Economics 106P

UCLA
E. McDevitt STUDY QUESTIONS-SET #3

LINEAR PROGRAMMING
1. A firm produces tires by separate processes that require different quantities of capital (K), labor (L), and
raw materials (M). Process 1 requires one unit of K, four units of L, and two units of M to produce a tire
yielding a $4 profit. Process 2 requires one unit of K, two units of L, and four units of M to produce a tire
yielding a $6 profit. The available supply of capital is 10; of labor, 32; and of raw materials, 32.
a. Set up the objective function.
b. Give the constraints.
c. Find the profit maximizing combination T1 and T2 by graph. First draw the constraints, showing the
feasible region. Then draw a graph of the feasible region superimposing the objective function on it.
After doing this, confirm your answer using the slope rule.
d. Find the shadow prices of raw materials and labor. At what available quantity would the shadow price of
raw materials be zero?
2. To maintain his health, a person must fulfill certain minimum daily requirements for several kinds of
nutrients. Assume that only three kinds of nutrients need to be considered: calcium, protein and calories.
Also assume that the person's diet is to consist of only two food items, x1 and x2, whose price and nutrient
contents are shown in the table below, where we have also listed the minimum daily requirement for each
nutrient. What is the cost-minimizing combination of x1 and x2 that will satisfy the daily requirements?
Provide a graphical solution. Support your graphical solution using the slope rule. What is total cost
at this optimal combination? What is cost at the other corners?
x1 x2 Minimum Daily Requirement .
Price $0.60/unit $1/unit
Calcium 10 4 20
Protein 5 5 20
Calories 2 6 12

TRANSFER PRICING
1. Using a set of graphs, answer each question below.
a. Find the profit-maximizing quantity of the upstream product and downstream product and the optimal
transfer price assuming there is no outside market for the intermediate good.
b. Find the profit-maximizing quantity of the upstream product used and produced by the upstream
division, quantity of downstream product, and optimal transfer price. Assume a competitive outside market
for the intermediate good and that the competitive price is below the transfer price that would exist if there
were no outside market.
c. Find the profit-maximizing quantity of the intermediate good produced, the quantity sold to outside
buyers, the quantity "sold" to downstream division, and the optimal transfer price.
2. The House Products Division of Acme Corporation manufactures and sells digital clock radios.
A major component is supplied by the electronics division of Acme. The Cost functions for the radio and
the electronic component divisions are, respectively,
TCr = 30 + 2Qr and TCc= 70+6Qc + Qc2

Note that TCr does not include the cost of the component. Manufacture of one radio set requires the use of
one electronic component. The firm's demand curve for the digital clock radio is Pr = 108-Qr.
a. Assuming no outside market for the components, how many of them should be produced to maximize
profits for Acme as a whole? What is the optimal transfer price?
b. Suppose the components can be bought and sold in a competitive market for $50/unit. Find the profit-
maximizing quantity of components, the transfer price and the total number of components used by
downstream division.
c. Suppose Acme is the only producer of these components but can sell them in an outside market.
The demand for the component from the outside market is Pc,o = 80-Qc,o. Find total quantity of components
produced, the transfer price, the quantity of components sold outside (and the price) and the quantity
of components "sold" to the downstream division.
AUCTIONS AND BIDDING
1. Consider auctions where bidders have independent private values. Explain why it is rational to bid one's
true valuation in a second-price, sealed-bid auction.
2. Consider an auction where bidders have independent private values. Each bidder perceives that
valuations are evenly distributed between 0 and $10. Mary knows her own valuation is $2. Determine
Mary's optimal bidding strategy when: a. a first-price, sealed-bid auction is used with 1 other bidder, (b)
a first-price, sealed-bid auction is used with 3 other bidders, (c) a second-price, sealed-bid auction with 20
bidders.
3. Assume a private-value auction with English auction bidding rules. Suppose there are just two bidders
either of whom could have a value of $10 or $100 for the item in question. Each case is equally likely.
The minimal bid increment is $1 and ties are settled by flipping a coin. Explain how the seller can increase
its expected revenue by having a minimum bid and indicate what the minimum bid is.
4. a.Assume that all bidders are risk-neutral (only care about expected values) and draw their values from
the same interval (for simplicity assume the interval is 0 to $100). Given these assumptions, explain the
Revenue Equivalence Theorem for private value auctions.
4.b. Will the expected revenue also be the same for a common value auction?
4.c. How would your answer to parts (a) and (b) change if bidders were risk averse?
5. Explain what Winner's Curse is in common value auctions. How can a potential bidder use the
information provided by his or her estimated value to avoid the winner's curse?

ALLOCATION OF JOINT COSTS

1. An upstream division produces y at a marginal cost of $10 per unit. A unit of good y can then be used to
jointly produce a unit of good x1 and a unit of good x2. The marginal cost of converting y into x1 is
$4 in downstream division 1and the marginal cost of converting y into x2 is $2 in downstream division 2.
The demands for each product are: P1=100-Q1 and P2=80-Q2.
a. Suppose every unit of y is converted both into one unit of x1 and one unit of x2. Under this assumption,
what are the profit-maximizing quantities, prices and profit?
b. Can the firm do better by producing unequal amounts of x1 and x2? Explain. What is profit in this
case?
c. A social planner desires to maximize the social total. How many units of x1 and x2 should be produced
to accomplish this goal?
2. Assume the following interdependent demands for electricity: P1 = 26-Q1-0.5Q2 and P2 = 22-Q2-0.5Q1 .
MC1 = $1, MC2=$1, and MCjoint = $2. Using the “MR approach” find the profit-maximizing set of prices
and profit.
ANSWERS 1.a. π = $4T1 + $6T2.
b. For K: 1T1 + 1T2 < 10.
For L: 4T1 + 2T2 < 32.
For M: 2T1 + 4T2 < 32.
T1 > 0.
T2 > 0.
c. A graph of constraints.
T2 constraint for L
16

constraint for K

10

8 constraint for M

4 6 8 10 16 T1
The feasible region is the shaded area. The intersection points were found by first solving 2T1 + 4T2 = 32
(M constraint set as if binding) and 1T1 + 1T2 = 10 (K constraint set as if binding) to get T1 and T2.
T1 = 4 and T2 =6 (point b). And then solving 4T1 + 2T2 = 32 (L constraint set as if binding) and 1T1 +
1T2 = 10 (K constraint set as if binding) to get the T1 and T2 forming the other corner point. T1 = 6
and T2 =4 (point c).

T2

8 a

b
6

4 c
objective function

4 6 8 T1.
By trial-and-error:
Corner T1 T2 Profit = $4T1 + $6T2.
a 0 8 $48 .
b 4 6 $52 .
c 6 4 $48 .
d 8 0 $32 .
The profit-maximizing combination consists of 4 of T1 and 6 units of T2. Notice that the labor constraint is
not binding at this combination.

Or by using the slope rule:

For K: 1T1 + 1T2 = 10, or 10-T1 = T2 and thus the slope is -1.
For L: 4T1 + 2T2 = 32, or 16-2T1= T2 and thus the slope is -2.
For M: 2T1 + 4T2 = 32, or 8-0.5T1= T2 and thus the slope is -0.5.
For objective function: π = 4T1 + 6T2 , or (π/6)- 2/3T1= T2 and thus the slope of the objective function is -
2/3. Since this slope lies between -0.5 and -1, the corner formed by the K and M constraints is the optimal
corner.
d. Since the labor constraint is not binding, the shadow price is zero.
Step 1: For raw materials, suppose we increase the amount available by 4 units so that the new constraint is
For M: 2T1 + 4T2 < 36.
Step 2: find the new intersection point. The new intersection (call it point b' ) is found by solving 2T1 + 4T2
= 36 and 1T1 + 1T2 = 10 to get the T1 and T2 forming the new corner point.
Doing this, we get T1 = 2 and T2 = 8.
Step 3: Calculate profit at this new combination: π = 4(2)+ 6(8)= $56. Thus, Δπ = $4.
Step 4: Dividing Δπ by the change in M gives us its shadow price. Δπ/ΔM = $4/4=$1.
The vertical-axis intercept of the K constraint is 10. If M increased enough so that its vertical axis intercept
is greater than 10, then the M constraint would lie entirely outside of the K constraint and therefore no
longer be binding (which means a zero shadow price). To find the M at which this happens we set up the
following equation 0 + 4(10)= M' and solve for M'. Since M' = 40, it follows that if M is greater than 40
it is no longer binding.
2. The problem is to minimize cost subject to the daily requirement constraints.
First, we need to set up the objective function. Objective function. Cost = 0.6X1 + 1X2.
Second, we need to express the constraints.
Calcium 10X1 + 4X2 > 20
Protein 5X1 + 5X2 > 20
Calories 2X1 + 6X2 > 12 and X1, X2 > 0
For the graphical solution, we need first construct a graph. The feasible region is shown below as the
shaded area.
5

Calcium C.
4
b
Protein C.
2

0.67 2 3 4 6 X1

Calorie Constraint
To find intersection points, treat each constraint as if it were binding:
pt. b is at intersection of the calcium and protein constraint, so we use 10X1 + 4X2 = 20 and
5X1 + 5X2 = 20. We get X1 = 0.67 and X2 = 3.33. And at pt.c, where the protein and calorie constraints
cross, we get X1 = 3 and X2 = 1.
The bold line represents the objective function. Its point of contact is at pt.c --the optimal combination of
food items.
The optimal combination can also be found using the slope rule.
Slope of calcium constraint = -2.5 (read off of above graph). Slope of protein constraint= -1.
Slope of the calorie constraint= -0.33. The slope of the objective function, which can be written as
X2=C - 0.6X1, is -0.6. This lies between the slopes of the calorie and protein constraints. Therefore the
optimal combination occurs at the intersection of those two constraints.
Total cost at each corner:
Corner X1 X2 cost = 0.6X1 + 1X2.
a 0 5 $5
b 0.67 3.33 $3.73
c 3 1 $2.80
d 6 0 $3.60
ANSWERS
1. Case 1: No Outside Market

Pd
MC1

Ptransfer MCdownstream

Q
Qd=Qu
NMR1 MR

Case 2: Competitive Outside Market

Pd
MC1

MCdownstream
Ptransfer MCu*
D

Q
Qprod. Qused
NMR1 MR

Case 3: Monopoly Supplier of Upstream product to Outside Market

Pd

MC1
P1,out
Ptransfer
MCdownstream

Dout D
MRout
Total NMR

Q
Q1,out Q1,used Q1,produced
NMR1 MR
2.a.Before we begin to answer this question, let us first point out that MR = 108 -2Qr , MCr = 2 ,
MCc = 6+2Qc and MPc = 1(since one component is used to produce 1 radio).
Step 1. Find the net marginal revenue equation for components.
Recall that NMRc = (MR-MCr)(MPc), but since MPc = 1 we can write this as NMRc = (MR-MCr). Thus,
we have NMRc = MR - MCr = (108 - 2Q) - 2 = 106 -2Qc
Step 2: Set NMRc equal to MCc and solve for quantity:
106 -2Qc = 6+2Qc or Qc = 100/4= 25.
Step 3: The optimal transfer price is the marginal cost of these 25 components: MCc = 6+2(25) = $56.
2.b. The transfer price is equal to the competitive price of $50.Since $50 is below the transfer price that is
optimal when there is no outside market, the firm should buy some engines from outside.
Step 1. Set this $50 marginal cost of components equal to NMR for components to find TOTAL quantity
of components used and radios produced: 106 -2Qc = $50 Æ Qc = 28
Step 2. To find the quantity of components produced by upstream components division, set the transfer
price of $50 equal to the upstream's division marginal cost of producing components.
$50 = 6+2Qc which gives us a quantity of 22. Thus, the firm buys 28-22=6 components from outside
the firm and produces 22 components in its upstream components division.
c. First, notice that Pc,o = 80-Qc,o gives us a marginal revenue curve of MRc,o = 80 - 2Qc,o.
Step 1. Find the TOTAL NMR for components by horizontally summing the MR curve for the outside
market with the net marginal revenue from "sales" to the downstream division.
For outputs beyond 13 units (that is where the kink occurs in the sum of the marginal revenue two curves),
this is: Q = 93 - MR or when solving for MR gives us NMRc,total = 93 - Qc,total.
[NOTE: to add the curves horizontally, we must first solve for Q, we would give us :
Qc,o = 40-0.5MRc,o and Qc = 53-0.5MRc. Once you have done this, you then add them. After you add them,
you then solve for MR again.]
Step 2. Set the total NMR equation equal to MCc and solve for Q:
93 - Qc,total = 6+2Qc,total which gives us Qc,total = 29. This gives us total quantity of components
produced by the upstream division.
Step 3. To find the optimal transfer price we must determine the MC of producing 29 units.
MCc =6+2(29) = $64.
Step 4. To find sales to the outside market , set this transfer price equal to marginal revenue from outside
sales: 64 = MRc,o = 80 - 2Qc,o and solve for quantity. Qc,o = 8. If we plug this quantity into
the outside market demand curve we get Pc,o = 80-8 = $72.
Step 5. To find "sales" within the firm, set this transfer price equal the net marginal revenue from
"sales" to downstream division: 64= 106 -2Qc , or Qc = 21.
AUCTIONS AND BIDDING
1. First, consider the case in which a bidder bids above his true value-- bi > vi.
Either (a) bi > vi > b2 Æ win bidding and gain (vi-b2), but win anyway and still gain vi-b2 if bi = vi > b2
or (b) bi > b2 > vi Æ win bidding but suffer loss since b2>vi, whereas if b'> bi=vi avoid loss.
or (c ) b' > bi > vi Æ lose bidding, but lose anyway if bi=vi .
So there is nothing to gain and possibly something to lose if you bid above your true value.
Second, consider the case in which the bidder bids below his true value-- bi < vi.
Either (a) vi > bi > b2 Æ win bidding and gain vi-b2, but win anyway and still gain vi-b2 if bi = vi > b2
(b) vi > b' > bi Æ lose bidding, but if bi = vi > b' you would win bidding and gain vi - b'
(c ) b' > vi > bi Æ lose bidding, but lose anyway if b' > bi = vi .
So there is something to possibly gain and nothing to lose by bidding your true value.
2. a. The optimal bid in a first-price, sealed-bid auction is given by bi =n/(n+1)vi where n is the number
of other bidders (and the lower bound of the distribution is zero), bi is bidder i's bid, and vi is bidder i's
valuation. Thus, Mary's optimal bid is (0.5)(2) = $1.
b. Optimal bid is (3/4)(2)= $1.50.
c. Mary should be bid her true valuation of $2.
3. There are four cases which are equally probable: bidder 1 and bidder 2 might each have values of 10,
bidder 1 may have a value of 10 and bidder 2 has a value of 100, bidder 1 has a value of 100 and bidder 2
has a value of 10, and both bidders have a value of 100. We can express this as (10,10), (10,100), (100,10),
and (100,100). We must now determine the winning bids in each of the above cases. They would be
(10,11,11,100). The expected revenue to the seller is (1/4)(10+11+11+100)= $33.
Now suppose the seller sets the minimum bid at $100. Three-quarters of the time the seller will sell the
item for $100, and one-quarter of the time there will be no winning bid. Therefore, the
expected revenue is 3/4(100)+1/4(0) = $75. This exceeds the expected revenue yielded by the English
auction with no minimum bid.
4. This theorem states that the expected revenue generated in a private value auction will be the same for
an English auction, Second-price sealed-bid auction, First-price sealed-bid auction, and a Dutch auction.
The expected revenue generated will equal the value of the second-highest bidder in each case.
i. English auction--Suppose there are three bidders with values of $100, $80, and $60. Then the winning
bid will be slightly greater than $80 (say, $81). At that bid, only one bidder remains and so wins.
ii. Second-price,sealed bid--in this type of auction, the winner pays the amount equal to second highest bid.
Since it is optimal to bid your true value in this type of auction, the value of the second highest bidder will
be the price paid by the winning bidder. So just as in the case of an English auction, the revenue generated
equals the second highest bid. (See answer to question 3 for more details on this type of auction).
iii. First-price, sealed-bid (and Dutch auction)-- in this case the optimal bid equals an amount slightly above
the second highest bidder. We have already determined from previous discussion that for two bidders the
optimal bid would be bi = 1/2vi , for three bidders it is bi = 2/3vi, and so on. Assume bidder i has the highest
value. So that with two bidders the situation looks like this:

0 100
1/2 vi vi
From bidder i's point of view, the second highest bid is some value below vi, but each these values
is equally likely. Given that each is equally likely, then, on average, the second highest bid will be 1/2vi.
Thus, by choosing a bid strategy of bi = 1/2vi a bidder has effectively bid the value of the second highest
bidder. So we have the same outcome, on average, as in an English auction or a second-price, sealed-bid
auction. Again, assume bidder i has the highest value. So that with three bidders the situation looks like
this:

0 1/3vi 2/3vi vi 100

From bidder i's point of view the other bids lie below vi, but each of these values are equally likely.
Therefore, on average, the highest of these other two bids will be 2/3vi. Bidder i can therefore
expect that on average the second highest bid with two other bidders will be 2/3vi. By choosing a strategy
of bi = 2/3vi when there are two other bidders means that bidder i has effectively bid the value of the second
highest bidder. In fact, with total bidders of nT , and assuming you have the highest bid, then the next
highest bid will be, on average, [(nT-1)/(nT)]vi. Since nT = n+1 (where n is the number of OTHER
bidders), we can see by substitution that [(nT-1)/(nT)]vi = [(n+1-1)/(n+1)]vi = [n/(n+1)]vi. It was
demonstrated in class that the optimal bidding strategy is bi = [n/(n+1)]vi, and now we can see that this
is simply equal to expected second-highest bid, given that you have the highest bid. We have thus finished
our demonstration that the expected revenue generated in a private value auction will be the same for an
English auction, Second-price sealed-bid auction, First-price sealed-bid auction, and a Dutch auction.
4b. No. A comparison of expected revenues in common value auctions (with risk neutral bidders) will look
like English>Second price>First price = Dutch auction. Bidders will shrink their bids below his individual
value in order to avoid the winner's curse. The greater the uncertainty about the true value of the object, the
greater the likelihood of an overbid, and therefore the greater the incentive for bidders to reduce his bid.
There is less uncertainty in an English auction than in a sealed bid auction because a bidder can observe the
prices at which other bidders drop out of the bidding--this provides information about their valuations.
4c. For risk-averse bidders, a comparison of expected revenues in private-value auction looks like this:
First-price= Dutch> Second-price = English. Risk aversion plays no role in determining the optimal
bidding strategy in English or second-price auctions. In English style auctions, risk-averse bidders will still
remain active until the price exceeds their valuation and then drop out. The winner will still pay a price
equal to the second-highest valuation. Likewise, in a second-price auction,it is still optimal for a risk-averse
bidder to bid his true valuation. However, risk aversion induces bidders to bid more aggressively in a first-
price auction. In the risk neutral case, bidders shade down their bids to equal the expected value of the
second highest bid. However, by doing so, this increases the chance that some other bidder outbids them.
Risk-averse bidders are less willing than risk-neutral bidders to take this chance and therefore will
shrink their bids by a smaller amount.
For risk averse bidders, a comparison of expected revenues in common-value auction:
English>Second Price > or < First price = Dutch.
The information revealed in an English auction lessons the winner's curse. This reduction in risk induces
risk-averse bidders to bid more aggressively, on average, in an English auction than in a second price
auction. Therefore, the English auction always generates greater expected revenue than a second price
auction. It may even generate higher revenues than a first-price or Dutch auction.
5. Suppose bidders bid their estimated value. Winning the bidding conveys information to the winner that
all the other firms think the asset is worth less than the winner paid for it. The probability that all the other
firms are wrong and the winning bidder is correct is small. In other words, given that all estimates are
subject to errors, the winning bidder is likely to be the person with the largest overestimate of the value of
the asset. Notice that if all bidders could "pool" their information and average it, they would have a more
precise estimate of the true value of the asset.
To avoid the winner's curse, a bidder should revive downward his or her private estimate of the value to
account for the fact that the winner is the bidder who is most optimistic about the true value of the asset.
Suppose the true value of an asset is uniformly distribution between 0 and U , where U is the upper limit of
the distribution. On average, estimated values are equally spread out over this interval. With the expected
value of the highest of the nT order statistics from the interval [0,U] equal to [(nT)/(nT+1)][U] , while the
second highest is [(nT -1)/(nT+1)][U], and so on (nT = number of total bidders). Suppose you receive an
estimated value and desire to avoid the winner's curse. To do so, follow these steps:
Step 1. Draw an estimate of the value of the asset. Assume you have the highest estimated value of all the
bidders. For example, assume the estimated value is 500 and that there are six bidders in all.
Step 2. If this is the highest estimate of the six bidders, and given that we know that the expected value of
highest order statistic is [(nT)/(nT+1)][U] = [6/7][U], we can then estimate U. To so, [6/7][U] = [500] or,
solving for U, we get U = 583.33.
Step 3. Our best estimate of the mean of the distribution, and therefore of the true value of the asset, is the
midpoint of the distribution [0, 583.33]. The midpoint is (0+583.33)/2 = 291.67. This should be our bid.
ALLOCATION OF JOINT COSTS
1. a. Step 1: Set up the profit function given that Qy=Q1=Q2.
π =P1Q1 + P2Q2 - MC1Q1-MC2Q2-MCQy = (100-Q)(Q) + (80-Q)(Q)-4Q-2Q-10Q
= 164Q-2Q2.
dπ/dQ = 164-4Q = 0
Q = 41.
P1= 100-41=$59 and P2= 80-41=$39.
π = (59)(41)+(39)(41)-16(41)=$3,362
Alternatively, given the assumption Qy=Q1=Q2, we could have MR1+MR2 = MC1+MC2+MCjoint and solved
for Q. (100-2Q) + (80-2Q) = 16, or 180-4Q=16. Solving for Q gives us Q = 41.
10. b. Plug these quantities into the marginal revenue functions for each good. If MRi>MCi + MCjoint for a
good i then we can increase profit by producing more (the other good will have MR<MC and so produce
less of it), but we must allocate the cost of the joint input to that good ONLY.
MR1 = 100-(2)(41) = $18 > MC1=$4 and $18> $4+ $10 Æ not producing enough of x1 (since 18 > 4 and
the $10 joint cost).
MR2=80-(2)(41)= -$2 < MC2=$2 Æ producing too much of x2.
To determine the profit-maximizing amount of each good, we allocate the cost of producing y to
good x1. For good 1: π1 = P1Q1-MC1Q1-MCQy = (100-Q1)Q1-4Q1-10Q1 = 86Q1-Q12
dπ1/dQ1 = 86-2Q1 = 0 or Q1= 43.
P1= 100-43=$57 and π1 = (57-14)(43) = $1,849
For good 2: π2 = P2Q2-MC2Q2 = (80-Q2)Q2-2Q1 = 78Q1-Q12
dπ2/dQ2 = 78-2Q1 = 0 or Q2= 39.
P2= 80-39=$41 and π2 = (41-2)(39) = $1,521
Alternatively, we could have solved for the quantities by have MR1 = MC1+MCjoint and MR2=MC2 .
100-2Q1 = 14, we gives us Q1 = 43, and 80-2Q2 = 2 and so Q2 = 39.
π1 + π2 = $1,849+$1,521= $3,370. Thus, total profit is greater $3,370> $3,362.
c. First. TV1 = 100Q1-0.5Q12 and TV2 = 80Q2=0.5Q22 . Let us first assume that the firm converts each unit
of y into one unit of x1 and one unit of x2. Thus, the objective function is:
Social Total (ST) =(100Q-0.5Q2 ) + (80Q=0.5Q2 ) –10Q-4Q-2Q = 180Q-Q2 -16Q = 164Q-Q2 .
dST/dQ = 164-2Q = 0, or Q = 82. So ST = 164(82)-(82)2 = $6,724.
Alternatively, given that the firm converts each unit of y into one unit of x1 and one unit of x2 we could
have solved for these quantities by MV1+MV2 = MC1+MC2+MCjoint . Thus,
(100-Q)+(80-Q) = 16, or 180 –2Q = 16. This gives us Q= 82.
Second. But the social planner can do better than this. To see why, plug Q=82 into MV1 and MV2 and
compare with MC1 and MC2. MV1 = 100-Q1= 100-82= $18 > MC1 = $4, and $18> $4+$10. Also, MV2 =
80- Q2= 80-82= -$2 < MC2 = $2. Thus, the firm is producing too little of good 1 (MV1 exceeds MC1 plus
MCy) and too much of good 2 (MV2 is below MC2 for good 2). We should therefore allocate the entire
joint cost to good 1 and proceed as follows: Maximize ST1 = (100Q1-0.5Q12 ) – 14Q1 with respect to Q1
and Maximize (80Q2-0.5Q22 ) – 2Q2 with respect to Q2 .
dST1/dQ1 = 86-Q1 = 0, or Q1 = 86. And dST2/dQ2 = 78-Q2 = 0, or Q2 = 78.
Alternatively, we could have solved for quantities by MV1= MC1+ MCjoint and MV2 = MC2.
100-Q1 = 14, or Q1 = 86, and 80-Q2=2, or Q2= 78.
Third. The social planner would charge P1 =100-Q1 = 100-86=$14 and P2 =100-Q2 = 80-78=$2.
ST1+ST2 = [86(86)-0.5(86)2 ]+ [78(78)-0.5(78)2] = $3698 + $3,042 = $6,740.
2. First, assume Q1=Q2=Q. Thus, we solve MR1+MR2 = MC1 + MC2+ MCjoint.
TR1 = 26Q1-Q12-0.5Q2Q1 and TR2 = 22Q2-Q22-0.5Q2Q1 .
MR1= ∂TR1/∂Q1 + ∂TR2/∂Q1= 26-2Q1-0.5Q2-0.5Q2 = 26-2Q1-Q2 .
and MR2 = 22-2Q2-0.5Q1-0.5Q1 = 22-2Q2-Q1 .
So, given Q1=Q2=Q, we can write (26-2Q1-Q2 ) + (22-2Q2-Q1 ) = 4, or 48 –6Q= 4. Q=7.333.
P1 = 26-Q1-0.5Q2 = 26-7.333-0.5(7.333) = $15 and P2 = 22-Q2-0.5Q1 = 22-7.333-0.5(7.333)= $11.
Profit= (15)(7.333)+(11)(7.333) – 4(7.333) = $161.33.
Second , plug into respective MR functions and compare with MC functions.
MR1= 26-2Q1-Q2 = 26-2(7.333)-7.3333= $4 > MC1 =1 and $4 > MC1+MCJoint = $3. This tells us the firm is
not producing enough of good 1.
MR2= 22-2Q2-Q1 = 22-2(7.333)-7.333 = 0 < MC2 = $1. Producing too much of good 2. We should
therefore allocate the entire joint cost to good 1.
Third, MR1 = MC1 + MCjoint and MR2 + MC2.
26-2Q1-Q2 = 3 and 22-2Q2-Q1 =1. The first equation gives us 11.5-0.5Q2 = Q1 and the second equation
tells us 21-2Q2= Q1. Setting these equal to each other and solving for good 2:
11.5-0.5Q2 = 21-2Q2 Æ 9.5 = 1.5Q2 or Q2 = 6.333. So us 21-2(6.333)= Q1 = 8.333.
Fourth, P1 = 26-Q1-0.5Q2 = 26-8.333-0.5(6.333) = $14.5 and P2=22-Q2-0.5Q1 –22-6.333-0.5(8.333)= $11.5
Profit = (14.5)(8.333)+(11.5)(6.333)-3(8.333)-1(6.333) = $120.83+72.83-25-6.333= $162.33.