You are on page 1of 2


Spring 2013
Tutorial 4 Solutions

1. Textbook Q12

a. The total cost of emissions reduction is minimized only when the
marginal costs are equal across all polluters, therefore a cost-effective
solution requires that MC
or that 300e
. Substituting
for e
in the formula e
=40 (since the policy goal is to reduce
emissions by 40 units) yields the solution. It is cost-effective for Firm
1 to reduce emissions by 10 units and for Firm 2 to reduce emissions
by 30 units.
b. In order to achieve cost-effective emission reductions, the emissions
fee should beset equal to $3,000. With this emissions fee, Firm 1
reduces 10 units and Firm 2 reduces 30 units, but Firm 1 has to pay
$3,000 for each unit of pollution they continue to produce, which gives
them a tax burden of $3,000 x 90 (Firm 1 generated 100 units in the
absence of government intervention) or $270,000. Firm 2 has a lower
tax burden because it is reducing emissions from 80 units to 50units.
Firm 2 pays $3,000 x 50 =$150,000. As the text concludes, the firm
that cuts back pollution less isnt really getting away with anything
because it has a larger tax liability than if it were to cut back more.
c. From an efficiency standpoint, the initial allocation of permits does not
matter. If the two firms could not trade permits, then Firm 2 would
have to undertake all of the emissions reduction. Initially, Firm 1s
MC is zero, while Firm 2s MC is $4,000, so there is a strong incentive
for Firm 2 to purchase permits from Firm 1. Trading should continue
until MC
, which is the cost-effective solution. This means that
the market price for permits will equal $3,000, the same as the
emissions fee. At this price, Firm 2 will purchase 10 permits from
Firm 2, allowing Firm 2 to reduce emissions by 30 rather than 40 and
requiring Firm 1 to reduce emissions by 10. This solution is the same
as the solution achieved with the emissions fee. However, Firm 1 is
better off because instead of having to pay taxes, it will receive a
payment of $30,000 for its permits. Firm 2 must pay $30,000 for the
extra permits, but it also avoids the payment of taxes. The government
lost $420,000 in tax revenue. The firms must still pay the cost of
emissions reduction, plus Firm 2 must pay for the permits purchased
from Firm 1.
2. On Pogue and Sgontz (1989).

a. Alcohol consumption generates positive surplus for consumers but excessive
consumption also generates negative externalities. So the objective of the
study is to determine the social welfare maximising alcohol tax rate for the US
market using basic micro tools of analysis.
b. Taxes are indiscriminate; they hurt i.e. reduce alcohol consumption of
alcohol abusers who generate negative externalities but also of non-abusers
who generate no external damage.
c. P is constant average and marginal private cost curve for alcohol supply.
E is marginal external cost or damage curve from alcohol supply
So P+E captures marginal social (private +social) costs of alcohol supply.
This curve is increasing at an increasing rate because external damages are
assumed to accelerate with increasing consumption. So for a mild abuser, an
additional drink is assumed to cause less additional harm than one additional
drink for a heavy abuser.
d. Equation (1) states that the welfare gain from any alcohol tax will be equal to
the reduced external damage less the reduced consumer surplus of alcohol
abusers less the reduced consumer surplus of non-abusers. Equations (2) and
(3) propose that the change in consumption by both groups can be calculated
using the percentage change in the alcohol price due to the tax (T/P)
multiplied by the own price elasticity of demand for alcohol (s). This is so
far for just one individual so these two quantities are multiplied by the
population in each group (xs). Equation (4) is a re-expression of (1) using (2)
instead of x
and (3) instead of x
. Equation (4) is now quadratic in T i.e. it
contains some T squared terms, hence it will look something like an upside
down U, with a local maximum. This can be calculated by differentiating (4)
with respect to T and then setting this equal to 0: this is what is done in (5).
Solving (5) for T gives us (6), the welfare maximising alcohol tax rate as a
function of external damage E, alcohol price P, elasticitys and the population
in each group.
e. As usual in applied studies, there is a deal of uncertainty over the data: In
column (1) the authors set out their best guesses re the values of the variables
needed to estimate the alcohol tax (case 1) but test the sensitivity of their
results by making other assumptions about elasticitys, populations in each
group, marginal damages, etc.
f. The authors conclusions are as follows: the 1983 optimal tax rate,
conservatively estimated, appears to be about double the actual rate in the
standard welfare model.. That is, in 1983 the alcohol tax rate was too low
hence the price of alcohol was too low relative to the welfare maximising
price. Too much alcohol consumption generating more than optimal levels of
external damage.
g. Do you get it? Do you now have a better understanding of how economic
theory can be used to provide policy advice? Do you think the assumptions of
the model are reasonable? Can you think of any criticisms of the model and/or
assumptions? Could this framework be applied to Australian data?