ECON 255, HOMEWORK 4.

AIR POLLUTION
DUE MAR 13

(1) L&T Chapter 14 (Pollution Control) Self Test Exercise 1. Two firms
can control emissions at the following marginal costs: M C1 = 200q1 ,
and M C2 = 100q2 , where q1 and q2 are, respectively the amount of
emissions reduced by the first and second firms. Assume that with
no control at all, each firm would be emitting 20 units or a total
of 40 units for both firms. Compute the cost-effective allocation of
control responsibility if a total reduction of 21 units of emissions is
[Link] We need a solution, (q1∗ , q2∗ ), that satisfies two
conditions. First we need the total reduction to be 21.

q1∗ + q2∗ = 21

Second, we the allocation will not be cost-effective unless the mar-

ginal costs of each firm evaluated at the allocation of emissions re-
ductions are equal:

M C1 (q1∗ ) = M C2 (q2∗ )
⇒200q1∗ = 100q2∗

Substituting the first condition in the equation for the second we

get

200q1∗ = 100(21 − q1∗ )

⇒300q1∗ = 2100
⇒q1∗ = 7
⇒q2∗ = 14

(2) L&T Chapter 14 (Pollution Control) Self Test Exercise 2. Assume

that the control authority wanted to reach its objective in the previ-
ous problem by using an emissions charge system. (a) What per-unit
charge should be imposed? (b) How much revenue would the control
1
2 ECON 255, HOMEWORK 4. AIR POLLUTION DUE MAR 13

authority collect? ANSWER Notice that if we plug either q1∗ or q2∗

back into their respective MC functions we find that marginal cost
at the cost effective solution is $1400 per unit:

M C1 (q1∗ ) = 200 ∗ 7 = 1400

M C2 (q2∗ ) = 100 ∗ 14 = 1400

Suppose the control authority set a tax equal to $1400 for each
unit emitted. This would incentive each firm to reduce emissions
until their MC is that level and therefore we would get exactly the
same allocation of emission reduction as in the last problem. Notice
that while firm 1 would reduce emission by 7 units, it would continue
to emit 13 and would therefore pay r1 in emission charges as follows.

$1400
r1 = ∗ 13 units = $18200
unit
Meanwhile firm 2 would reduce by 14 and continue to emit 6 units.

$1400
r2 = ∗ 6 units = $8400
unit
therefore the total collected by the control authority would be
r1 + r2 = 18200 + 8400 or $26,600.
(3) L&T Chapter 14 (Pollution Control) Self Test Exercise 3. In a region
that must reduce emissions, three polluters currently emit 30 units
of emissions. The three firms have the following marginal abatement
cost functions that describe how marginal costs vary with the amount
of emissions each firm reduces?

M AC1 (q1 ) = .75 + .5q1

M AC2 (q2 ) = .5 + q2
M AC3 (q3 ) = 1.5 + q3

The table below (reproduced from the text) shows the marginal
cost of discrete units for each firm . Suppose this region needs to
reduce emissions by 14 units and plans to do it using a form of
cap-and-trade that auctions allowances off to the highest bidder.
ECON 255, HOMEWORK 4. AIR POLLUTION DUE MAR 13 3

(a) How many allowances will the control authority auction off?
Why? ANSWER. Since they want to reduce emissions by
14 from current level of 30, they would want to create 16 al-
lowances. Auctioning off all 16 would maximize revenue.
(b) Assuming no market power, how many of the allowances would
each firm be expected to buy?
(c) Assuming that demand equals supply, what price would be paid
for those allowances? ANSWER We can convert each MAC
curve to a demand curve for pollution. For example take firm 1.
Let y1 = 10−q1 be the amount of pollution emitted. (Since 10 is
the original amount emitted and q1 is defined in the problem as
the amount abated). The gives us an inverse (vertical) demand
curve for firm 1 as follows.

p1 (y1 ) = .75 + .5(10 − y1 )

p(y1 ) = 5.75 − .5y1
Converting this to a (horizontal) demand equation, we have

y1 (p) = 11.5 − 2p
Similarly for firm 2 and 3 we would have

y2 (p) = 10.5 − p
y3 (p) = 11.5 − p
If we add up the demands for each firm we get

Y (p) = y1 (p) + y2 (p) + y3 (p)

= 33.5 − 4p
Setting this equal to 16 - the target level of pollution we can
find the price that will get us there.

Y (p∗ ) = 16
33.5 − 4p∗ = 16
p∗ = 4.375
4 ECON 255, HOMEWORK 4. AIR POLLUTION DUE MAR 13

(d) If the control authority decided to use an emissions tax rather

than cap-and-trade, what tax rate would achieve the 14-unit
reduction cost-effectively? ANSWER It would need to set
the tax to the same market clearing price from the allowance
market. Not that weather policy is a tax set to p∗ = 4.375
or a system of tradeable allowances the allocation of pollution
emissions (y) and reduction would be identical. Namely

y1 (p) = 11.5 − 2(4.375) = 2.75 ⇒ q1∗ = 7.25

y2 (p) = 10.5 − 4.375 = 6.125 ⇒ q2∗ = 3.875
y3 (p) = 11.5 − 4.375 = 7.125 ⇒ q3∗ = 2.875

Yes this would represent cost-effective reduction, since it equal-

izes the marginal cost of abatement across all firms. Note that
one would get almost the exact same answer by simply staring
at the table showing the marginal abatement cost schedule of
each firm below. The method in that case would involve finding
the lowest marginal cost unit of abatement, crossing it out and
repeating until 14 units were crossed out. In the table below
that would mean crossing out all of the units whose marginal
cost of abatement is $4 or less. (The price to achieve this would
be somewhere above $4 and below $4.5.) Why does this sound
a little different from the solution above? Its just because if
you use the table inspection method just described, you are im-
plicitly assuming that units have to be abated in discrete whole
units, while in the method that led to a price of $4.375, we im-
plicitly assumed that pollution could be abated in any amount.
 ECON 255, HOMEWORK 4. AIR POLLUTION DUE MAR 13 5

Firm Firm 1 Firm 2 Firm 3

Emissions Marginal Marginal Marginal
Reductions Cost Cost Cost
1 $1.00 $1.00 $2.00
2 $1.50 $2.00 $3.00
3 $2.00 $3.00 $4.00
4 $2.50 $4.00 $5.00
5 $3.00 $5.00 $6.00
6 $3.50 $6.00 $7.00
7 $4.00 $7.00 $8.00
8 $4.50 $8.00 $9.00
9 $5.00 $9.00 $10.00
10 $5.50 $10.00 $11.00

(4) L&T Chapter 15 (Stationary Source Air Pollution) Self Test Exer-
cise 1. Would imposing the same tax rate on every unit of emissions
normally be expected to yield a cost-effective allocation of pollution
control responsibility? Does you answer depend on whether the en-
vironmental target is an aggregate emissions reduction or meeting an
ambient standard? Explain. ANSWER, yes it depends. In the case
where the target is simply an aggregate emission level (appropriate,
for example, in the case of green-house gas emissions), then a single
price induced by a uniform tax or a competitive market of trade-
able permits will always be the most cost effective way to achieve
the target. However, if the control authority is trying to achieve an
ambient standard then cost-effectiveness would likely demand that
prices be adjusted to local conditions. In places where there is more
pollution or in places where abatement is more difficult (because,
for example, of different mix of local industry), prices would need to
be higher to achieve the same ambient levels as in places with fewer
sources or sources that have easier ways of abating. One interesting
implication, however, in the case where prices adjust to keep pollu-
tions below levels that would violate the ambient standard is that
the pollution price then becomes a factor in the location decisions of
firms which, in the long run, would keep prices in different regions
from diverging too much.
6 ECON 255, HOMEWORK 4. AIR POLLUTION DUE MAR 13

(5) Suppose that the marginal damage caused by some form of pollution
were $75 / ton. Demand for the right to pollute has the following
form

Pd (Q) = a − Q

where Q is the quantity of pollution. (Pd (Q) can be interpreted

as the price at which a given quantity, Q, would be demanded or
the marginal willing to pay, MWTP, for the next or last unit of
pollution). However there is uncertainty about the size of demand.
In other words, we do not know whether a = 100 or a = 150. We
think that it could be either with probability 21 each.
(a) If the regulator uses a cap-and-trade system, what is the optimal
choice for the cap? What is the expected dead-weight loss?
ANSWER. First note that if the true demand turns out to be
the low (L) case (a=100), then the efficient quantity would be
L = 25, since that is the quantity that would equate MC and
qef f
marginal willingness to pay in that case. Similarly qef H = 75.
f
Intuitively the right choice when we are unsure of demand is
to split the different and set it equal to 50. Let’s check our
intuition. The dead weight loss of setting the quantity to any
target value, qτ in the high case (H) is

1
DW LH = (pH (qτ ) − M D)(qefH
f − qτ )
2
1
= (150 − qτ − 75)(75 − qτ )
2
1
= (75 − qτ )(75 − qτ )
2

This is just the area of the standard DLW triangle of setting

the target below the efficient level. Similarly if demand turned
out to be low then DWL would be
ECON 255, HOMEWORK 4. AIR POLLUTION DUE MAR 13 7

1
DW LL = (M D − pL (qτ ))(qτ − qef
L
f)
2
1
= (75 − (100 − qτ ))(qτ − 25)
2
1
= (qτ − 25)(qτ − 25)
2

The control authority’s problem is to minimize expected DWL

which is 12 DW LH + 21 DW LL since each true state of demand
(H or L) occurs with probability 12 . Therefore they want to
minimize the following amount:

1 1
E(DW L) = DW LH + DW LL
2 2
11 11
= (75 − qτ )(75 − qτ ) + (qτ − 25)(qτ − 25)
22 22
1 1
= (75 − qτ )2 + (qτ − 25)2
4 4

Setting the first derivative of that equal to zero we get

dE(DW L) 1
= − (75 − qτ ) + (qτ − 25) = 0
dqτ 2
⇒qτ = 50

Intuition confirmed! Note this works out so nicely as the amount

exactly half way between the H and L efficient target levels
because (i) demand is linear (albeit uncertain) (ii) MD is linear
and (iii) the probabilities are equal. If any of those assumptions
were altered the expected DWL minimizing choice would not be
exactly the midpoint, but its a good benchmark to keep in mind.
Can you say how one should adjust, if for example the marginal
damage curve is convex?
(b) If the regulator uses an emissions charge or tax, what is the
optimal choice for the rate? What is the expected dead-weight
loss? ANSWER. In this case, the efficient policy is simply to
set the tax equal to MD. Then no matter what demand turns
out to be, the tax will achieve the efficient pollution level. DWL
8 ECON 255, HOMEWORK 4. AIR POLLUTION DUE MAR 13

is zero! Note this illustrates the basic message in Weitzman’s

famous paper “Prices versus Quantities”. The ideal approach,
in general, is the one where the artificial supply curve induced
by the policy most closely mimics the MD curve. Since the MD
is flat at $75, the most efficient policy is the one that generates
a constant price of $75 which is exactly what a tax or emissions
charge would do. Suppose instead that the MD curve were very
steep? Would we get the same result?