Political Economy Chapter 4: Agency

1
Political EconomyChapter 4Agency

• Persson Tabellini

Marcelo Tyszler Lorena Barberia Carol Fonseca
18/mar de 2005
2
Outline

• Politicians pre-commit to policies before
  elections
• Efficient Electoral Competition
• Inefficient Electoral Competition (ex-post
  discretion on implementation)
• Politicians with difficulties in pre-commitment
• Enforceable, verifiable promises (e.g. Judiciary)
• Enforceable, non-verifiable promises
• Non-enforceable promises
• Politicians do not pre-commit to policies
  (non-enforcement)
• Electoral Accountability and re-election of
  incumbents
• Incumbents have power
• Asymmetric information
• Career Concerns and re-election of incumbents
• Two-period model
• Electoral cycles

3
Agency Politicians versus Voters

• Principal-Agent Theory
• The person (the agent) whose job it is to act to
  the benefit of someone else (the principal), but
  who may require some incentive to do so.
• E.g., agent is politician and the principal is
  the voters

• Goal of chapter
• Explore to what how agency operates in
  majoritarian, two-party elections for the
  delivery of optimal government spending (based on
  voters utility preferences) when candidates have
  potential to extract endogenous rents.
• We want how elections discipline candidates
  (enforcement) in terms of guaranteeing spending
  on government services and how voters limits on
  information may influence the potential
  disciplining force of competition.

4
The Voters Problem
5
The Voters Problem with Ideology
6
The Candidates Problem
7
Agency The Problem
8
4.1 Efficient Electoral CompetitionKey Variables

• q Policy
• g government spending
• tax rate
• r endogenous rents
• R exogenous ego-rents
• Wi(q) Indirect utility function
• H(g) increasing function part of indirect
  utility function function of publicly provided
  goods
• transaction costs associated with rent
  appropriation (exogenous)
• m median voter

9
4.1 Efficient Electoral CompetitionAssumptions

• Income distribution is continuous
• Candidates have no ideological attributes and are
  identical in all respects
• Timing
• 1) Platforms are announced (g) level of govt
  spending
• 2) Elections are held between two parties or two
  candidates.
• 3) Winner implements platform (g).

10
4.1 Efficient Electoral Competition

• Chapter 3
• Median voter equilibrium (Condorcet winner)
• gm Hg-1(ym/y)
• All parties converge
• gm gA gB
• Political equilibrium median y
• Govt spending average y

• Chapter 4
• Median voter equilibrium still holds.
• gm Hg-1(ym/y)
• All parties converge
• gm gA gB
• rm rA rB 0
• RENTS ZERO

11
4.1 Efficient Electoral CompetitionImplications

• Equilibrium public goods are provided (gA gB).
• Size of government spending, g, is determined by
  average income.
• Political equilibrium (convergence to same
  policy choice for g) is determined by median
  income.
• Political competition drives endogenous rents,
  r, to zero.

12
4.1 Inefficient Electoral Competition Key
Variables

• J income groups
• siJ individual specific parameter (ideological
  bias if gt0 ideological bias in favor of
  candidate), uniformly distributed.
• Densities summarize how responsive voters are to
  policy, number of swing voters
• average relative popularity of candidate,
  uniformly distributed

13
4.2 Inefficient Electoral CompetitionAssumptions

• Income distribution is discrete among three
  groups J.
• Candidates have ideological attributes
• Voters have preferences on these attributes,
  distribution of these preferences is identical,
  number of swing voters is identical for all J
• Timing
• 1) Platforms are announced (g) level of govt
  spending
• 2) The actual values of is realized.
• 3) Elections are held between two parties or two
  candidates.
• 4) Winner implements platform (g).
• Sources of inefficiency rent-seeking and
  electoral competition in which ideology weakens
  competition

14
4.2 Inefficient Electoral Competition

• Chapter 3
• Voter J is indifferent
• Probability Candidate A wins
• Equilibrium determined by swing voter

• Chapter 4
• Both candidates convergence to equilibrium
  spending on public goods.
• However, equilibrium rents are gt0

Because y y, gs g All groups get equal
weight, end result is to maximize average voters
utility. Convergence to equilibrium spending on
public goods.
15
4.2 Inefficient Electoral CompetitionImplication
s

• Equilibrium public goods are provided. (gA gB).
  
• Political equilibrium (PA PB ) is determined
  by income .
• Political competition does not necessarily drive
  endogenous rents, r, to zero.
• Greater endogenous rents
• Higher uncertainty of election outcome (marginal
  decrease in probability of winning)
• Lower exogenous value of holding office, R
• Lower transaction costs associated with rent
  appropriation

16
4.3 Enforceability, Verifiability, and
Observability Assumptions

• yi y all voters earn same income
• the cost of transforming private output
  into public goods cost of public goods
• Timing
• 1) Platforms are announced (g) level of govt
  spending
• 2) Elections are held between two parties or two
  candidates.
• 3) , cost of transforming public goods, is
  realized.
• 4) Winner implements platform (g).
• Sources of inefficiency rent-seeking and
  electoral competition in which ideology weakens
  competition candidates have ex-post discretion
  on implementation

17
4.3.1 Enforceability, Verifiability, and
Observability Implications if promises are
enforceable and verifiable ( e.g. Existence of
Judiciary)

• Voters prefer candidate that give them highest
  utility. Probability of winning varies depending
  on utility derived from A vs. B.

18
4.3.1 Enforceability, Verifiability, and
Observability Implications if promises are
enforceable and verifiable ( e.g. Existence of
Judiciary)

• The two candidates or parties converge to same
  equilibrium policy
• Equilibrium supply of public goods varies with
  cost of transformation from private to public.
  The efficient supply of public goods is
  decreasing in terms of cost of public good.
  Higher cost, lower level of g.

19
4.3.1 Enforceability, Verifiability, and
Observability Implications if promises are
enforceable and verifiable ( e.g. Existence of
Judiciary)

• Equilibrium rents
• Equilibrium level of taxes increases with cost
  of transformation.

20
4.3.2 Enforceability, Verifiability, and
Observability Assumptions if promises are
enforceable, but promises are not verifiable

• Assume either two states are possible
• low cost of public goods, lower taxes,
  higher spending on g
• high cost of public goods, higher taxes,
  lower spending on g
• Source of inefficiency Uncertainty and
  Non-verifiability. Politicians have incentive to
  claim high cost and benefit from rents they keep
  ( if reality low cost). However, if real state is
  high cost, candidates must deliver goods and can
  not capture rents for themselves.
• Claim The best voters can hope for is policy in
  expensive state (higher taxes!), but guaranteed
  g.

21
4.3.2 Enforceability, Verifiability, and
Observability Implications if promises are
enforceable, but promises are not verifiable

• Voters maximize their utility and prefer
  outcome with higher cost of public goods and
  guaranteed level of goods,
• Taxes will be collected at high-level

22
4.3.2 Enforceability, Verifiability, and
Observability Implications if promises are
enforceable, but promises are not verifiable

• However, equilibrium level of rents depend on
  realized state

23
4.3.3. Enforceability, Verifiability, and
Observability Implications if promises are
not-enforceable,

• Without enforcement, politicians claim high cost
  and deliver nothing to voters- -Levithan state
• Politicians will claim all y and taxesy
• Politicians will keep rents, r (theta)y
• No spending on g (theta) 0

24

4.4.1 Electoral Accountability and Elected
Incumbents Assumptions

• Opponent is identical in all aspects to
  incumbent.
• Voters punish incumbent for bad behavior and
  reward for good behavior.
• Timing
• 1) Cost of providing public goods is
  realized and observed by all.
• 2) Voters set a reservation utility for
  reelecting incumbent.
• 3) Incumbent sets policy qI
• 4) Elections are held between incumbent and
  opponent.

25

4.4.1 Electoral Accountability and Elected
Incumbents Implications

• If utility of voters is greater or equal to
  reservation utility, probability incumbent is
  reelected is 100.
• Equilibrium rents are the same, irrespective of
  the state of the world.
• Politician maximizes rents, subject to
  constraint that
• . If current
  future rents are greater than the amount he can
  extract from taxes, he will spend on g. If not,
  he will keep all of taxes for himself. The
  optimal amount of rents

26

4.4.1 Electoral Accountability and Elected
Incumbents Implications

• If voters coordinate their vote and set their
  optimal strategy, they can obtain the optimal
  level of public goods in every state. However,
  voters utility is state contingent. Voters set
  utility equal to the average reservation utility
  for given state of the world
• Where equilibrium rents, decreasing with cost of
  rent extraction and increasing with exogenous
  value of public office, R

27

4.4.2 Electoral Accountability and Asymmetric
Information Assumptions

• Opponent is identical in all aspects to
  incumbent.
• Voters punish incumbent for bad behavior and
  reward for good behavior, but do not have
  information on state of world that is realized.
• Timing
• 1) Cost of providing public goods is
  realized, only incumbents observe.
• 2) Voters set a reservation utility for
  reelecting incumbent.
• 3) Incumbent sets policy qI
• 4) Elections are held between incumbent and
  opponent.

28

4.4.2 Electoral Accountability and Asymmetric
Information Implications

• If utility is greater than or equal to average
  reservation utility, probability incumbent is
  reelected is 100.
• Supply of public goods depends on state of the
  world, with lower levels of public goods as the
  price of providing goods increases

29

4.4.2 Electoral Accountability and Asymmetric
Information Implications

• Equilibrium rents now depend on the state of the
  world. If theta is the optimal outcome where
  utility is equal to voters reservation utility,
  rents will be greater

30

4.5 Electoral Accountability and Career Concerns
in a Two-Period Model Assumptions

• Taxes are fixed.
• The marginal utility from consumption is
  constant, risk neutrality.
• Government budget constraint is
• which are uniformly distributed
• Voters punish incumbent for bad behavior and
  reward for good behavior, but do not have
  information on state of world that is realized.
• Timing
• 1) Incumbent in office in period 1 chooses rents
  in period 1 without knowing his competence.
• 2) The value of competence is realized.
• 3) Elections are held between incumbent and
  opponent.
• 4) Period 2 rents are set and public goods are
  residually determined.

31

4.5 Electoral Accountability and Career Concerns

• Roles of election
• 4.1 and 4.2 ? Select among economic policies
• 4.4 ? Hold incumbent accountable ex-post for bad
  behavior
• 4.5 ? Select the most competent or talented
  politician
• Additional Incentive to perform well
• Want to appear talented to the voters
• Past Economic performance as signal of future
  competence
• Electoral Cycles shortly before elections, the
  incentives are higher to appear competent

32

4.5.1 Electoral Accountability and Career
Concerns in a Two-Period Model Assumptions

• Taxes are fixed.
• The marginal utility from public consumption is
  constant, risk neutrality.
• Government budget constraint is
• which are uniformly distributed, with
• Voters can punish incumbent for bad behavior and
  reward for good behavior, but do not have
  information on state of world that is realized.
• Objective function for politician at period 1

33
4.5.1 Electoral Accountability and Career
Concerns in a Two-Period Model

• Timing
• 1) Incumbent in office in period 1 chooses rents
  in period 1 without knowing his own competence.
• 2) The value of competence is realized.
• 3) Voters observe their utility, but not
  competence neither rents
• 4) Elections are held between incumbent and
  opponent.
• 5) Period 2 rents are set and public goods are
  residually determined.
• At period 2, there is no incentive for good
  behavior
• The higher the competence, the better off are the
  voters

34

Electoral Accountability and Career Concerns in a
Two-Period Model Implications

• Lets call
• as an estimate
  to competence
• he has
  probability of winning reelection
• He has to set r1 such that

35

4.5.2 Electoral Accountability and Career
Concerns in a Two-Period Model Implications
36
(No Transcript)