The trade-off between governance and checks and balances

Transcription

1 The trade-off between governance and checks and balances Alvaro Forteza FCS-UDELAR, Uruguay Juan S. Pereyra ECARES - Université libre de Bruxelles & FNRS decon-fcs Marzo, / 33

2 Political Accountability and C&B Principal Agent Problem: Voters vs Politicians Voters should control politicians to minimize politicians rents and, to align policies to citizens preferences. One important device: Elections (Barro, 1973; Ferejohn, 1986; Besley, 2005) but they may be insufficient Separation of Powers and C&B complement elections 2 / 33

3 Separation of Powers and Checks and Balances Source: Red Millennial.com 3 / 33

4 General Goal Formal model and case studies. Formal model of the trade-off: Governance vs CB ( Horizontal Accountability) How do voters manage this trade-off? Why and when do voters undermine checks on the executive? Acemoglu et al. (2013): Why Do Voters Dismantle CB? CB make politicians cheaper to bribe by elites Plausible only under pro-redistribution executives, but what if the executive is pro elite? Did voters support the dismantling of CB to make leaders more expensive to bribe? (controversial) 4 / 33

5 Are these guys so sophisticated and cynical? 5 / 33

6 Empirical Evidence Democracies have moved to stronger C&B (Besely and Persson, 2011). However, there have been reversals in some countries, let s see... 6 / 33

7 Peru under Fujimori : Economic & Political Crisis 90: Elections April 92: Auto-coup: suspending congress & Overriding judiciary 71% approve closing legislature 80% agree on the restructure of the judiciary July 92: Constitutional Congress (Strong Majority) July 93: New Constitution (less C&B) Unicameral legislature Reelection Presidential legislative power 1995: Reelection 7 / 33

8 Rhetoric The lack of identification of some fundamental institutions with the national interests, like the legislature and the judiciary, blocks the actions of the government oriented to national reconstruction and development. 8 / 33

9 Argentina under Menem : Economic Crisis 1989: Elections Main concern: Judiciary : Reform of the Supreme Court Expand the number of judges (5 to 9) Opaque appointing procedures Automatic Majority April 94: Constitutional convention (majority) Aug 94: New Constitution: Option for reelection 1995: Reelection 9 / 33

10 Judiciary Independence? My only bosses are Perón and Menem I cannot dictate a ruling that is against the government. I only issue rulings that are favorable to administration officials. (Judge Rodolfo Barra) 10 / 33

11 The 2000s: Anti-markets reforms Venezuela under Chavez: : Constitutional Reform (approved by 72%) Unicameral legislature Right to rule by decree Ecuador under Correa: : New constitution (approved by 68%) Allow for reelection 2011: Judiciary Reform:...the President of the Republic is not only the head of the executive power, he is the head of the whole Ecuadorian state, and the Ecuadorian state is the executive power, the legislative power, the judicial power, the electoral power,... Bolivia under Morales: : New Constitution Get rid of super-majorities Remove one-term limit 11 / 33

12 Common themes Strong leaders claiming the need of reforms Rhetoric: CB block reforms Citizens support the loosening of CB Executives vis-a-vis legislative and judiciary Not exclusive of left-wing government: 90s pro-market vs 2000s anti-market But then, Which elements convinced voters to support the loosening of CB? 12 / 33

13 Model: Outline Imperfect Information Game. Exec. and Leg. decide on policy issue and rent extraction. Agencies have a policy bias, like their own rents and dislike rents extracted by the other agency. Agencies observe the state of nature and make proposals (commitment). Voters are uncertain about the state of nature. Referendum to decide on the constitutional rule: CB or SP CB No reform + No rents SP Reform + rents 13 / 33

14 Contribution in a nutshell (cont d) X observes s and may call for referendum on special powers Voters vote for SP iff their expected gains from reform are larger than the costs of rent extraction. Expected gains from reform depend on Importance of policy agenda. Probability that the reform is beneficial. Executive credibility. Commitment. 14 / 33

15 Contribution in a nutshell (cont d) Two types of equilibria with special powers Truth-telling: aligned preferences. Only needed reforms are implemented. Always reform: strong pro-reform bias. Extensions Judiciary instead of Legislature No commitment Elections: endogenous divided government. Model 15 / 33

16 Model I Three players: Voters, Executive (X ), Legislature (L). Two policy dimensions: a policy issue p {0, 1} and r X, r L Two states of Nature s {0, 1}. (q = Pr{s = 1}) Payoffs: Citizens [ v = E s (p s) 2 ] c(r X + r L ), c > 0 Politicians u j = E s [ (p (s + δj )) 2] + a j r j b r j, a j > 0, j {L, X }, and b > 0 16 / 33

17 Model II Constitutional rules: Checks & Balances: Policies: p = p X p 0 if p X = p L if p X p L Executive chooses r [0, r] Legislature chooses r L [0, r]; r X [0, r r L ] Special Powers: Policies: p = p X Executive chooses r X [0, r]; r L [0, r r X ] We focus on a case in which: The status quo policy is p0 = 0. The legislature is conservative (the interesting and likely) case: p L = / 33

18 Timing p 0, s, X, L p X, p L Nature Executive & Legislature CB or SP Voters p, r X, r L t Executive & Legislature 18 / 33

19 Solving the model using backward induction After the referendum: perfect information games between the executive and the legislature 1. With CB: p = p 0 = 0, and r L = r X = With SP: p = p X, r L = 0, and r X = r. Before the referendum: imperfect information game between voters and the executive Equilibrium concept: Perfect Bayesian Equilibrium Extensive form / 33

20 The extensive form referendum game δ 2 X, 0 CB SP δ 2 X + a X r, c r Voters (1 + δ X ) 2, 1 CB SP (1 + δ X ) 2 + a X r, 1 c r 2 δx, 0 CB p X = 0 X p X = 1 s = 0, (1 q) SP (1 δ X ) 2 + a X r, 1 c r N Voters (1 + s = 1, (q) δx ) 2, 1 CB p X = 0 X p X = 1 SP 2 δx + a X r, c r 20 / 33

21 Equilibria with Special Powers Proposition 1 There are two, and only two, PBE with SP: 1. (0, 1, CB, SP) is a PBE iff c r 1 and 1+a X r 2 δ X 1 a X r 2 2. (1, 1, CB, SP) is a PBE iff c r 2q 1 1 and δ X 1 a X r 2 21 / 33

22 Equilibria without Special Powers Proposition 2 There are five, and only five, PBE where SP are never part of the equilibrium outcome. 1. (0, 0, CB, CB) iff c r 2Pr(s = 1 p X = 1) 1; 2. (0, 0, CB, SP) iff c r 2Pr(s = 1 p X = 1) 1 and δ X 1+a X r 2 ; 3. (1, 1, CB, CB) iff c r 2q 1; 4. (0, 1, CB, CB) iff c r 1; 5. (1, 0, CB, CB) is a PBE. (weakly Pareto dominated) 22 / 33

23 Equilibrium Outcomes δ X a X 2c 1 2 SP & reform 2q 1 c 1 c CB & 0 r CB & no reform if s = 0 SP & reform if s = 1 no reform δ X a X 2c 1 2 CB & no reform 1 a X 1 c 0 r CB & no reform if s = 0 SP & reform if s = 1 CB & no reform 1 2 a X 2c CB & no reform 1 2 a X 2c Figure: If q 1 2 Figure: If q > / 33

24 Special Powers at equilibrium Corollary 1 SP are part of the equilibrium outcome if, and only if, 1. Voters expected gains from reform are larger than their utility cost from rent extraction, i.e. if and Pr(s = 1 p X = 1) Pr(s = 0 p X = 1) c r, 2. the incumbent is not strongly pro-status-quo biased: δ X 1 + a X r 2 24 / 33

25 Extensions Judiciary No Commitment Or directly to the Concluding Remarks 25 / 33

26 Extension I: Judiciary Two regimes: 1. CB: 2. SP: p = p 0 r = 0 p = px Executive chooses r X [0, r] Conservative judiciary: always chooses p = p 0. u X = E s [ (p (s + δx )) 2] + a X r X, Back v = E s [ (p s) 2 ] c r X, 26 / 33

27 Extension II: No Commitment There is no information transmission between politicians and voters in the model without policy commitment. δ X SP CB 1 2 SP 0 r SP 2q 1 c q c 1 c CB CB 1 2 CB 27 / 33

28 Extension II: No Commitment (Welfare) Proposition 3 (The effects of commitment on welfare) Without commitment: 1. Voters and executive are weakly better off iff δ X 1 2 a X r 2 2. Voters are weakly worse off, executive is weakly better off iff δ X 1 2 a X r 2 and r q c Graph 3. Both voters and the executive are weakly worse off iff δ X 1 2 a X r 2 and q c r 1 c 4. With other parameter values, commitment has no impact on welfare. Back 28 / 33

29 Concluding Remarks Back to case studies: Why did voters support SP? q rose before these events: 80s: Lost decade 90s: Drawbacks of pro-market reforms Pro-reform leaders (opportunities for corruption?) Qualitative change: From CB and no reform to always reform equilibrium Next Step. Introducing Elections: Political Accountability Through Elections and Checks and Balances 29 / 33

30 Thank you 30 / 33

31 No Commitment I: Better off without commitment δ X SP & reform CB & 1 2 2q 1 c no reform 1 2 a X 2c c 0 r CB & no reform if s = 0 SP & reform if s = 1 CB & no reform q c 1 2 a X 2c Figure: If q > 1 2 Back 31 / 33

32 No Commitment II: Voters worse off without commitment & Executive better off δ X 1 2 SP & reform CB & no reform 1 2 a X 2c c 0 r CB & no reform if s = 0 SP & reform if s = 1 CB & no reform q c 1 2 a X 2c Figure: If q > 1 2 Back 32 / 33

33 No Commitment III: worse off without commitment δ X 1 2 SP & reform CB & no reform 1 c 0 r 1 2 CB & no reform if s = 0 SP & reform if s = 1 CB & no reform q c 1 2 a X 2c Figure: If q > 1 2 Back 33 / 33