The Making of Economic Policy: A Transaction-Cost Politics Perspective

1
The Making of Economic Policy A Transaction-Cost
Politics Perspective

• Avinash K. Dixit
• MIT Press, 1996

2
Motivation

• The reality of most countries trade policies is
  so blatantly contrary to all the normative
  prescriptions of the economist that there is no
  way to understand it except by delving into the
  politics.

3
Transaction-Costs

• In economics, it has come to mean a very general
  class of information, negotiation, and
  enforcement problems that affect the internal
  organization of firms and the outcome of market
  and nonmarket relations among firms, workers and
  so on.
• A similar and even more severe class of
  transaction costs (which deals with
  principal-agent problems, commitment and
  credibility) pervades political relations and
  affect political outcomes.
• The policy process can be better understood and
  related to each other by thinking of them as the
  result of various transaction costs and of the
  strategies of the participants to cope with these
  costs.

4
Common Agency

• Within this general framework, a particularly
  important feature of the political process of
  making economic policy is the common agency,
  where several players in the political game try
  to influence the actions of one decision-maker.
• It leads to a severe diminution of the power of
  the incentives that can be offered to this
  decision-maker.

5
Market versus Government

• The tradition dichotomy of market versus
  government, and the question of which system
  perform better, largely lose its relevance.
• Both of them are facts of the imperfect economic
  life, and they unavoidable interact in complex
  ways.
• The most we can do is to understand how the
  combined economic-political system evolves
  mechanisms to cope with the variety of
  transaction costs it must face that precludes a
  fully ideal outcome.

6
Principal-Agent Model
7
The players in the Regulatory Process and their
inter-relations
Regulatory agency
8
The players in the Regulatory Process and their
inter-relations
Firms and Investors
Regulatory agencies
9
The players in the Regulatory Process and their
inter-relations
Firms and Investors
Regulatory agency
Consumers
10
The players in the Regulatory Process and their
inter-relations
Firms and investors
Regulatory agency
Executive
Consumers
Congress and Committees
11
The players in the Regulatory Process and their
inter-relations
Firms and Investors
Regulatory agency
Executive
Consumers
Congress and Committees
12
The players in the Regulatory Process and their
inter-relations
Firms and investors
Regulatory agency
Executive
Voters
Consumers
Congress and Committees
13
The players in the Regulatory Process and their
inter-relations
Bureaucracy and other agencies
Firms and Investors
Regulatory agencies
Executive
Voters
Congress and Committees
Consumers
14
The players in the Regulatory Process and their
inter-relations
Bureaucracy and other agencies
Courts
Firms and Investors
Regulatory Agency
Executive
Voters
Consumers
Congress and Committees
15
The players in the Regulatory Process and their
inter-relations
Bureaucracy and other agencies
Courts
Supreme Court
Firms and Investors
Regulatory agency
Executive
Voters
Consumers
Congress and Committees
16
Set of Principal Agent Models
Supreme Court
Court
Bureaucracy and other agencies
Regulatory Agencies
Executive
Firms and Investors
Voters
Consumers
Interest Groups
Congress and Committees
17
Principal-Agent model

• It is relevant to analyze relationships with
  following conditions
• Delegation
• Asymmetric information
• Imperfect relation between the effort and its
  results
• High costs of monitoring
• No alignment of preferences or objectives

18
Principal-Agent model

• In general terms, there exist two solutions to
  the principal-agent relation
• Involves a structure of remuneration aiming at
  approaching the incentives of both parts involved
  (paying a tip).
• Rules and institutions capable of avoiding
  opportunistic behavior of the agent.

19
Positive versus Normative Political Theory

• To what extent those theories are excluding,
  complementary, or competitors?
• Both are based on the premise that agents are
  rational and look for the own interests.
• Dixit suggests a synthesis, labeled
  transaction-cost politics that views
  policymaking as a process in real time.

20
Normative Political Theory

• Market failure (i.e. natural monopoly)
• Requires a criterion to fix with the best way
  this failure aiming at maximizing the social
  welfare.
• Pareto-optimality (efficiency criterion)
• First-best and second-best solutions (overcome
  the market imperfections)
• Moral hazard (level of effort of the firm)
• Adverse selection (the regulator has no complete
  information about the costs of the firm)
• The regulator is obliged to pay informational
  rents.

21
Normative Political Theory

• As a consequence of asymmetric information
  Laffont and Tirole offer as a solution a menu of
  contract
• Price-caps
• Repartition of profits
• Thus, the firm would have incentives to revel its
  true effort

22
Normative Political Theory

• The normative solution is rarely observed in the
  real life
• The normative solution requires a high level of
  discretion of the regulator. This generates
  incentives for opportunistic behavior
• Dixit (1996) observes that the normative theory
  understand the formulation and implementation of
  policies as a technical or organizational
  problems. As if political failures could be
  avoided through good management, namely, giving
  power to make and implement economic policy to an
  economist
• In other words, it does not take into account
  political and economic institutions.
• The benevolent, omnipotent, and omniscient
  dictator would maximize the social welfare.

23
Positive and Normative Political theory

• Dixit 1996
• Dictator benevolent, omnipotent and omniscient

second-best literature
24
Positive and Normative Political theory

• Dixit 1996
• Dictator benevolent, omnipotent and omniscient

second-best literature
Informational Economics literature
25
Positive and Normative Political theory

• Dixit 1996
• Dictator benevolent, omnipotent and omniscient

second-best literature
Informational Economics literature
PPTR tries to relax this unrealistic premises
26
Positive and Normative Political theory

• Dixit 1996
• Dictator benevolent, omnipotent and omniscient

second-best literature
Informational Economics literature
Economic relations Usually involves Multiple
principals
27
Positive and Normative Political theory

• Dixit 1996
• Dictator benevolent, omnipotent and omniscient

second-best literature
Informational Economics literature
Economic agents, Including regulatory
agencies, Bureaucracy, and government are made by
people that try to Maximize their own interests
28
Positive Theory

• Also starts from a market failure
• Regulation necessarily leads to a redistribution.
  Rarely there are Pareto-optimum corrections or
  ways of implementing compensations
  (side-payments)
• Given the behavioral premises, individuals and
  groups would try to influence this
  redistribution.
• Their capacity to do so would depend on the
  institutions (i.e. property rights)

29
Positive Theory

• Many situations apparently inefficient could be
  understood as a consequence of restrictions
  imposed by transaction costs among agents. The
  great majority of the PPTR tries to explain why
  such inefficient situations are observed even
  when there are obvious and better ways to deal
  with that. This involves to identify the source
  of restriction which are generating transaction
  costs and to show how it affects the agent
  choices. Almost always it requires to take into
  account the political institutions.

30
Why does the regulation (government intervention)
tend to be inefficient?

• Economic reasons
• Asymmetric information
• Uncertainty about effects, costs, and benefits
• It does not mean that the regulation is not
  necessary however, those problems should be
  taken into account

31
Why does the regulation (government intervention)
tend to be inefficient?

• Political reasons
• Regulation necessarily means income
  redistribution
• Quotas, licenses, subsides, establish a price,
  etc. transfer rents and incomes
• Interest-groups will demand this redistributions
  and politicians offer
• Generally, the regulation tends to be inefficient
• Rent-seeking
• Few beneficiaries and lots of opponents

32
A Synthesis A Policy Process in Real Time

• Constitutions are incomplete contracts
• The constitution never lays down the clear, firm,
  and comprehensive set of rules that the
  contractarian approach depicts so there is room
  for maneuver in individual acts.
• Inability to foresee all the possible
  contingencies and to adjust to them
• Last longer than most business relationships
• Constitutions are not made behind a veil of
  ignorance

33
Coases Theorem

• Once the property rights over a disputed resource
  (ex. Rents of a monopoly) have been established,
  and given that the transaction costs are equal to
  zero, the private negotiations between agents
  approaches to the efficient level of allocation.
  (Coase, Ronald, (1960) The problem of social
  cost. Journal of Law and Economics, 3, 1-44).

34
Example

• A firm pollutes a river which provides a cost of
  500 to a farmer along the river.

35
Example

• A firm pollutes a river which provides a cost of
  500 to a farmer along the river.
• It costs 300 to the farmer to build a station to
  purify the water.

36
Example

• A firm pollutes a river which provides a cost of
  500 to a farmer along the river.
• If costs 300 to the farmer to build a station to
  purify the water.
• The firm can eliminate the harmful effect of
  pollution changing its production process at the
  cost of 100.

37
Questions

• Should be allowed the firm to pollute the river?
• Should the firm be obliged to change its
  production process?
• Who should pay for that?
• What are the economic implications if the firm
  would have the property rights of polluting and
  the farmer would have to compensate him in order
  to not pollute?

38
Example

• And, if the property rights belong to the farmer?

39
The Coases Theorem

• Coase argues that, for the economic efficiency
  point of view, the result would be the same
  regardless the of the property rights given that
  the transaction costs are null.

40
Numeric Example I

• Cost of pollution over the farmer 500
• Cost of cleaning from the farmer 300
• Cost of cleaning from the firm 100
• Case I
• The farmer has the property rights.
• The firm will pay for the right to pollute, It
  pays 100 at maximum
• The farmer would accept more than 300 only since
  he/she would have to clean the water in order to
  avoid the cost of 500
• Thus, the is no agreement and the firm purifies
  the water itself.

41
Numeric Example I

• Cost of pollution over the farmer 500
• Cost of cleaning from the farmer 300
• Cost of cleaning from the firm 100
• Case II
• The firm has the property rights.
• The farmer will pay for the firm to not pollute.
  It pays 300 at maximum
• The farmer would accept to pay no more than 100
  since he/she could clean the water at this price.
• Thus, the farmer pays between 100 and 300 and
  the firm purifies the water itself.

42
The solution is the same in both cases

• The solution is economically efficient

43
Efficiency vs. Equity

• Although the reached solution are not affected by
  the property rights the distribution between the
  parts involved would be.

44
Efficiency vs. Equity

• When the property rights belong to the farmer,
  the firm has a cost of 100.

45
Efficiency vs. Equity

• However, when the property rights belong to the
  firm, the farmer pays at least 100 and part of
  the exceeding of 200. This rent of 200 will
  be allocated by negotiations between the parts.

46
What is the fairest solution?

• Coase recommends that we should not conclude too
  fast.
• He calls our attention to the reciprocal nature
  of the problem.

47
Graphic Example.
MB for the firm to pollute
MC of the pollution over the farmer
c
a
d
b
?3
?
?1
?2
48
If the firm has PR it pollutes until ?3. The
farmer can convince the firm to pollute until ?2
offering more of d. The farmer would thus win
cd.
MB for the firm to pollute
MC of the pollution over the farmer
c
a
d
b
?3
?
?1
?2
49
If the farmer has the PR he/she prefers ?1 (if
there is no negotiation). The firm can compensate
the farmer to pollute at ?2 offering more of
b.
MB for the firm to pollute
MC of the pollution over the farmer
c
a
d
b
?3
?
?1
?2
50
Necessary conditions to reach the negotiated
result

• Well defined property rights. It implies
  political and judicial institutions that work
  well.
• Low transaction costs
• Information about all costs of each possible
  result for each part involved.

51
Polemic Implications of the Coases Theorem

• In many cases there are no necessity of
  governmental interventions in order to solve
  problems of externalities since the private
  negotiations reach the economically efficient
  solution.

52
Limitations of the Coases Theorem

• The transactional costs usually are not small,
  especially when several players are involved.
• There exist public and private externalities. In
  the public ones there are the free-ridding
  problem.
• There exist the problem of inter-temporal
  transactions with future generations.
• The initial distribution of the property rights
  can affect the capacity of negotiation of each
  part.
• Even when the theorem fits there are room for
  government role.

53
Role of the government is to help greasing the
private negotiation

• Property rights have to be well defined and
  allocated clearly.
• The government has to monitor the environment
  aiming at identifying who are the polluters and
  who are victims of the pollution. This
  information has to be publicized for all parts
  that have been affected.
• The government must establish, quantify, and
  publicize the legal rules that regulates the
  environment in order to make it easier the
  private negotiations.
• Access to the judicial system has to be
  facilitated and cheap

54
Regulation and the Coases Theorem

• As in practical terms those conditions are rarely
  found, the coases solution is not viable (this
  is exactly what Coase claims!)
• Therefore, the regulation becomes necessary.
• The transactions costs are the key. Whose are the
  political property rights? What are the political
  transaction costs?
• The results usually are not efficient

55
Regulation and the Coasee Theorem

• Reasons why it is difficult to make side
  payments
• Informational problems, uncertainty, difficulties
  of measuring costs and benefits
• Collective action problems
• Inexistence of a market of political property
  rights (political risks)

56
Delegating Powers

• A Transaction Cost Politics Approach to Policy
  making under Separate Powers

David Epstein Sharyn OHalloran
57
(No Transcript)
58
Puzzle

• Why does Congress delegate broad authority to the
  executive in some policy areas but not in others?
• Can Congress perfectly control delegated
  authority through administrative procedures,
  oversight, and administrative law?

59
Question

• When does Congress specify the details of policy
  in legislation, and when does it leave these
  details to regulatory agencies?
• What are the implications of this institutional
  choice for policy outcomes?
• Basic question of policy making in a
  separation of powers system.

60
Transaction Cost Politics

• Policy can be made (i) in Congress, (ii) through
  delegation authority to the executive agencies,
  or a (iii) mixture of these two.
• When deciding where policy will be made, Congress
  trades off the internal policy production costs
  of the committee system against the external
  costs of delegation
• (make-or-buy decision).

61
Theory

• A policy will be made in the politically most
  efficient manner that is, maximizing
  legislators political goal reelection.
• Legislators will prefer to make policy themselves
  as long as the political benefits they derive
  from doing so outweighs the political costs
  otherwise, they will delegate to the executive

62
Abdication versus delegation?

• Implications
• To what extent can Congress control unelected
  regulatory agencies when delegating authority?
• Lowi (1969)
• Congress has surrendered their legislative
  authority to bureaucrats
• MCNOLGAST (1987, 1989)
• Legislators can control bureaucrats through
  oversight, administrative procedures and
  enfranchising third parties into the
  decision-making process.

63
Authors Claim

• A theory of oversight divorced from a theory of
  delegation will overlook the fact that those
  issue areas in which the executive makes policy
  differ from those in which Congress makes policy
  itself.

64
Legislative organization and delegation

• Distributive versus informational theories
• This is a false dichotomy (whether committee
  serves pork barrel projects or legislators
  desire to avoid the consequences of ill-informed
  policies)
• Some policy areas are more characterized by
  informational concerns and other by distributive.
  The first ones tend to be delegated and the
  second tend to be made in Congress

65
Divided Government and Delegation

• To what extend will an opposing Congress delegate
  more powers to the executive?
• Mayhew (1991) the same number of legislation get
  passed under unified and divided government. It
  doesn't lead to gridlock.
• However, the authors call our attention to the
  quality of bills passed
• Congress delegate less and constrains more under
  divided government procedural gridlock, that is,
  producing executive branch agencies with less
  authority to make policy and increase oversight
  from congressional committees.

66
What does Transaction Cost mean?

• Anything that makes reality deviate from this
  Coasian world
• anything that impedes the specification,
  monitoring, or enforcement of an economic
  transaction is a transaction cost (Dixit 1966).

67
Common elements of the Transaction Costs analysis

• The contract is the unit of analysis
• The contracts are enforceable by some neutral
  third-party
• Assumes the existence of multiple governance
  structures
• Economic actors are assumed to be boundedly
  rational (agents cannot eliminate transaction
  costs by simply contracting)
• Holdup problems
• Contracts are incomplete and ambiguous

68
Theory of Transaction Cost politics

• Delegation to a bureaucracy is subjected to the
  political equivalent of a holdup problem so
  Congress will delegate to the executive when the
  external transaction costs of doing so are less
  than the internal transaction costs of making
  policy through the normal legislative process of
  the committees.

69
Reasons not to delegate

• With transaction costs equal to zero Congress
  might like to delegate broad discretionary
  authorities and agencies might like to receive
  such authority (please everyone). However,
• Difficult to control and monitoring
• Incomplete contracts
• Uncertainty of future opportunistic behavior
• Impossibility of binding the actions of agency
  successors

70
Reasons to delegate

• Save time/Reduce congress workload
• Take advantage of agency expertise
• Protect special interests
• Blame-shifting (avoiding the inefficiencies of
  committee system)
• Delays, logrolling, informational inefficiencies,
  etc.
• These do not explain why delegate in some
  policy areas rather than others.

71
Neither alternative is perfect

• However, in some specific cases delegate or not
  (how mach delegate) would be relatively more
  attractive than from the median legislators
  preference.

72
Basic Assumptions

• Two alternative modes of policy making
• Congress
• Committee System
• Regulatory Agencies
• Delegation to Executive
• Legislators decide where policy is made.
• Legislators primary goal is reelection
• Authority will be allocated across the branches
  in a politically efficient manner

73
Division of Labor

• Each alternative mode of policy making has its
  own set of inefficiencies
• Legislative Policy Making (Committees)
• Logrolling, delay, informational problems
• Agency Policy Making (Delegation)
• Principal-agent problems of oversight
• So when deciding where policy will be made,
  legislators trade off these internal and external
  inefficiencies.

74
Make-or-Buy

• Congresss decision to delegate is like a firms
  make-or-buy decision.
• Legislators can either produce policy internally,
  
• OR
• Legislators can subcontract out (delegate) to the
  executive.

75
Discretion Continuum
D
Legislative Policymaking
Agency Policymaking
High Discretion
Low Discretion
76
Discretion Continuum
D
Legislative Policymaking
Agency Policymaking
High Discretion
Low Discretion
Trade Off
77
Discretion Continuum
D
Legislative Policymaking
Agency Policymaking
High Discretion
Low Discretion
Trade Off

• Amount of discretion delegated to the executive
  balances these costs at the margin.

78
The formal decision to delegate

1. The legislation (bill) is the unit of analysis
2. There exist enforceable mechanisms for
   inter-branch contracts
3. Multiple possible governance structures
4. Bounded-rationality
5. Holdup problem
6. Internal transaction cost of production
7. A decision mechanisms predicting which governance
   structure will be chosen

79
Model Elements

• Actors Floor, Comm., Pres., Agency
• Policy space X ?1, status quo x00
• Preferences (x-xi)2 xF0, xPgt0
• Implies that all actors will be risk averse
• Outcomes x p ? , ? U-R,R
• Agencies have perfect information of ?
• Committees only know sign of ?
• Floor knows ex ante distribution of ?

80
Model

• Players
• Median floor voter in Congress (F)
• Congressional Committee (C)
• President (P)
• Agency (A)
• One-dimension policy single-pick preferences,
  median voter (pivotal)
• Each player has a most-preferable policy or ideal
  point and are risk-averse (dislike uncertainty)
• Policy outcomes are not necessarily the same as
  the policies that emerged from the policymaking
  process.
• That is, X p ?

81
Model

• The players start the game with probabilistic
  beliefs about the value of ?.
• These beliefs are uniform in some interval -R,
  R.
• During the game, players will try to gather more
  information (expertise) about the exact value of ?

82
Game Tree
F
l
o
o
r

s
e
t
s

p
o
l
i
c
y

p
F
x

p

?
F
F
D
o
n
'
t

D
e
l
e
g
a
t
e
C
o
m
m
i
t
t
e
e

l
e
a
r
n
s

?
gt
0

o
r

?
lt
0
,

?
r
e
p
o
r
t
s

b
i
l
l

(
b
)
N
F
C
D
e
l
e
g
a
t
e
x

S
Q

p

?
A
F
A
g
e
n
c
y

l
e
a
r
n
s

F
l
o
o
r

s
e
t
s

P
r
e
s
i
d
e
n
t

s
e
t
s

e
x
a
c
t

v
a
l
u
e

o
f

?
?


d
i
s
c
r
e
t
i
o
n

(
d
)
,


A
g
e
n
c
y

i
d
e
a
l

s
t
a
t
u
s

q
u
o

(
S
Q
)
p
o
i
n
t

x

s
e
t
s

p
o
l
i
c
y

p
A
s
.
t
.

A


F
i
n
a
l

I
n
s
t
i
t
u
t
i
o
n
a
l

C
h
o
i
c
e
P
o
l
i
c
y

M
a
k
i
n
g

P
r
o
c
e
s
s
O
u
t
c
o
m
e
s
83
Sequential Signaling game

• Institutional choice
• While the agency observes the exact value of ?,
  the committee observes only whether ? lies in the
  interval range -R, 0 or the range 0. R that
  is, the committee sees only the sign of ?.
• Policy-making Process
• Choice of buying or making the policy.
• Final outcome
• Committee X p ?
• Agency X SQ pa ?

84
Solving the Game

• Agency sets policy as close to its ideal point as
  possible, given limits on discretion.
• President sets Agencys ideal point to equal her
  own xAxP.
• Congress sets status quo and discretion
• Set SQ to get its ideal point in expectation,
  given beliefs about ?
• Set discretion dR-xP.

85
Theoretical Propositions

• In equilibrium, Congress will delegate more
  discretionary authority to the executive
• 1. The lower the level of congressional-
  executive conflict
• 2. The higher the level of committee-floor
  conflict
• 3. The higher the level of uncertainty in the
  policy environment.

86
Deliberate Discretion?

• The Institutional Foundation of Bureaucratic
  Autonomy

John Huber Charles Shipan
87
What this book is about?

• How elected officials use the statutes to
  establish policy details in effort to achieve
  desired outcomes.
• Develop and test a theory of delegation that
  explains the choice of how to delegate aiming at
  understanding how institutional context affects
  delegation strategies.

88
Two types of designing statutes

1. To write long statutes with extremely detailed
   languages in a effort to micromanage the
   policymaking process.
2. To write vague statutes that have many details
   unspecified, thereby delegating policymaking
   authority to other actors, usually bureaucrats.

89
Vague (Germany) versus specific (Ireland)
statutes about sexual harassment

• Why has the same policy issue been treated
  differently by politicians in distinct political
  systems?
• Why these choice matter?
• Is the transferred discretion a deliberate
  choice?

90
Delegation as double-edge sword

• the central reason for granting policymaking
  autonomy to bureaucrats their technical
  expertise also creates a big problem, as
  bureaucrats can use their knowledge against
  politicians. Bureaucratic expertise is thus a
  double-edge sword, creating both the incentive
  for legislatures to give policymaking power to
  bureaucrats and the opportunity for these
  bureaucrats to act counter to legislative
  preferences (Peters 1981)

91
Two prevalent views

• Bureaucratic discretion (bureaucratic dominance)
• versus
• Deliberate discretion (politicians dominance.
  That is, politicians would have capacity to
  control bureaucrats and achieve their goals by
  delegating)

92
Alternative view

• Rather than asking to what extent politicians
  control bureaucrats, the authors assume that in
  any political context there typically exists a
  politician who has greater opportunities than
  other political actors to influence bureaucratic
  behavior.
• In parliamentary democracy Cabinet
• In presidential system president or governors

93
4 factors that affect politicians incentives to
micromanage policymaking

1. Level of political conflict between politicians
   who adopt the statute and bureaucrats who
   implement them
2. The capacity of politicians to write detailed
   statutes
3. The bargaining environment (e.g. the existence of
   vetoes or bicameral conflicts) in which the
   statutes are adopted
4. The politicians expectation about non-statutory
   factors such as the role of courts or legislative
   oversight.