Majoritarianism and Transparency

1
Majoritarianism and Transparency

• B. Peter Rosendorff, NYU
• James R. Vreeland, Yale

IPES, Stanford University, November 2007
2
Majoritarianism

• Majoritarianism
• Pro-consumer rather than pro-producer
• Policies support broad rather than narrow
  interests
• Rogowski and Kayser (2002), Linzer and Rogowski
  (2008)
• the greater vote-seat elasticities of
  majoritarian electoral systems will tilt policy
  in favor of consumers (p. 526).

3
Transparency

• Transparency in policymaking
• Voters are better able to monitor the incumbent
• Rosendorff and Vreeland (2007), Doces and
  Rosendorff (2007)
• greater electoral accountability is associated
  with the adoption of transparent institutions.

4
Majoritarianism and Transparency

• Might one infer
• Majoritarian systems are more transparent?
• NO.
• Key Result
• More majoritarian systems are likely to be less
  transparent.

5
Theory

• Majoritarian polities
• Convert votes into seats in the legislature in a
  more decisive fashion
• Winner-take-all
• Single member majoritarian system 50 of the
  votes in 50 of the districts yields a majority
• 25 of the votes ? 50 of seats
• Pure PR system
• 25 of the votes ? 25 of seats.
• More seats for a given fraction of the votes.

6
Incumbent Seats and Seignorage.

• Incumbent cares about seats, S and returns from
  extraction (seignorage) v
• Incumbent chooses inflation rate p
• Marginal utility of seats declines
• Seats are determined by votes
• Votes are determined by the voting behavior of
  the voters

7
Voters Non-transparency

• Voters only see their money balances at election
  time
• Cant distinguish between inflation and bad
  shocks
• Inflation erodes their consumption possibilities
• Aggregate iid shock affects holdings
• Given any reelection threshold, U
• A voter will vote in favor of the incumbent if
  utility exceeds the threshold.
• Ex ante probability of voting for incumbent

8
Votes to Seats

• Incumbent is share of seats in legislature
• Strict PR
• Plurality SMD
• British House of Commons / US House of
  Representatives
• US Electoral College
• Larger t, more majoritarian

9
Equilibrium

• Given U, best response for incumbent
• Yields
• Comparative static

10
(No Transcript)
11
Intuition

• For a given number of votes
• Majoritarian systems will have more seats than
  non-majoritarian
• Marginal utility of seats declines
• Marginal utility of last seat is lower for
  majoritarian systems
• More likely to forgo that seat for increased
  extraction.
• For majoritarian system
• Inflation rate the balances marginal benefit with
  marginal cost is higher
• Incumbent payoff is higher.

12
Transparency

• Incumbent announces a policy
• Independent agency (World Bank) affirms
• Voters

13
Equilibrium under Transparency
In any equilibrium, incumbent chooses an
inflation rate below the non-transparent
level agency confirms announcement voters
reelect with probability 1 Incumbent receives
maximal seats Payoff to the incumbent is now not
a function of p is not a function of t
14
(No Transcript)
15
Key Result

• If the information environment is a matter of
  choice to the incumbent
• Then high majoritarian systems
• more likely to choose non-transparency
• Non-majoritarian systems
• more likely to choose transparency

16
Non-majoritarian Transparency
Majoritarian Non-transparency
17
Switching

• Benefits to switching to transparency
• Lower risk of unfair punishment
• When aggregate shocks are negative
• Costs of switching
• Fewer opportunities for extraction
• Commit to a lower inflation rate

18
Intuition

• Majoritarian
• Low marginal utility of last seat
• Preserve opportunities for extraction
• Extraction not that costly
• Risk of bad shock not large
• NON-TRANSPARENT
• Non-majoritarian systems
• High marginal utility of last seat
• Willing to give up opportunities for extraction
• In return for a reduction in the risk of unfair
  punishment
• TRANSPARENT

19
Last time with Rosendorff Vreeland
20
Lo and behold!
21
Table 1 The effect of majoritarianism on data
dissemination
Unemployment data
 
Inflation data
 
Cox
Dynamic Logit
Region Dummies Logit
Country F.E Logit
Logit
Logit
 
Logit
Logit
Variable
-0.48
-1.22
-0.83
-11.08
-1.38
-1.57
 
-0.65
-1.06
Majoritarian
(0.24)
(0.28)
(0.25)
(1.13)
(0.15)
(0.12)
 
(0.32)
(0.18)
 
0.0001
0.0001
0.0003
0.0012
0.0003
 
 
0.0001
 
GDP/capita
(0.0000)
(0.0000)
(0.0000)
(0.0003)
(0.0000)
 
 
(0.0000)
 
 
0.2832
-0.287
0.83
1.06
1.18
 
 
-0.07
 
IMF
(0.39)
(0.33)
(0.20)
(0.44)
(0.16)
 
 
(0.28)
 
 
 
-1.69
0.002
 
-1.02
1.42
 
3.40
3.23
Constant
 
(0.29)
(0.82)
 
(0.17)
(0.08)
 
(0.38)
(0.14)
 
349
435
1347
514
1417
1570
 
2122
2291
of obs
22
Clarifying

• Inflation
• Majoritarian systems are 4 to 9 less likely to
  report data than non-majoritarian systems.
• Results are not robust to fixed effects or
  duration dependencebut democracies already
  report 90 of the time.
• Unemployment
• Majoritarian systems are 30 to 39 less likely
  to report data than non-majoritarian systems.
• Results are robust to fixed effects, regional
  effects, duration dependence.

23
Conclusion

• Peter is right.
• Rogowski and Kayser
• Majoritarian systems protect broad interests
• This paper
• Majoritarian systems less transparent
• Key difference
• RK Conflict is distributional
• Between narrow and broad interests
• Here Conflict is between voter and executive
• Between median voter and policymaker
• More majoritarianism
• More representative policy comes at the cost of
  
• More extraction by the incumbent.