Chapter 3
Government Intervention in
Market Failure
© 2004 Thomson Learning/South-Western
Topics in Chapter 3
1.
2.
3.
4.
5.
Should the Government Intervene? Are
there private solutions that will work?
Types of Government Intervention – general
introduction
The “optimal” level of environmental quality
Government intervention: Command and
Control policies
Government intervention: Economic
incentives
2
Should the Government Intervene?
Pigouvian Taxes



A.C. Pigou (1938) argued that an externality
cannot be mitigated by contractual
negotiation between the affected parties.
Pigou argued that direct coercion by the
government or judicious use of taxes should
be used against the offending party.
These taxes are referred to as Pigouvian
taxes.
3
Pigouvian Taxes



The basic principle behind the use of
externality taxes is that the tax eliminates the
divergence between the Marginal Private
Cost (MPC) and the Marginal Social Cost
(MSC).
Q1 represents the market equilibrium (where
MPC=MPB), and
Q* represents the optimal level of output
(where MSC=MSB).
4
An Externality Tax on Output
5
An Externality Tax on Output
MSC = MPC + MDpollution
$
a
MPC1
b
Demand
Q*
Q1
Quantity of steel
6
Pigouvian Taxes





An externalities tax equal to the divergence between
MPC and MSC would raise the steel firms’ private
costs.
The tax would shift the MPC curve by an amount
equal to the distance from a to b in Figure 3.1.
The market would arrive at an optimal equilibrium of
Q*.
This is known as internalizing an externality.
More precisely, the tax should be placed on the
externality itself (the amount of pollution emissions)
rather than on output (amount of steel).
7
Coase Theorem




Ronald Coase (1960) argued that not only is
a tax unnecessary, it is often undesirable.
Coase argued:
The market will automatically generate the
optimal level of the externality.
This optimal level of the externality will be
generated regardless of the initial allocation
of property rights.
8
Coase Theorem





One example to illustrate his theory is based on the
interaction of a cattle rancher and a crop farmer.
Cattle occasionally leave rancher’s property and
damage farmer’s crop.
Coase argued that the farmer and rancher will reach
an agreement that will make them both better off.
Either the rancher will accept payment to reduce the
size of the herd or farmer will accept payment to
cover cost of crops lost.
And this will happen without government
intervention.
9
Another example: Dorm room stereos and
studying
$
MC loudness to you
D = MB loudness to partier
Q0
Q*
QL
Loudness
10
Dorm room stereos and studying

If property rights belong to partier, where is
initial noise level? QL

But there are gains from trade until move
back to Q*

If property rights belong to partier, where is
initial noise level? Q0
Again, gains from trade until get to Q*
Gains to be split between two parties are
denoted “A” and “B” in diagram


11
Another example: Dorm room stereos
and studying
$
D = MB loudness to partier
MC loudness to you
A
B
Q0
Q*
QL
Loudness
12
Coase Theorem



If there are no transaction costs and property rights
are well defined, then voluntary transactions will
eliminate any distortions in resource allocation
stemming from an externality and the outcome is
independent of the property rights
This version of the “theorem” is from Baumol and
Oates text “The Theory of Environmental Policy.”
Emphasizes private behavior and importance of
transaction costs
13
Coase Theorem





What happens if impose a Pigouvian tax on the
generator of the externality, would this result in an
efficient outcome?
Set a tax equal to marginal damage at the optimal to
shift the demand for loudness
After the tax, are there still gains from trade?
Would tax be a good idea?
This is basis for Coase’s argument that government
intervention could make things worse
14
Coase with a tax per unit of Loudness
$
MC loudness to you
D = MB loudness to partier - tax
tax
Q0
QN
Q*
D = MB loudness to partier
QL
Loudness
15
Criticisms: Coase Theorem



Two important assumptions: transactions
costs are insignificant and property rights well
defined.
Transactions costs are costs associated with
arriving at an agreement (the costs of
negotiation).
These may be small for a 2 party agreement
but would be very large for an externality
such as sulfur dioxide emissions across
North America.
16
Coase Theorem



The number of participants makes
transactions costs important.
One way to reduce transactions costs is to
appoint an agent who acts in behalf of a large
number of people.
The use of agents is associated with its own
problems:


Free riders – don’t share in cost, but share
benefits.
Often it is difficult for individuals to identify the
agent that will best represent their view point.
17
Coase Theorem


Another problem associated with the Coase
example can occur when the allocation of
property rights would signal entry and exit in
response to those rights.
If ranchers have the right to let their cattle
roam without worrying about paying
damages, then there can be an increase in
the number of ranchers, and more damage.
18
Bottom Line on Coase arguments



Probably are cases where private
negotiations can be effective
In those cases, government should stay out
But, probably plenty of cases where
transaction costs and other issues lead to
need for intervention
19
Types of Government Intervention

There are five broad classes of government
intervention:






Moral suasion
Direct production of environmental quality
Pollution prevention
Command and control regulations
Economic incentives
Each of these represents a different
philosophy toward the role of government in
society.
20
Moral Suasion



This term is used to describe government
attempts to influence behavior without
actually stipulating any rules.
Effectiveness depends upon the extent to
which individuals believe it is in their
collective interest to do so.
Successful programs include Woodsy Owl’s
“Give a hoot, don’t pollute” and Smokey
Bear’s “Only you can prevent forest fires.”
21
22
Direct Production of Environmental
Quality

Includes







reforestation,
breaching of dams,
stocking of fish,
creation of wetlands,
treatment of sewage, and
toxic waste site cleanup.
These are sometimes ameliorative actions.
23
Pollution Prevention



Designed to address market failure of imperfect
information, in some cases there may be
technologies that could be developed that save
firm’s money and improve environment
Basic premise is that combined efforts of
government agencies, national laboratories,
university and private firms can lead to development
of innovative and beneficial technologies.
These programs emphasize being proactive in
reducing pollution, encourage R&D and adoption of
“green technologies” .
24
Command and Control Regulation



These place constraints on the behavior of
households and firms.
Constraints generally take the form of limits
on inputs or outputs in the consumption or
production process.
Examples include:


Requiring sulfur-removing scrubbers on the
smokestacks of coal-burning utilities.
Prohibitions against dumping of toxic substances.
25
Economic Incentives


Economic incentives make self interest
coincide with social interest.
Examples include:






Pollution taxes
Pollution subsidies
Marketable pollution permits
Deposit-refund systems
Performance bonds
Liability systems
26
Choosing the Correct Level of
Environmental Quality

Zero pollution is not possible/desirable for two
reasons:




The reduction of pollution will have opportunity costs.
The Law of Mass Balance makes a choice of zero
physically impossible.
The Law of Mass Balance states that the mass of
outputs of any activity are equal to the mass of
inputs.
Any consumption or production activity must
produce waste.
27
Choosing the Correct Level of
Environmental Quality
Definitions first:
 Stock pollutants: pollutants for which
environment has little ability to absorb: non
biodegradable bottles, heavy metals, toxics
 Fund pollutants: environment has some
ability to absorb, pollutant doesn’t accumulate
indefinitely; organic pollutants, CO2 absorbed

by plants, etc.
Focus now on Fund Pollutants
28
Choosing the Correct Level of
Environmental Quality



The desired level of pollution will be a
function of the social costs associated with
pollution.
The first of these is the damage that pollution
creates by degrading the physical, natural,
and social environment.
The second is the cost of reducing pollution
and includes the opportunity costs of
resources used to reduce pollution and the
value of foregone outputs.
29
The Marginal Damage Function

The marginal damage function represents the
damages that pollution generates by
degrading the environment.

Even if these impacts are not quantifiable, the
marginal damage function is useful for
thinking about the relationship between
environmental change and social welfare.
30
Figure 3.3 Marginal Damage Function
31
Marginal Damage Function

The marginal damage function specifies the
damages associated with an additional unit of
pollution.

The total damages generated by a particular
level of pollution is represented by the area
under the marginal damage function.
32
Marginal Damage Function

The increasing slope of the marginal damage
function indicates how damage changes with
each additional unit of pollution.

An upward sloping marginal damage function
indicates that as the level of pollution
becomes larger, the damages associated
with the marginal unit of pollution become
larger.
33
Marginal Abatement Cost Function


Abatement Costs are those costs associated
with reducing pollution to a lower level so that
there are fewer damages.
Abatement costs include:




Labor
Capital
Energy needed to lessen emissions
Opportunity costs from reducing levels of
production or consumption.
34
Marginal Abatement Cost Function



The marginal abatement cost function
represents the costs of reducing pollution by
one more unit.
In the following figure, Eu represents the level
of pollution that would be generated in
absence of any government intervention.
As pollution is reduced below Eu, the
marginal abatement cost increases.
35
Marginal Abatement Cost Function
36
Marginal Abatement Cost Function



Marginal abatement costs rise as cheaper
options for reducing pollution are exhausted
and more expensive steps must be taken.
The decreasing slope indicates that the costs
of reducing pollution increases at an
increasing rate.
A high vertical intercept indicates that the
cost of eliminating the last few units of
pollutants would be extremely high.
37
The Optimal Level of Pollution

Optimal level of pollution minimizes the total
social costs of pollution (the sum of total
abatement costs and total damages).

This level occurs at the point where marginal
abatement costs are equal to marginal
damages.
38
The Optimal Level of Pollution
39
The Optimal Level of Pollution




If the level of emissions is less than E1, then the
marginal abatement costs are greater than the
marginal damages that the unit of pollution would
have caused.
It doesn’t make sense to reduce pollution.
If the level of emissions are greater than E1, then the
marginal damages are greater than the marginal
abatement costs associated with reducing pollution
by one unit.
Society is better off eliminating that unit of pollution.
40
Social Costs When Pollution Level is
Greater than Optimal
41
Social Costs When Pollution Level is
Greater than Optimal




The optimal level of pollution is E1.
The actual level of pollution is E2.
Total costs associated with pollution have
been increased by the area of triangle abc.
This represents marginal damages greater
than marginal abatement costs for the range
of pollution emissions between E1 and E2.
42
Social Costs When Pollution Level is Less
Than the Optimal
43
Social Costs When Pollution Level is Less
than Optimal




The optimal level of pollution is E1.
The actual level of pollution is E3.
Total costs associated with pollution have
been increased by the area of triangle ade.
This represents marginal abatement costs
greater than marginal damage for the range
of pollution emissions between E1 and E3.
44
Optimal Level of Pollution, an alternative
graphical representation
damages,
costs, $
MAC
MDF = MB
abatement
A1
Abatement
45
Optimal Level of Pollution, an alternative
approach




Plot functions against “abatement” instead of
pollution
Abatement is the amount of pollution reduced
These are analagous approaches, just
sometimes more convenient to think in terms
of abatement vs. pollution
Answers are the same.
46
Optimal Level of Pollution, two
approaches on one graph
= MB emissions
= MB abatement
abatement
47
Optimal pollution (abatement) levels and
costs of control

Two goals of environmental policy
1.
Get the optimal amount of pollution
(abatement) – just discussed
Achieve that level at the lowest possible
cost
Goals are actually inter related, but helpful
to think about them in two steps, once have
identified optimal amount of pollution, how
to achieve it at least cost
2.

48
Optimal pollution (abatement) levels and
costs of control




Suppose optimal to control (abate) 100 units
of pollution that are generated by two firms
How much control should each firm
undertake to minimize total costs
Plot abatement levels by the two firms
against each other with a total of 100 units
of abatement
Cost of control is at a minimum when the
marginal abatement costs are equal
49
Least cost allocation of abatement
between two sources (firms)
damages,
costs, $
MAC1
MAC2
a
b
0 10 20 30 40 50 60 70 80 90 100
100 90 80 70 60 50 40 30 20 10 0
Abatement firm 1
Abatement firm 2
50
Optimal pollution (abatement) levels and
costs of control




In this example firm 1 should control 40
units and firm 2 should control 60 to achieve
least cost of control
This solution takes into account the fact that
different firms have different costs of control
Can consider both goals on one graph
Can see both that optimal abatement is 100
and efficient (least cost) allocation is 40, 60
51
Optimal pollution and least cost
allocation of abatement
damages,
costs, $
MAC1
MAC2
MAC = MAC1
+ MAC2
MB
0 10 20 30 40 50 60 70 80 90 100 110 120 130
Abatement
52
Pursuing Environmental Quality with
Command and Control Policies



One way to achieve an optimal level of
pollution is to mandate action to achieve the
desired level of pollution.
Critics have argued that command and
control regulations generate more abatement
costs than necessary.
Suppose there is a desire to reduce pollution
by half, each firm might be required to control
half of its emissions, would this be the least
cost way to accomplish this reduction?
53
Pursuing Environmental Quality with
Command and Control Policies
54
Pursuing Environmental Quality with
Command and Control Policies




Recall, the aggregate marginal abatement cost
function is the horizontal summation of the
individual marginal abatement cost functions.
Note we are back to plotting against emissions
With no environmental regulation, polluter 1
would emit 10 units and polluter 2 would emit 6.
A requirement to reduce emissions by 50%,
regardless of cost, would reduce polluter 1 to 5
units and polluter 2 to 3 units.
55
Pursuing Environmental Quality by
Equating Marginal Abatement Costs



When both polluters are required to reduce
emissions by 50%, regardless of marginal
abatement costs, polluter 2 incurs a higher
cost ($3) than polluter 1 ($2).
Society’s total abatement costs can be
lowered by keeping total emissions constant,
but reallocating level of emissions by
marginal abatement costs.
The optimal level of emissions will be where
marginal abatement costs are equal, for a
given level of emission.
56
Pursuing Environmental Quality by
Equating Marginal Abatement Costs
57
Pursuing Environmental Quality by
Equating Marginal Abatement Costs




Since polluter 2 has higher marginal abatement
costs, polluter 2 should be allowed to emit more,
and polluter 1 will be required to pollute less.
Polluter 1 reduces pollution by one half unit (to 4 ½)
and polluter 2 increases pollution by one half unit (to
3 ½).
Polluter 1’s marginal abatement costs increase and
polluter 2’s marginal abatement costs decrease.
Total abatement costs are minimized.
58
The Role of Command and Control
Policies

Despite their typical inability to equate marginal
abatement costs across polluters, command and
control policies may still be the most desirable policy
instrument under the following circumstances:
 When monitoring costs are high.
 When the optimal level of emissions is at or near
zero.
 During random events or emergencies that can
change the relationship between emissions and
damages.
59
The Role of Command and Control
Policies



While it might be possible to achieve an
optimal amount of litter through the use of a
tax or per person allocation, this would
require the “litter police”.
It is easier to make ALL littering illegal and
establish a punitive fine for those caught
littering.
The fine multiplied by the probability of being
caught would be factored into the choice to
litter.
60
The Role of Command and Control
Policies




When the optimal level of pollution is zero or at zero,
direct controls make sense.
This is the case for extremely dangerous pollutants,
such as heavy metals and radioactive waste.
Damages associated with these pollutants are quite
severe.
Direct controls also make sense in other cases
where initial damages are quite high compared to
initial marginal abatement costs.

An example is CFC’s, where accumulated amounts
are dangerous but there are low cost alternatives.
61
The Role of Command and Control
Policies



Emergency situations may make direct
controls the preferable policy instrument.
These events occur in random and
unpredictable fashion.
Examples include smog alerts and droughts.
62
Pursuing Environmental Quality with
Economic Incentives

Economists advocate policies based on
economic incentives for two primary reasons:


Economic incentives minimize total abatement
costs by equating marginal abatement costs
across polluters and encouraging a broader array
of abatement options.
Economic incentives encourage more research
and development into abatement technologies
and alternatives to the activities that generate the
pollution.
63
Economic Incentives and Minimized Total
Abatement Costs






Consider the following graph.
A polluter is polluting at an unregulated level of 10
units.
The government imposes a tax equal to t dollars per
unit of pollution.
The polluter compares the tax of t dollars to the
marginal abatement cost (MAC) of reducing
pollution.
As long as the MAC is less than the tax, polluter will
reduce level of emissions.
Each polluter will chose an emission level which
equates MAC and the tax.
64
Economic Incentives and Minimized Total
Abatement Costs
65
Economic Incentives and the Certainty of
Attaining a Target Level of Pollution



If the aggregate marginal abatement cost
function is know, then achieving a targeted
level of pollution is easily accomplished.
If the aggregate marginal abatement cost
function is not known, the appropriate tax
level is much harder to determine.
Consider Figure 3.16, where evidence
suggests that the true MAC function lies
between an upper and lower bound set of
MAC’s.
66
Economic Incentives and the Certainty of
Attaining a Target Level of Pollution
67
Economic Incentives and the Certainty of
Attaining a Target Level of Pollution




Suppose policymakers believe MAC1b is the true
MAC. In an effort to achieve an emissions level of
E1, they impose a tax of t1.
However, if MACt describes how polluters will
respond, the emissions level will be E2.
E2 is higher than the desired level of pollution.
Because the choice of pollution abatement and
production technologies is sensitive to specific tax
structures, it may not be easy to change the tax to
achieve the desired level of pollution emissions.
68
Economic incentives and incentives for
research


If a firm is faced with a tax on its pollution, it
has the incentive to find ways to reduce its
pollution cheaply
The motivation that taxes provide for
technology development is an advantage of
taxes over command and control
69
Economic incentives and incentives for
Firm cost initially =
research for a firm
a + b + c (tax bill) +
d + e (abatement cost)
damages,
costs, $
Firm cost after R&D =
MAC initially
a (tax bill) +
b + e (abatement cost)
MAC
d
e
after R&D
t
c
b
a
Abatement
70
In summary:



Pollution taxes are preferable to command
and control techniques since pollution taxes
minimize abatement costs and provide
incentives for R&D
But, taxes do not put the level of pollution
under direct control so when there is
uncertainty in abatement costs one might not
get the desired level of pollution.
Marketable permits might achieve both…?
71
Marketable Pollution Permits



Marketable pollution permits are permits
which give a firm the right to emit a specific
number of units of pollution.
Polluters are free to buy and sell these rights
to pollute.
A marketable pollution permit system can
both minimize total abatement costs, provide
flexibility in the choice of mechanisms used to
meet pollution goals, and achieve the desired
level of pollution emissions.
72
Marketable Pollution Permits




A system of marketable pollution permits
begins with the determination of the target
level of pollution.
The next step is to allocate pollution across
polluters.
This allocation can be based on historic
pollution levels, auctions, a lottery, or some
other allocation scheme.
The buying and selling of pollution permits
will reallocate the emission rights.
73
Marketable Pollution Permits





Marketable pollution permits equate marginal
abatement costs across polluters.
Each polluter compares his/her marginal abatement
costs with the price of a permit.
If the marginal abatement costs are higher than the
price, they have an incentive to buy.
If the marginal abatement costs are lower, they have
an incentive to sell.
Buying and selling will continue until the equilibrium
price is reached which equates marginal abatement
costs across all firms.
74
Marketable Pollution Permits and
Geographic Considerations



Geographic location of emissions can have a
profound impact on the damages the
pollution generates for some categories of
pollution.
Central to the importance of location of
emissions is the manner in which the
pollution disperses when it enters the
environment.
Pollution controls must take into
consideration the geographic variation in the
effect of pollution on society.
75
Marketable Pollution Permits and
Geographic Considerations



A pollution control system based on taxes
could take variation into account by charging
higher taxes in areas where emissions are
more damaging.
A marketable pollution permit system must
divide the overall region into subregions.
These subregions can account for
geographic variability in one of two ways:
development of a receptor-based system or
development of separate markets for
subregions.
76
Marketable Permits and Geography:
Ambient-based Permit System




A receptor-based or ambient-based system allocates
pollution receptors across the subregion.
Locations relatively close to, and downwind from, the
polluter may require more permits.
Dispersion coefficients are used to help define the
terms of trade in this type of marketable pollution
permit market.
In the following figure, the location of a particular
polluter is denoted by a star and receptors are
designed by letters. This polluter may have to buy
some combination of 15 different types of permits.
77
Marketable Pollution Permits and
Geographic Considerations:
Ambient-based Permit System
78
Marketable Permits and Geography:
Emissions-based Permit System




An alternative to the ambient-based system is to
divide the subregions into separate markets.
Polluters need only purchase permits for the
subregion in which they are located.
The inability to trade across subregions may mean
that firms with lower abatement costs will not be
able to trade permits with higher abatement cost
firms in another subregion.
A compromise would be to have one type of permit
and allow trade across all regions, as long as the
trade does not result in ambient quality standards
being violated at any receptor point.
79
Marketable Pollution Permits and
Geographic Considerations:
Emissions-based Permit System
80
Other Types of Economic Incentives




Deposit-refund systems are a good way of
employing economic incentives when monitoring
costs are high.
This system is based on requiring a payment up
front for undesirable acts and then building in a
refund when a desirable action occurs.
The most common example of this is the depositrefund system in place for beverage containers.
This system has also been used for cars and
batteries in other countries.
81
Other Types of Economic Incentives




Bonding systems are closely related to depositrefund systems.
A bonding system requires a potential degrader of
the environment to place a large sum of money in
an escrow account.
This money is returned if the environment is
undamaged (or returned to its original condition) and
will be forfeit otherwise.
Bonds need to be large enough to provide an
incentive to use appropriate safeguards and/or
cover the cost of clean up if damage occurs.
82
Other Types of Economic Incentives



Liability systems are based on defining legal liability
for the damages caused by certain types of pollution
discharges and facilitating collection of these
damages.
The Comprehensive Environmental Response,
Compensation and Liability Act of 1980 (CERCLA)
defines legal rights to natural resources for local,
state and federal governments and defines how
damages can be recovered.
A related system defines legal liability and then
requires potential polluters to obtain full insurance
against any damages. There is a potential moral
hazard problem with this option.
83
Other Types of Economic Incentives



A system of pollution subsidies would pay each
polluter a fixed amount of money for each unit of
pollution reduced.
The polluter would reduce pollution to the point
where the subsidy is equal to the marginal cost of
abatement.
While the outcome is the same as a tax on polluters,
there are distributional effects, problems with
political acceptability and the possibility that
strategic behavior would lead to higher initial levels
of pollution in order to obtain the subsidy. In
addition, the subsidy could potentially attract more
polluters into the industry.
84
Conclusion




Market failures associated with environmental
externalities generate losses in welfare.
Command and control policies are the basis of
current policy but do not equate marginal abatement
costs across polluters.
Economic incentives, such as taxes or marketable
pollution permits do equate marginal abatement
costs.
While there are some problems with economic
incentives, they do create additional motivation for
technological innovation to reduce pollution.
85