L = xα
1 x2 + λ[w − p1 x1 − p2 x2 ]

The first order conditions, then, are

αx1α−1 xβ2 − λp1 = 0
and
β−1
βxα
1 x2 − λp2 = 0
and, of course, the budget constraint.
p1 x1 + p2 x2 = w
We can solve the first equation for λ and substitute into the second to get
αx1α−1 xβ2 βxα β−1
1 x2
=
p1 p2
or
α
p1 x1 = p 2 x2
β
Substituting this into the budget constraint yields
α
p2 x2 [1 + ]=w
β
or
βw
x2 =
p2 [α + β]
and
αw
x1 =
p1 [α + β]
So,  
αw βw
x(p, w) = (x1 , x2 )(p1 , p2 , w) = ,
p1 [α + β] p2 [α + β]
√
(b) u(x1 , x2 ) = x1 + 2 x2
Let’s make good 1 the numeraire so that its price is 1 and let p2 be the price of 2. The first order condition
for an interior max is
1 1
p2 = √ i.e. x2 = 2 = p−2 2
x2 p2
What about corner solutions? There will never be a corner solution where x2 = 0, since the marginal
utility of x2 approaches infinity as x2 approaches 0. But there will be a corner solution with x1 = 0 if
p2 (p−2 −1
2 ) = p2 > w or equivalently if p2 < 1/w. Hence, the demand is given by

w 1

(0, )
 if w <
x(p, w) = (x1 , x2 )(p, w) = p2 p2
1 1 1
(w − , 2 )
 if w ≥
p2 p2 p2
1 Please refer a good textbook in Microeconomics for the detailed treatment of the concepts. Few of those are listed here: Intermediate

Microeconomics by Varian, Microeconomic Analysis by Varian, A Course in Microeconomic Theory by Kreps, Advanced Microeconomic
Theory by Jehle and Reny, Microeconomic Theory by Mas-collel, Whinston and Green.
2 Contact me: [email protected]

1
 (c) u(x1 , x2 ) = min{x1 + 2x2 , 2x1 + x2 }
Lets plot an indifference curve for say u = α > 0 and examine all the equilibrium possibilities,
IC(α) = {(x1 , x2 ) ∈ R2+ | min{x1 + 2x2 , 2x1 + x2 } = α}

↑ ↑ ↑
x2 x2 x2
(0, α)
6 6 6
A A A
A A A
A A w
A
A α α (0, ) H A w w
A• ( , ) w A p
) P A•H 2
•H
( , )
(0,
H AH
HH 3 3 p2 PPH Hp1 + p2 p1 + p2
HH PP
HH HH
H- PPH- H-
(α, 0) x → w w
(0, 0)
1
(0, 0) ( , 0) x1 → (0, 0) ( , 0) x1 →
p1 p1
p1 1 p1 1
IC(α) < =
p2 2 p2 2
↑ ↑ ↑
x2 x2 x2
6 w 6 w 6
(0, ) (0, )
A p2 A p2 BA
w A A BA
(0, )
p2 @ A A BA
@A w w A w w BA
A• ( , ) AA• ( , )
B A•HH
@HH p1 + p2 p1 + p2 HH p1 + p2 p1 + p2
@H A HH B H
@ HH - A H- B HH -
w w w
(0, 0) ( , 0) x1 → (0, 0) ( , 0) x1 → (0, 0) ( , 0) x1 →
p1 p1 p1
1 p1 p1 p1
< <2 =2 >2
2 p2 p2 p2

We can summarize the above information to get the demand in the following way
 w p1 1
 {( , 0)} if <
p1 p2 2



w w p1 1


2
{(y 1 , y2 ) ∈ R+ |p1 y1 + p2 y2 = w, ≤ y1 ≤ } if =





 p1 + p2 p1 p2 2
x(p, w) = (x1 , x2 )(p, w) ∈ {( w , w )} if
1
<
p1
<2
 p1 + p2 p1 + p2
 2 p2
 w p1
{(y1 , y2 ) ∈ R2+ |p1 y1 + p2 y2 = w, 0 ≤ y1 ≤ } if =2



 p1 + p2 p2


{(0, )} w p1
if >2

p2 p2

2. Laxmi is a poor agricultural worker. Her consumption basket comprises three commodities: rice and two
vegetables - cabbage and potato. But there are occasionally very hard days when her income is so low that
she can afford to buy only rice and no vegetables. However, there never arises a situation when she buys only
vegetables and no rice. But when she can afford to buy vegetables, she buys only one vegetable, namely the
one that has the lower price per kilogram on that day. Price of each vegetable fluctuates day to day while
the price of rice is constant. Write down a suitable utility function that would represent Laxmi’s preference
pattern. Explain your answer.
Check that the following utility function represents the preferences of Laxmi:
√
u(xc , xp , xr ) = max{xc , xp } + xr

3. A monopolist has contracted to sell as much of his output as he likes to the government at Rs. 100/- per unit.
His sale to the government is positive. He also sells to private buyers at Rs. 150/- per unit. What is the price
elasticity of demand for the monopolist’s products in the private market?
Since monopolist’s sale to the government is positive, his marginal revenue at the point of sale in the private
market must be Rs. 100/-. Now price in the private market is Rs. 150/-. We can compute the price elasticity
of demand in the following way:

TR(x) = p(x) · x

2
 Differentiating TR(x) w.r.t. x, we get
dp(x)
MR(x) = p(x) + x
dx
x dp(x)
⇔ MR(x) = p(x) + p(x)
p(x) dx
 
1
⇔ MR(x) = p(x) 1 +
ξ
Now substituting p(x) = 150 and MR(x) = 100 in above we get elasticity, ξ = −3
4. (ME-2.3(2008)) Two firms 1 and 2 sell a single, homogeneous, infinitely divisible good in a market. Firm 1 has
40 units to sell and firm 2 has 80 units to sell. Neither firm can produce any more units. There is a demand
curve: p = a − q, where q is the total amount placed by the firms in the market. So if qi is the amount placed
by firm i ∈ {1, 2}, q = q1 + q2 and p is the price that emerges. a is positive and a measure of market size. It
is known that a is either 100 or 200. The value of a is observed by both firms. After they observe the value
of a, each firm decides whether or not to destroy a part of its output. This decision is made simultaneously
and independently by the firms. Each firm faces a constant per unit cost of destruction equal to 10. Whatever
number of units is left over after destruction is sold by the firm in the market. Solve for Nash equilibrium for
both values of a.
Firm 1’s objective:

max (a − q1 − q2 )q1 − 10(40 − q1 )

q1
subject to 0 ≤ q1 ≤ 40.

Differentiating the objective we get,

(a − 2q1 − q2 ) + 10
Given q2 , if (a − 2q1 − q2 ) + 10 ≥ 0 at q1 = 40, then the best response of firm 1 is 40.
If (a − 2q1 − q2 ) + 10 ≤ 0 at q1 = 0 then the best response of firm 1 is 0.
If (a − 2q1 − q2 ) + 10 = 0 at some 0 ≤ q1 ≤ 40 then the best response of firm 1 is q1 = (a + 10 − q2 )/2. To
summarize, best response correspondence of firm 1 is:

q1 (q2 ) = 40 if (a − 70 ≥ q2 )
=0 if (a + 10 ≤ q2 )
= (a + 10 − q2 )/2 if (a − 70 ≤ q2 ≤ a + 10)

Firm 2’s objective:

max (a − q1 − q2 )q2 − 10(80 − q2 )

q2
subject to 0 ≤ q2 ≤ 80.

Similar to above, best response correspondence of firm 2 is:

q2 (q1 ) = 80 if (a − 150 ≥ q1 )
=0 if (a + 10 ≤ q1 )
= (a + 10 − q1 )/2 if (a − 150 ≤ q1 ≤ a + 10)

Put a = 200, the Nash equilibrium is: (q1 , q2 ) = (40, 80)

Put a = 100, the Nash equilibrium is: (q1 , q2 ) = (110/3, 110/3)
5. (ME-2.3b(2007)) Consider an IS-LM model for a closed economy with government where investment (I) is a
function of rate of interest (r) only. An increase in government expenditure is found to crowd out 50 units of
private investment. The government wants to prevent this by a minimum change in the supply of real money
balance. It is given that dI/dr = −50 , dr/dy (slope of the LM curve) = 1/250, and slope of the IS curve,
dr/dy = −1/125 and all relations are linear. Compute the change in y from the initial to the final equilibrium
when all adjustments have been made.

I = A − 50r
M
= ky − hr = k(y − 250r) . . . because dr/dy = k/h = 1/250
P
(1 − c)y = A − 50r + G . . . Since dr/dy = −(1 − c)/50 = −1/125, this gives us c = 0.6
Thus y = 2.5A − 125r + 2.5G

3
Since an increase in government expenditure is found to crowd out 50 units of private investment. This implies
r has gone up by 1 as a result of increase in G. Solving IS-LM for r, we get:
1
∆r = ∆G
150
Putting ∆r = 1, we get ∆G = 150. Solving IS-LM for y, we get:
5
∆y = ∆G = 250
3
The above solution is not adjusted for the money supply change to rectify the crowding out: So we need to
increase the money supply such that rate of interest again fall back to its original value i.e. we need to do the
following: we will move along the new IS curve in such a way that r falls by 1 from its new value. Given the
slope of the IS curve
dr/dy = −1/125
we get,
∆y = 125
So the total change in income due to fiscal policy and corresponding monetary policy is 250+125 = 375.

4
Mathematics and Statistics Solved Problems
1. (Ref: DSE 2010(21)) Let N be the number of observations that lie in [a,b] when a sample (x1 , x2 ) is drawn.
Expected value of N is

E(N ) = 0 · Pr(x1 ∈
/ [a, b], x2 ∈
/ [a, b]) + 1 · Pr(x1 ∈ [a, b], x2 ∈
/ [a, b]) +
1 · Pr(x1 ∈
/ [a, b], x2 ∈ [a, b]) + 2 · Pr(x1 ∈ [a, b], x2 ∈ [a, b])
= 0 + 1 · Pr(x1 ∈ [a, b]) · Pr(x2 ∈
/ [a, b]) + 1 · Pr(x1 ∈
/ [a, b]) · Pr(x2 ∈ [a, b]) +
2 · Pr(x1 ∈ [a, b]) · Pr(x2 ∈ [a, b]) . . . since the draws are independent
Z b Z b ! Z b ! Z
b Z b Z b
= 1· f (x)dx · 1 − f (x)dx + 1 · 1 − f (x)dx · f (x)dx + 2 · f (x)dx · f (x)dx
a a a a a a
!
Z b Z b Z b Z b
= 2· f (x)dx · 1− f (x)dx +2· f (x)dx · f (x)dx
a a a a
Z b
= 2· f (x)dx
a

2. (Ref: DSE 2010(22)) Lets first find the distribution of Y = max{X1 , X2 , . . . , Xn }. For 0 < y < 1,

Pr(Y ≤ y) = Pr(max{X1 , X2 , . . . , Xn } ≤ y)
= Pr(X1 ≤ y, X2 ≤ y, . . . , Xn ≤ y)
= Pr(X1 ≤ y) · Pr(X2 ≤ y) · · · Pr(Xn ≤ y) . . . by independence of Xi ’s
= yn

Probability density of Y , denoted by f (·), is the derivative of y n with respect to y,

f (y) = n · y n−1

Expectation of Y is
Z 1
E(Y ) = yf (y)dy
0
Z 1
= y(n · y n−1 )dy
0
Z 1
= ny n dy
0
 1
n
= y n+1
n+1 0
n
=
n+1

3. (Ref: DSE 2010(23)) Given that X takes values in {−1, 0, 1} and Pr(X = −1) = Pr(X = 0) = Pr(X = 1) = 1/3.
Define Y = X 2 . Pr(Y = 1) = Pr(X = 1) + Pr(X = −1) = 2/3 and Pr(Y = 0) = Pr(X = 0) = 1/3. Clearly
E(X) = 0 and E(Y ) = 2/3. To determine the correlation between Y and X, we first compute the covariance
between X and Y .
1 1 1
Cov(X, Y ) = E(XY ) − E(X)E(Y ) = E(X 3 ) = (−1)3 ( ) + (0)3 ( ) + (1)3 ( ) = 0
3 3 3
Since the covariance is 0, X and Y are uncorrelated. To check for independence consider the event (X = 0, Y =
0).
1 1 1 1
Pr(X = 0, Y = 0) = Pr(X = 0) = 6= = · = Pr(X = 0) Pr(Y = 0)
3 9 3 3
Hence X and Y are not independent.
4. (Ref: DSE 2010(26)) For x ∈ [−1, 1],

Pr(X ≤ x) = Pr(sin θ ≤ x)
= Pr(θ ≤ sin−1 x) . . . because sine is an increasing function over [−π/2, π/2]
1  −1 π
= sin x +
π 2

5
 5. (Ref: DSE 2008(58)) It can be easily shown (graphically, or otherwise) that x > 1 + log x, ∀x 6= 1. Thus, for
x = log π we have
log π > 1 + log log π
⇔ elog π > e1+log log π
⇔ π > e1 elog log π
⇔ π log e > e log π
⇔ log eπ > log π e
⇔ eπ > π e

6. (Ref: DSE 2009(II(17))) Jai and Vijay are taking a exam in statistics. The exam has only three grades A, B
and C. The probability that Jai gets a B is 0.3, the probability that Vijay gets a B is 0.4, the probability that
neither gets an A, but at least one gets a B is 0.1. What is the probability that neither gets a C but at least
one gets a B?
In what follows, I am going to write Jai’s grade first, followed by Vijay’s grade. For example, if I mention the
sample point AB, then that would mean that Jai’s grade is A and Vijay’s grade is B.
Pr(Jai gets a B) = Pr(BA or BB or BC) = Pr(BA) + Pr(BB) + Pr(BC) = 0.3 —(1)
Pr(Vijay gets a B) = Pr(AB or BB or CB) = Pr(AB) + Pr(BB) + Pr(CB) = 0.4 —(2)
Pr(Neither gets an A, but at least one gets a B) = Pr(BC or BB or CB) = Pr(BC) + Pr(BB) + Pr(CB) = 0.1
—(3)
Pr(Neither gets an C, but at least one gets a B) = Pr(BA or BB or AB) = Pr(BA) + Pr(BB) + Pr(AB) = 0.6
(Add (1) and (2) and then subtract (3) from it)
7. (Ref: DSE 2010(II(21))) Let N denotes the number of observations in the sample that fall within the specified
interval [a, b]. Since we are drawing a sample of size 2, we have:
 Rb 2
0, with probability (1R− a f (x)dx) R

b b
N = 1, with probability 2( a f (x)dx)(1 − a f (x)dx)
 Rb
2, with probability ( a f (x)dx)2


Z b Z b Z b Z b
2
E(N ) = (0)(1 − f (x)dx) + (1)2( f (x)dx)(1 − f (x)dx) + (2)( f (x)dx)2
a a a a
Z b
= 2( f (x)dx)
a

8. (ME-1.18(2010)) Let X be a Normally distributed random variable with mean 0 and variance 1. Let Φ(·) be
the cumulative distribution function of the variable X. Then the expectation of Φ(X) is
To find the Expectation of Φ(X), we will first recover its distribution: For 0 < a < 1,
Pr(Φ(X) ≤ a) = Pr(X ≤ Φ−1 (a)) . . . because Φ is strictly increasing
= Φ(Φ−1 (a))
= a
Thus Φ(X) is distributed uniformly over (0, 1). Hence the expectation of Φ(X) is 1/2.
9. (ME-1.6(2006)) Let f (x) be a function of real variable and let ∆f be the function ∆f (x) = f (x + 1) − f (x) .
For k > 1, put ∆k f = ∆(∆k−1 f ) . Then ∆k f (x) equals
For k = 1,
∆1 f (x) = ∆f (x) = f (x + 1) − f (x) (This will rule out options (a) and (d))
For k = 2,
∆2 f (x) = ∆(∆f (x)) = ∆f (x + 1) − ∆f (x) = f (x + 2) − f (x + 1) − f (x + 1) + f (x) = f (x + 2) − 2f (x + 1) + f (x)
(This will help us rule out option (b)).
So, the answer must be (c).
10. (ME-1.14(2006)) The number of positive pairs of integral values of (x, y) that solves 2xy −4x2 +12x−5y = 11 is

2xy − 4x2 + 12x − 5y = 11

(2x − 5)y = 4x2 − 12x + 11
4x2 − 12x + 11
y =
2x − 5
6
y = 2x − 1 +
2x − 5

6
 Now 6 is divisible by 2x − 5 for the following values of x: 1, 2, 3, 4. And y is positive when x is either 3 or 4.
Hence the answer is 2.
Z √x21 +x22
2 2 2
11. Consider the function f (x1 , x2 ) = e−(w /(x1 +x2 )) dw with the property that f (0, 0) = 0. Show that
0
the function is homogeneous of degree 1.
Z √(λx1 )2 +(λx2 )2
2
/((λx1 )2 +(λx2 )2 ))
f (λx1 , λx2 ) = e−(w dw
0

w
Substituting = v,
λ
Z √x21 +x22
2
/(x21 +x22 ))
f (λx1 , λx2 ) = λ e−(v dv = λf (x1 , x2 )
0
p p
(Calculating
p range of v from range of w : 0 ≤ w ≤ λ x21 + x22 ⇐⇒ 0 ≤ λv ≤ λ x21 + x22 ⇐⇒ 0 ≤ v ≤
x21 + x22 )

12. The number of ways in which a beggar can be given at least 1 rupee from four 25-p coins, three 50-p coins and
2 one-rupee coins.
Total number of ways in which a beggar can be given money is (including the case when he is not given any
money) = 5 × 4 × 3 = 60 (There are 5 ways in which you can give 25-p coins (give him none of it, give 1 coin,
give 2 coins, give 3 coins, give 4 coins), similarly for 50-p and one-rupee coins). The number of ways in which
he is given at least 1 rupee is the number of ways he can be given money minus the number of ways he is given
at most 75 paise. He is given at most 75 paise in 6 ways (give him nothing, give him one 25-p coin, give him
two 25-p coins, give him one 50-p coin, give him three 25-p coins, give him one 25-p and one 50-p coin). So,
the total number of ways in which a beggar can be given at least 1 rupee is 60 − 6 = 54.
13. A randomly chosen group is tested for a disease. Within this group, each individual has a probability of 0.1 of
having the disease. A test is performed to identify the individuals with the disease. The test has two outcomes:
positive or negative. If the individual does not have the disease, the test outcome is “negative” 90 percent of
the time. If the individual has the disease, the test outcome is “negative” 20 percent of the time. Individuals
who test “positive” are sent to a hospital for further treatment. What is the probability that an individual
sent to the hospital indeed has the disease?
Pr(disease) = 0.1
Pr(negative|no disease) = 0.9
Pr(negative|disease) = 0.2
Pr(disease, positive) Pr(positive|disease) Pr(disease)
Pr(disease|positive) = =
Pr(positive) Pr(positive, disease) + Pr(positive, no disease)
Pr(positive|disease) Pr(disease)
=
Pr(positive|disease) Pr(disease) + Pr(positive|no disease) Pr(no disease)
0.8 × 0.1 8
= =
(0.8 × 0.1) + (0.1 × 0.9) 17
14. Let the function f : R++ → R++ be such that f (1) = 3 and f 0 (1) = 9, where R++ is the positive part of the
 1
f (1 + x) x
real line. Then lim equals
x→0 f (1)
  x1
f (1 + x) f (1+x)
lim e x log( )
1
lim = f (1)
x→0 f (1) x→0
f (1+x)
elimx→0 x log( f (1) ) . . . by continuity of exponential function
1
=
f 0 (1 + x)
 
1 f (1 + x)
lim log = lim . . . by L’Hospital’s Rule
x→0 x f (1) x→0 f (1 + x)
= 3
  x1
f (1 + x)
Thus, lim = e3
x→0 f (1)

15. Consider the two variable linear regression model where error Ui are iid with mean 0 and variance 1.

Yi = A + B(Xi − X̄) + Ui

7
 i is 1, 2, 3, . . . n
Let a and b be OLS estimates of A and B. Find correlation coefficient between a and b.
p
Cor(a, b) = Cov(a, b)/ Var(a)Var(b)

Now OLS estimates of A and B are

Pn
i=1 Yi
a =
Pnn
Yi (Xi − X̄)
b = Pi=1
n 2
i=1 (Xi − X̄)

Using Yi = A + B(Xi − X̄) + Ui we can rewrite a and b as

Pn Pn
i=1 A + B(Xi − X̄) + Ui Ui
a = = A + i=1
Pn n n Pn
i=1 (A + B(X i − X̄) + U i )(Xi − X̄) Ui (Xi − X̄)
b = P n 2
= B + Pi=1
n 2
i=1 (X i − X̄) i=1 (Xi − X̄)

Therefore, E(a) = A and E(b) = B because E(Ui ) = 0, and Xi s are non-stochastic.

Cov(a, b) = E((a − E(a))(b − E(b)))

Pn Pn
i=1 Ui Ui (Xi − X̄)
= E( Pi=1
n 2
)
n i=1 (Xi − X̄)
n Xn
1 X
= Pn E( Ui Uj (Xj − X̄))
n i=1 (Xi − X̄)2 i=1 j=1
n X n
1 X
= Pn 2
E(Ui Uj (Xj − X̄))
n i=1 (Xi − X̄) i=1 j=1
n X n
1 X
= Pn 2
(Xj − X̄)E(Ui Uj )
n i=1 (Xi − X̄) i=1 j=1
n
1 X
= Pn 2
(Xi − X̄)E(Ui2 )
n i=1 (X i − X̄) i=1
[E(Ui Uj ) = 0 for i 6= j because Ui s are iid with E(Ui ) = 0]
n
1 X
= Pn (Xi − X̄)
n i=1 (Xi − X̄)2 i=1
[because E(Ui2 ) = 1]
= 0

Therefore, Cor(a, b) = 0.
16. In a rectangular array (matrix) of distinct positive numbers, which has m rows and n columns, let x denote
the largest of the smallest number in each column, and y the smallest of the largest number in each row. Then
one can infer:
(a) x ≥ y
(b) y ≥ x
(c) x = y
(d) none of the above
Let aij denote the element in the ith row and jth column of the matrix where 1 ≤ i ≤ m and 1 ≤ j ≤ n. Let
x denote the largest of the smallest number in each column i.e.,

x = max [ min aij ]

1≤j≤n 1≤i≤m

y denote the smallest of the largest number in each row i.e.,

y = min [ max aij ]

1≤i≤m 1≤j≤n

8
 Now, Smallest number in a column is smaller than every number in that column. Given j,

min aij ≤ aij ∀1 ≤ i ≤ m

1≤i≤m

Therefore, the row produced by collecting the smallest numbers from each column will have a smaller maximum
than maximum of any row in the matrix.

x = max min aij ≤ max aij ∀1 ≤ i ≤ m

1≤j≤n 1≤i≤m 1≤j≤n

Now x is smaller than maximum of every row, hence it is smaller than the smallest of them.

x = max min aij ≤ min max aij = y

1≤j≤n 1≤i≤m 1≤i≤m 1≤j≤n

17. Evaluate √ √ √
1+ 2 + ··· + n−1
lim
n→∞ n3/2
Now,
R n−1 √ √ √ √ Rn√
0
xdx 1+ 2 + ··· + n−1 1
xdx
≤ ≤ ∀n (Think why?)
n3/2 n3/2 n3/2
Taking limits, R n−1 √ √ √ √ Rn√
0
xdx 1+ xdx2 + ··· + n−1
lim ≤ lim ≤ lim 1 3/2
n→∞ n3/2 n→∞ n→∞ n n3/2
Now Applying L’Hopital’s rule,
R n−1 √ √ r
0
xdx n−1 2 1 2
lim = lim 3 √ = lim 1− =
n→∞ n3/2 n→∞
2 n n→∞ 3 n 3

And Rn√ √
1
xdx n 2
lim = lim 3 √ =
n→∞ n3/2 n→∞
2 n
3
Hence, √ √ √
1+ 2 + ··· + n−1 2
lim =
n→∞ n3/2 3
18. (ME-1.13(2004))

x2 + y 2
 
−1
u = sin
x+y
 2
x + y2

sin u =
x+y
∂u 2x(x + y) − (x2 + y 2 ) x2 − y 2 + 2xy
cos u = 2
=
∂x (x + y) (x + y)2
∂u 2y(x + y) − (x2 + y 2 ) y 2 − x2 + 2xy
cos u = 2
=
∂y (x + y) (x + y)2
∂u ∂u x3 − xy 2 + 2x2 y + y 3 − yx2 + 2xy 2
x cos u + y cos u =
∂x ∂y (x + y)2
x3 + x2 y + y 3 + xy 2
=
(x + y)2
x2 (x + y) + y 2 (x + y)
=
(x + y)2
x2 + y 2
=
x+y
= sin u
∂u ∂u sin u
Hence, x +y = = tan u
∂x ∂y cos u

9
 1 1 1
19. (ME-1.16(2009)) Given n ≥ 9, µ = n 2 + n 3 + n 4 .
1 1 1
Clearly at n = 9, µ < n because n 4 < n 3 < n 2 = 3.
dµ 1 −1 1 −2 1 −3 1 −1 1 −1 1 −1 1
= n2 + n3 + n4 < n2 + n2 + n2 ≤ ∀n ≥ 9
dn 2 3 4 2 2 2 2
dµ
Since < 1 at all n ≥ 9 and at n = 9, µ < 9 therefore, µ < n ∀n ≥ 9.
dn
20. Suppose .5 units of good 2 are required in the production of one unit of good 1 and .5 units of good 1 are
required in the production of good 2.
(a) What is the input-output matrix A for this economy?
(b) What is the total production vector x = (x1 , x2 ) required for the economy to supply consumption vector
y = (3, 6) to consumers?

21. A price taking firm makes machine tools Y using labour and capital according to the following production
function
Y = K 0.25 L0.25
Labour can be hired at the beginning of every week, while capital can be hired only at the beginning of every
month. It is given that the wage rate = rental rate of capital = 10. Show that the short run (week) cost
function is 10Y 4 /K ∗ where the amount of capital is fixed at K ∗ and the long run (month) cost function is
20Y 2 .

10