EITM Institutions Week

1
EITM Institutions Week

• John Aldrich
• Duke University
• Arthur Lupia
• University of Michigan

2
Implications of Multiple Chambers

• Tsebelis on Veto Players and Policy Stability
• Diermeier Myerson on Bicameralism and the
• Organization of Legislatures
• Bawn on Choosing the German Electoral Law

3
Tsebelis, Decision Making in Political Systems

• M Compare democratic political systems in
  terms of policy stability/change.
• NH Regime type (presidentialism vs.
  parliamentarism) and party systems (two vs. many
  parties) are the crucial ways by which democratic
  political systems matter.

4
Tsebelis, Decision-Making, contd.

• P
• Veto players (one whose assent is necessary to
  change the status quo) are central.
• Institutional vetoers (e.g., president, chamber)
  and party are the crucial veto players.
• Change is less likely as
• The number of veto players increase
• Congruence of policy preferences among veto
  players decrease
• Cohesion (internal policy agreement) decreases.
• Size of win set is key to stability

5
Tsebelis, Decision-Making, contd.

• C
• As number of players increases, the win set of
  the status quo does not increase (stability
  decreases).
• As distance among veto players increases along
  the same line, the win set does not increase.
• As size of yolk of veto players increases, the
  win set increases
• N.B. determination of veto players complex
  even so, data from others consistent with
  derivations.

6
Tsebelis, Veto Players and Law Production in
Parliamentary Democracies

• Empirical tests of the following hypotheses
• Increase in the number of parties in government
  reduces the ability to produce significant
  legislation.
• Increase in ideological distance reduces the
  ability to produce significant legislation.
• Number of sig. laws increase with government
  duration.
• Number of sig. laws increases with increasing
  distance between the current and preceding
  governments.

7
EITM of Institutional Choice
8
Bawn (1993)

• M. Questions about institutional choice.
• NH. Institutional choice is a matter of social
  engineering, it is not political.
• P.
• Party preferences are defined over policy
  outcomes.
• Parties use information to predict vote shares.
• Parties know other parties policy preferences.
• C. Decisions on German election law in 1949 and
  1953 are consistent with Bawns political model.

9
1949. CDU favored FPTP. SPD small parties
favored PR.
What assumptions are needed to produce the
estimate?
10
1953. SPD small parties favored split ballot.
CDU did not.
Incumbents, SPD, and smaller parties gain.
11
The Logic of Institutional Preferences

• What would you do?

12
Diermeier and Myerson (1999)

• M. What explains variations in committee systems
  across countries?
• NH. Informational and distributional theories are
  sufficient to answer the question.
• P. Legislative chambers compete in a market for
  legislation. Procedural hurdles as legislative
  prices.
• C. Presidential veto power and bicameral
  separation encourage chambers to create internal
  hurdles.

13
Diermeier and Myerson (1999)

• The model is not game-theoretic ? a theorem.
• It builds on Groseclose and Snyder (1996, g-t).
• Agent 1 wants a bill to pass. Agent 0 wants it to
  fail.
• Each can pay money to individual legislators
  conditional on them supporting the bill.
• Legislator utility is payment from lobbyists.
• Q When are House members expected payoffs
  increased by raising the Houses hurdle factor?

14
Hurdle Factors

• Q?0,1 is the proportion of the legislature
  whose assent is required for a bill to pass.
• Any coalition larger than 1-Q can block the bill.
• Agent 1 must promise a sufficient amount to each
  potential blocking coalition.
• The hurdle factor is 1/(1-Q) plus N of veto
  players.
• Dictator in the chamber 1.
• Simple majority rule 2.
• 2/3 supermajority 3.
• Simple majority rule plus one internal veto
  player 3.

15
Diermeier and Myerson (1999)

• The Houses hurdle factor is s.
• The external hurdle factor is t.
• V is what Agent 1 can pay.
• W is what Agent 0 can pay.
• What value of s should the House choose?

16
Diermeier-Myerson main result
17
Bicameralism and Its Consequences for the
Internal Organization of Legislatures

• What would you do?

18
Candidate Entry and CompetitionThe Rationality
of Candidate Emergence in Seemingly Uncompetitive
Elections
19
A Later Example of EITM

• John Aldrich
• and
• Arthur Lupia

20
Rohde, Progressive Ambition

• M There are no variable types of (careerist)
  ambition, but one expected utility maximizing
  problem for all.
• NH Ambition for higher office is exogenous.
• P Incumbents are decision makers, not solving a
  career ambition game at equilibrium.
• Risk-taking is an exogenous, psychological
  trait.
• C At least for House incumbents, incumbency is
  attractive due to electoral safety, a Senate seat
  is more desirable than a Governorship, and the
  probability of winning the higher office seems
  most crucial.

21
Banks Kiewiet

• M Empirical research is insufficiently
  theoretically grounded to provide insights into
  candidate competition.
• NH Inexperienced challengers to incumbent
  Members of Congress make foolish decisions with
  respect to politics.
• P Game theoretic principles three person
  context.
• Quality means higher probability of
  defeating incumbent.
• C Weak challengers run for out-party nomination
  to oppose an incumbent because that yields a
  higher probability of winning office than
  opposing a strong challenger in the primary.

22
Aldrich Bianco

• M Choice of party affiliation is endogenous to
  career aspirations.
• NH Party is chosen by ideology or beliefs or
  up-bringing. It is, like party identification in
  the electorate, all but an unmoved mover.
• P Game theoretic premises. Affiliation changes
  can be due to change in electoral support for a
  party but may also be due to seeking to avoid
  contested primaries
• C As support for party declines, party
  affiliates face a collective action problem.

23
Carson

• M Estimate game theoretic model of candidate
  emergence with game theoretic-consistent
  methodology.
• NH Standard, decision theoretic probit
  estimation yields sufficiently accurate
  estimates.
• P Game theoretical setting, with endogenized
  disturbence term.
• QRE equilibrium concept.
• C Incumbent spending significantly reduces the
  probability of a subsequent challenger entering.
  Challengers pay attention to the incumbents
  voting record on floor. Both are different in
  standard, decision theoretic probit model.

24
Basic Game
(1-m)(1-b)
md
m(1-d)
(1-m)b
Probabilities in Equilibrium
25
Probit and Strategic Probit Models of Candidate
Competition in U.S. House Elections, 1990-2000
Robust standard errors in parentheses.
Significant at p lt 0.05
26
Shepsle and Muthoo (2003)

• M. Do staggered terms affect Senate dynamics?
• NH. Individual-level variations in time before
  election do not affect Senate bargains.
• P.
• The Senate contains an old player and a young
  player.
• Each period, they play an ultimatum game.
• ? a what have you done for me lately heuristic.
• C. Under broad conditions, most agenda setting
  power is allocated to the old player.

27
Shepsle/Muthoo
28
ShepsleMuthoo
29
WHYDFML
30
The Solution
Corollary 2. In case i, the old player still gets
more power.
31
Agenda Setting Powers in Organizations with
Overlapping Generations of Players

• What would you do?
• Dont answer yet.

32
Shepsle, Dickson and VanHouwelling (2003)

• M. What are the implications of staggered terms?
• NH. Individual-level variations in time before
  election do not affect Senate bargains.
• P
• The Senate contains an old player and a young
  player.
• Each period, they play a divide-the-dollar
  game.
• ? a what have you done for me lately heuristic.
• C. Under broad but different -- conditions,
  most agenda setting power is allocated to the old
  player.

33
Baron-Ferejohn format
With common knowledge and symmetric information,
the first proposal is accepted in equilibrium.
34
WHYDFML
There are now three generations of legislators in
the game.
35
Initial Conclusions
36
Ultimate Conclusions
37
Bargaining in Legislatures with Overlapping
Generations of Politicians

• What would you do?