Microeconomics 1

1
Microeconomics 1

• Mushtaq H. Khan
• Department of Economics

2
Market failures
Market failure is the common analytical framework
adopted by most economists to look at problems of
development and justifications of intervention
and collective action However, different
meanings are attached to this term and different
interpretations of critical market failures
affecting developing countries exist in the
literature Differences in interpretation can
lead to different explanations of the reasons for
economic backwardness in poor countries and imply
different areas for policy concern
3
What is a market?
Markets are systems of private contracting where
individuals voluntarily agree to contracts to
buy, sell or do specific things If these
decisions are voluntary, by definition market
contracts must make all individuals participating
in these contracts better off Of course the
definition of voluntary is problematic which
external constraints are considered legitimate
and which not. (The institutional framework
within which markets operate is the subject of
institutional economics) But if we accept
existing institutions, then a fundamental
proposition follows any contract that makes
society better off should be individually
negotiable because it should be possible to write
a contract that compensates all losers and still
leave some participants better off
4
The notion of market failure
Market failures are defined as situations where
private contracting fails to maximize net social
benefit Unfortunately, the benchmark exchange or
production alternative that would have maximized
net social benefit can be differently
conceptualized in different economic frameworks
We will look at some of the differences in
economic approaches to the conceptualization of
benchmarks that actual markets are failing to
achieve Secondly, the response to these market
failures depends greatly on the political and
perhaps philosophical assumptions of the
economist as to whether the market failures can
be corrected by intervention or not
5
The benchmarks what is the market supposed to
achieve?

• The neoclassical general equilibrium model
  perfectly competitive markets achieve Allocative
  Efficiencythe First Theorem of Welfare Economics
• A well-working efficient market with low
  transaction costs should achieve allocative
  efficiency, the failure to do so is a market
  failure due to different types of transaction
  costs. Externalities and public goods are
  examples of allocative efficiency market
  failures
• 2) However, inefficiency is not always just
  allocative Leibensteins X-Inefficiency.
  Non-competitive markets can result in firms not
  succeeding in minimizing costs.
• Here the problem is not about the allocation of
  resources but the extraction of effort from
  managers and workers
• The failure to achieve X-Efficiency is also a
  market failure because cost minimization is not
  just a zero sum game, but can in principle make
  society better off even after compensating the
  losers

6
The benchmarks what is the market supposed to
achieve?
3) Dynamic efficiency, Risk-taking, investment,
learning and innovation drive growth, and some
markets achieve these goals. Kaldors creative
function of markets based on Adam Smith and
Allyn Youngs analysis of cumulative causation.
Keynesian cumulative interaction between
investment and income. Dynamic aspects of
markets also central to Marx and Schumpeters
analysis of capitalist markets these markets can
promote innovation and productivity growth under
some circumstances Market failure in a dynamic
efficiency analysis is when markets fail to
achieve cumulative causation through innovation,
learning and productivity growth Here the
calculation of value lost due to market failure
is more difficult because gains are uncertain and
in the future, but dynamic failures are probably
the most important type of market failure from
the perspective of real economies, in particular
developing ones
7
The three notions of market failure are related
X-Efficiency and Dynamic efficiency are both
related to the allocation and use of resources in
particular ways so we could tautologically argue
that there is no difference between allocative
efficiency and other notions of
efficiency However, there are conceptual
differences between different notions of
efficiency
Allocative efficiency is the point P, and
typically neoclassical market failures refer to
production on other points on the production
function X-inefficiency refers to production at a
point such as A Dynamic efficiency refers to the
fastest rate of success in moving the production
function out to a point such as B
Source Arndt
8
Market failure as an organizing concept

• The response to market failures also depends on
  how we conceptualize the feasibility of the
  solution
• Intervention-Pessimist Market Supporters argue
  that an efficient market mechanism can be
  institutionally achieved, or that interventions
  will only make matters worse
• From this perspective, the reform agenda is to
  achieve the best possible market-promoting
  institutions that reduce transaction costs (the
  good governance agenda, the liberalization
  agenda, the strategy of right-sizing the state
  etc)
• ii) Market Pragmatists (for instance Stiglitz)
  argue that it is wrong to assume that there is a
  potentially efficient market out there or that
  inefficient markets cannot be improved upon.
• The concept of market failure can be used to
  identify specific problems that have specific
  solutions. But each solution requires
  interventions that create rents and so we are
  always going to be in a second best world where
  the costs of specific market failures are always
  being compared to costs of specific interventions
  to mitigate these failures

9
That is why economists use market failure in very
different ways
Differences between anti-interventionists and
pragmatists, and between different benchmarks can
result in important differences in policy focus.
More significantly, even if we are policy
pragmatists, prioritizing one benchmark may
apparently require sacrificing another For
instance, trying to enhance dynamic efficiency
and learning in developing countries through
conditional subsidies for growth sectors may
apparently worsen allocative efficiency Conversely
, enhancing allocative efficiency through
liberalization and subsidy withdrawal may lower
the capacity of learning either through
self-discovery (Hausmann and Rodrik) or through
the management of learning rents (Khan) Even more
important is that different explanations of the
sources of dynamic inefficiency are quite
different and have different policy implications
and identify different reasons for policy
failures in developing countries
10
Focus on detailed explanations and policy
implications
A pragmatic policy approach is to identify (from
a comparative analysis of countries) the main
constraints on development performance and the
types of market failures of different types
that may help to explain the persistence of these
problems The next step is to identify which of a
number of potential explanations is most likely
to be the primary cause of a persistent problem
there may be many possible mechanisms
theoretically responsible for a particular type
of market failure Each mechanism explaining a
market failure identifies different
implementation and capacity problems to resolve
it (and also identifies different explanations
for the differential performance of countries)
This analysis is particularly important because
intervention may not necessarily correct market
failure and in some cases can make the problem
worse government failure can be worse than
market failure so we need to carefully assess the
capacities of states to intervene to solve market
failures of particular types and build
appropriate capacities if necessary
11
Critical sources of market failures in poor
countries
Externalities are the most generic source of
market failure but the term is very broad and can
cover many different types of problems However, a
simple analysis of externalities is useful
because it points to a number of common problems
externalities are often difficult to solve
because of differences in the interpretation of
their severity due to different incomes and
preferences of different players, and they are
particularly difficult to solve because there are
usually multiple solutions which imply different
costs and benefits for different players Another
bundle of market failures very relevant for
developing countries are due to the absence of
risk-sharing institutions, imperfect information
and weak credibility and enforcement capacities
of states these prevent the solution of critical
learning problems for a different set of reasons
compared to a simple analysis of externalities
12
Externalities
An externality exists whenever the utility
of a consumer or the production possibility of a
producer is directly affected by the actions of
another agent in the economy.
By definition, externalities exist because
of missing markets, but missing markets in turn
exist because of transaction costs of transacting
in particular markets, or missing property rights
which are typically missing because of high
transaction costs of enforcement
Transaction costs are the costs of
discovering trade opportunities, reaching
agreements on price, monitoring the transactions
to prevent free riding and fraud, and if
necessarily enforcing contracts if default is
detected
Externalities can be consumption externalities
when they affect utility functions or production
externalities when they affect production
functions. They can be bilateral if they affect
two people or multilateral if they affect many.
They may be positive if they increase utility or
production, negative otherwise.
13
Externalities 2
A technological externality exists where a
utility or production function has the form
(shown in the case of a utility function) Ui
Ui (x1i, x2i,..... xni, h), where xni is the
level of each xn consumed by i and h is an
activity whose level is directly chosen by
another agent. For instance, h could be the level
of smoke in the air agent i breathes.
The derived utility function is more useful for
our purposes and is given by Vi Vi (p, wi, h)
since each xni in the utility function is a
function of p and wi, the price vector and the
initial endowments of agent i
Most externality analysis is based on an
important and implicit assumption that of zero
wealth effects. We will first follow this
tradition and then examine the consequences of
dropping this assumption later.
14
Externalities 3
If there are no wealth effects, the derived
utility function can be written as Vi (p, wi, h)
?i (p, h) wi. This means that greater or
lesser wealth has an effect on utility but no
effect on the precise quantities of each good
consumed.
If utility functions were in fact like this, the
optimal level of any activity, including
externality-generating activities, would not
depend on the wealth of the participants.
15
The Marginal cost and benefit of externalities
The marginal benefit of accepting the external
effect is the savings in net social value by not
removing the effect so one way of thinking about
the MB is as the marginal cost of abatement The
marginal cost of the external effect is more
straightforward it is the direct marginal cost
of not doing anything
16
The Efficient Level of Externalities
In the absence of any collective action to solve
the externality, the emittor maximizes his
personal utility and produces h where ?1'(h)
0 (MBzero). Efficiency and social optimality
requires the maximization of social (joint)
utility which requires output of ho where ?1'(h)
?2' (h) 0 (MBMC).
17
The Efficient Level of Externalities 2
By moving from h to ho, the loss for the emittor
(agent or group 1) is A, the gain for the
recipient (agent or group 2) is AB. Societys
gain is AB - A B.
18
Accepting wealth effects means there is no unique
efficient level of external effects
What is the efficient level of emissions? With
wealth effects, the answer depends on who is the
emittor and who is the recipient.
19
Policy Responses to Externalities Regulation
The simplest solution would be to regulate the
level of activity to ho. The problem is that the
functions ? for each agent would have to be
precisely known by the state to calculate this
level of h.
This is because at the optimal level of
externality production, each firm has to face
exactly the same marginal cost it must cost the
same to get rid of a unit of pollution for every
firm.
If firm A had a lower marginal cost of reducing
pollution (its marginal benefit from pollution
was lower) compared to firm B, then society will
be better off by allowing firm B to produce more
pollution and getting firm A to compensate by
reducing pollution
In other words, a situation where the marginal
costs of removing pollution (the marginal benefit
of continuing to pollute) are unequal cannot be
an optimal equilibrium
This is the problem for the regulator we have to
know each firms MB curve to determine the
optimal amount pollution target to set for each
20
Policy Responses to Externalities Regulation
MC/MB
15
Collective Output at each level Of MB
MB of Firm A
MB of Firm B
Pollution
Total
As pollution
Bs
In this diagram, if each firm had to pay 15 to
for each unit of pollution produced, firm A would
produce A, firm B would produce more pollution at
B (because it has a higher abatement cost/older
technology), and the total pollution produced at
this marginal cost would be the horizontal sum,
shown as Total pollution
The horizontal sum of the two MB curves shows the
total output of pollution when each firm faces
any given price at the margin for the pollution
it produces
21
Policy Responses to Externalities Regulation
MC/MB
Social Marginal Cost
Firm A
Collective Output at each level Of MB
Firm B
Pollution
As Regulatory Target
Social Optimum
Bs Regulatory Target
The collective output line allows us to determine
the optimal social amount of pollution when we
can identify the social marginal cost curve and
the distribution of this between the two firms
can be read off along the horizontal line from
that equilibrium
22
Policy Responses to Externalities Regulation
MC/MB
Social Marginal Cost
Firm A
b
a
Collective Output at each level Of MB
Firm B
Pollution
Average Regulatory Target
If instead the regulator divided the total
pollution at the social optimum equally between
the two firms, the result would not be a social
optimum
Firm A would be allowed to produce too much
pollution given its marginal benefit, with a
social loss shown by triangle a, and Firm B would
be forced to reduce pollution so much that its
marginal benefit would be higher than the social
marginal cost, with a loss shown by triangle b
23
Transaction Costs Associated with Regulation
i) Costs of collecting firm level information
about production possibilities and costs of
abatement.
ii) Costs of monitoring accurately the emissions
at the firm level. Note that government failure
which allows corrupt officials to be bribed would
significantly increase the transaction costs of
collecting accurate information.
iii) Costs of enforcement when rules are broken.
This includes legal costs and bureaucracies to
collect the fines etc.
iv) If the regulatory solution is to be made
acceptable, firms who are regulated should be
compensated to the extent of their loss. But
organizing this compensation also raises
transaction costs of collecting taxes and
distributing the subsidy.
24
Policy Responses to Externalities Taxes and
Subsidies
If the state knew the optimal level of ho, it
would be able to induce this by imposing a tax on
the emitter of th - ?2'(ho), the marginal cost
to the recipient at the optimal level ho.
The loss for 1 ATax ACD. The gain for 2
ABE. The gain for G (government) is the Tax
collected which is CD. The net social benefit
gained (ABE)(CD)-(ACD)BE as before.
This holds even if those taxed are otherwise
compensated to the extent of ACD.
25
Policy Responses to Externalities Taxes and
Subsidies 2
Paradoxically, a subsidy to the producer, 1, at
the same per unit level, not to produce beyond ho
will theoretically have the same effect. A tax
creates a disincentive to produce more than ho. A
subsidy creates an incentive to cut production
from h to ho.
In the case of the subsidy, the gain for 1
Subsidy - Lost production AB-AB. Gain
for 2 ABE. Loss for G AB (the
subsidy). Net social benefit gained B (ABE)
- (AB) BE.
26
Transaction costs of Tax-Subsidy Solutions
i) Collecting information about marginal costs
and benefits in the sector. Here firm level
information is not required so the information
requirement is less than in the case of
regulation.
ii) Monitoring the amount of emission in the case
of a tax and the amount of non-emission in the
case of a subsidy. The second is more difficult
to observe since each producer has the incentive
to lie and exaggerate how much he would have
produced in the absence of a subsidy. On the
other hand, if the emittor is poorer, enforcing a
tax on the emittor is likely to be politically
very costly, particularly in international
negotiations.
iii) Administering the collection of the tax or
delivery of the subsidy. This includes
bureaucratic costs.
iv) In the case of the tax, in the long run
compliance may require the emittor to be
compensated to the extent of ACD. This has to
be organized through subsidies unrelated to
emission control and involve further transaction
costs for the government.
27
Policy Responses to Externalities Enforceable
Rights
An alternative solution which saves on the
information requirement of the state was proposed
by Coase. This is based on creating the
conditions for individual bargaining between
emitters and recipients. The problem here is of
potentially very high transaction costs of
private exchange.
The theory is that if enforceable rights can be
created on the external effects (essentially
enforceable liability for damage in the case of
negative externalities), then provided
transaction costs of private bargaining are low,
private bargaining between individuals will
result in the optimal output of the activity,
irrespective of the initial allocation of
liability and the bargaining power of the
agents. This result is also known as Coase
Theorem.
28
Policy Responses to Externalities Enforceable
Rights 2
1) Suppose the right to pollute is given to the
polluter. We then start off at h. There are two
possibilities in terms of bargaining power (e
represents a very small number)
a) Polluter is more powerful and can demand
compensation of ab-e, making a net gain of (ab-
e)-a b- e The Recipient makes a net gain of ab
- (ab- e) e Social gain b- e e b
b) Recipient is more powerful and can offer
compensation of a e, making a net gain of
(ab)-(ae) b- e The Polluter makes a net gain
of a e - a e Social gain b- e e b (the
same as before)
29
Policy Responses to Externalities Enforceable
Rights 3
2) Suppose the right to pollute is given to the
recipient. We then start off at O. There are two
possibilities in terms of bargaining power
a) Polluter is more powerful and can offer
compensation of c e, making a net gain of cd-
(c e) d- e The Recipient makes a net gain of
(c e) - c e Social gain d- e e d
b) Recipient is more powerful and can demand
compensation of cd- e, making a net gain of
(cd- e) - c d- e The Polluter makes a net gain
of (cd) - (cd- e) e Social gain d- e e
d (the same as before)
30
Coase Theorem in the Presence of Wealth Effects
If rights are initially given to the polluter and
the starting point is h, wealth effects can mean
that the bargaining solution leads to h rather
than ho.
If rights are initially given to the recipient
and the starting point is O, wealth effects can
mean that the bargaining solution leads to h
rather than ho.
31
Coase Theorem in the Presence of Transaction Costs
Coase is often misrepresented as saying that
transaction costs are low in fact he was saying
that transaction costs are typically not low, and
therefore the allocation of rights does matter
Case I Rights given to Polluter and as a result,
assume TC has to be paid by recipients. Their
willingness to pay declines and bargaining leads
to h instead of ho.
Case II Rights given to Recipient and as a
result, assume TC has to be paid by polluters.
Their willingness to pay declines and bargaining
leads to h instead of ho.
32
Coase Theorem in the Presence of Transaction Costs
The transaction costs relevant here are the costs
of either recipients of the externality, or the
emittors or both to organize and enforce the
transaction that would reduce the externality to
the optimal level
For instance, for recipients, these include i)
the costs of discovering other potential
recipients, ii) organizing collective action with
them by agreeing the contributions that each will
make, iii) discovering the emittors and the
individual or collective payoffs that will induce
them to reduce their emission, iv) monitoring
subsequent emissions to ensure that contracts
have been adhered to, v) enforcing contract
through courts or other means when contract
violation happens
Clearly these costs can be very high and will
also be very high if emittors had to discover and
pay off recipients. However, the aggregate costs
of emittors and recipients can be very different
33
Coase Theorem in the Presence of Transaction Costs
In this case, the presence of transaction costs
and their differential impact means that who gets
the initial rights DOES matter, not just for
distribution, but also for determining the
optimal amount of pollution
If the allocation of property rights can
determine the optimal level of pollution, the
policy solution will be driven by the political
preferences of the policy-makers if
policy-makers have a strong dislike for this
externality they can achieve their outcome by
choosing a particular policy approach
Actually this makes policy towards externalities
very difficult to sell because many
constituencies will always rightly think that the
policy has been designed against their interests
However, transaction costs have a different and
more practical implication for policy if the
transaction costs associated with the property
rights (or any other) approach are too high,
regulatory or other direct interventions may be
better (provided their transaction costs are
lower)
34
Implications of Transaction Costs
The solution of any external problem has a
potential gain in net social benefit (but we have
seen that this net social benefit can depend on
the transaction costs involved in solving the
externality problem)
In principle, the transaction cost associated
with some solutions may be so great that it is
not feasible to solve the externality there will
be no net social benefit to capture
MC
MC/MB
Willingness To pay after Transaction Costs
Potential Net Social Benefit
MB
External effect
35
Why Do Externalities Exist?
If the transaction costs of solving an
externality are too large, it is clearly more
efficient to let the externality persist! Perhaps
this is why many externalities exist in the first
place
So if an externality persists there are two
immediate possibilities, neither of which are
very satisfactory a) The gain in net social
benefit is more than the potential transaction
cost involved in solving the problem, ?NSBgtTC,
but a failure of knowledge on the part of
individuals or the state prevents a solution.
b) The gain in net social benefit is less than
the potential transaction cost, ?NSBltTC, so the
externalities are not Pareto relevant.
Neither is satisfactory in terms of being able to
explain the vast range of externalities that are
perceived to exist. If these are the only two
possibilities, either most of these externalities
can be rapidly solved once economists have
pointed out the possibility, or they are
irrelevant since they cannot actually be solved.
36
Why Do Externalities Exist? 2
A third possibility that is often not recognized
is that the problem may be fundamentally
political a) There may be differences in the
valuation of costs and benefits because of
different income levels of the groups involved,
and the rich or the poor may not accept the
willingness to pay criteria based on existing
incomes as a true measure of social cost or
benefit.
b) The distribution of costs and benefits may be
different under different solutions and different
groups may hold out till their preferred
solutions are adopted. In both cases solutions
may be difficult to find even though the
externality was Pareto relevant and there was
no knowledge failure.
37
A further caveat the non-convexity problem
The marginal approach to optimization can also be
misleading in some circumstances where the
effects of an externality can be non-convex
Non-convexity is particularly a problem where the
recipients of an externally can take evasive
action, that is the problem is not purely a
technical one (though economies of scale can also
cause technical non-convexities)
In this diagram the recipients total profits
decline with emission but after a point the firm
starts to cut back its own production or moving
activities elsewhere so that its profits decline
at a slower rate
38
A further caveat the non-convexity problem
Looking at the marginal cost of the recipient in
this case shows that marginal costs initially
increase as in the usual way, but as the rate of
decline of total profits slows down, the marginal
cost actually starts to fall at the point A, the
point of inflexion in the previous diagram
Clearly in this case we have to be careful
equating marginal costs and benefits can give us
the global maximum C, but it could also give us
the global minimum D. Compared to that, the point
E where the externality is not tackled at all
gives a higher net social benefit (in this case
profit)