University:
Boston University
Course:
CAS PO 502 | Political Analysis
Academic year:

2022
Views:

371

Pages:

20
Author:

gotejatsgrr7

NECESSARILY φ(T̃ ) = δ. Since non-transgressors are always preferred to an entrant, Since an incumbent leader is willing and able to stand for power again with probability δ Consequently ( 0 if ρ < ρ∗ φ(T ) = δ otherwise (8) and ( σA ρ= 1 − σD if φ(T ) < φ∗ otherwise. (9) Equilibrium Proposition If 1 − αc · (1 − β · δ) ∈ (σA , 1 − σD ) then we have two stable norms. They are: 1 ‘non-permissive’: (0, σA ) and 2 ‘permissive’: (δ, 1 − σD ). Permissive σD ρ∗ 1 − σH δ φ∗ φ (Pr[cheater re-elected]) δ0 Non-Permissive ρ (Pr[politician cheating]) Figure 1: Multiple Norms - Phase Diagram Many examples of non-permissive equilibria Corruption accusations (unproved, unprosecuted) electorally costly in many developed democracies Roosevelt’s stacking the supreme court Weingast – Limits observed by King, enhanced power of Parliament, after the Glorious Revolution Permissive equilibria? 41 percent of Italy’s postwar (upto 1994) deputies were charged with serious crimes (excluding slander, libel, defamation of character). Yet most deputies were reelected during this period, and the difference in reelection probabilities between those who were charged and those who were not is relatively slight, 51 percent compared with 58 percent. Golden 2009 “India also has a culture in which people commonly assume that everyone in public life is corrupt unless they prove otherwise.” Jagdish Baghwati Globe and Mail, Jan 4 2011. Two Term Limits in Africa – Vencovsky (2008) Since 1990, 18 African presidents completed two terms in office (under 2-term constraints) 10 attempted constitutional amendments 7 were successful, 3 that were not hand picked successors Ideas Case 2: Introducing some behavioural types Not all citizens are rational, in every period some are ‘irrationally’ gripped by an idea The idea: leaders MUST respect citizen’s rights Recent examples: Anna Hazare – corruption Arab Spring – democratic representation So, every period, some proportion of citizens will stand against a transgressor no matter what Nitish Kumar (Bihar Chief Minister) “People of India had enough of corrupt politicians and are not going to tolerate it anymore.” Patna daily April 11 2011. Fukuyama (2011 p.3) “The shift to Democracy was a result of millions of formerly passive individuals around the world organizing themselves and participating in the political life of their societies.” Key question is how many people? There are two possible states: State 1 (Normal State): More rational than gripped State 2 (Gripped State): More gripped than rational (most citizens will stand against a transgressing incumbent no matter what) In the gripped state transgressing leaders are deposed. A citizen does not directly know how widespread the gripped feeling is A leader knows the state Formally The state space is {G , G̃ } When an incumbent leaves office, the new leader in power operates under the same state as the departing incumbent with probability s and draws a new state with probability 1 − s. The parameter s captures persistence in the state across successive leaders. When a new state is drawn it is G with probability λ. Citizen’s share common beliefs about the probability of being in state G , denoted by π. State Transitions: G → G , s + (1 − s)λ ≡ π 0 G → G̃ , (1 − s)(1 − λ) G̃ → G , (1 − s)λ ≡ π 0 G̃ → G̃ , s + (1 − s)(1 − λ) Value Functions Again characterize equilibria that feature “Political Norms”, (φ(a), ρ) Write down value functions given norms taking beliefs, π, as given Look for consistent norms and beliefs required to accompany them Value Functions Let G (T̃ ) be the expected value to citizens of having a non-transgressing incumbent in the gripped state, G (T ) be the expected value to citizens of having a transgressing incumbent in the gripped state, G̃ (T̃ ) be the expected value to citizens of having a non-transgressing incumbent in the non-gripped state, and G̃ (T ) be the expected value to citizens of having a transgressing incumbent in the non-gripped state. EG be the expected value to citizens associated with an entrant leader in the gripped state, and EG̃ be the expected value to citizens associated with an entrant leader in the non-gripped state. These six value functions are related as follows G (T̃ ) = 1 + α + β · [φ(T̃ ) · G (T̃ ) + (1 − φ(T̃ )) · EG ] G (T ) = 1 + α − c + β · EG (10) (11) G̃ (T̃ ) = 1 + α + β · [φ(T̃ ) · G̃ (T̃ ) + (1 − φ(T̃ )) · EG̃ ] (12) G̃ (T ) = 1 + α − c + β · [φ(T ) · G̃ (T ) + (1 − φ(T )) · EG̃ ] (13) EG = π 0 · [σA · G (T ) + (1 − σA ) · G (T̃ )] +(1 − π 0 ) · [ρ · G̃ (T ) + (1 − ρ) · G̃ (T̃ )] − α EG̃ = π 0 · [σA · G (T ) + (1 − σA ) · G (T̃ )] +(1 − π 0 ) · [ρ · G̃ (T ) + (1 − ρ) · G̃ (T̃ )] − α. Let π1 be the posterior belief (after observing leader’s current period behaviour) of being in the G state. Let π2 be the posterior belief (after observing the leader’s outcome) of being in the G state. Belief Updating politician acts π0 voting occurs π1 Markov transition π2 π00 Again, considering individual citizens who choose support based on value functions Letting V (T | π1 ) be the expected value to supporting a transgressor, Ṽ (T | π1 ) be the expected value to not supporting a transgressor, V (T̃ | π1 ) be the expected value to supporting a non-transgressor, and Ṽ (T̃ | π1 ) be the expected value to not supporting a non-transgressor, we have V (T | π1 ) ≡ π1 · G (T ) + (1 − π1 ) · G̃ (T ) (14) Ṽ (T | π1 ) ≡ π1 · EG + (1 − π1 ) · EG̃ (15) V (T̃ | π1 ) ≡ π1 · G (T̃ ) + (1 − π1 ) · G̃ (T̃ ) (16) Ṽ (T̃ | π1 ) ≡ π1 · EG + (1 − π1 ) · EG̃ . (17) Note that V (T̃ | π1 ) ≥ Ṽ (T̃ | π1 )∀π1 It’s always preferred to support a non-trangressor, implying that: φ(T̃ ) = δ in any regime. (18) Definitions Non-permissive norms are “consistent” at beliefs π1 if Ṽ (T | π1 ) > V (T | π1 ). Permissive norms are “consistent” at beliefs π1 if Ṽ (T | π1 ) < V (T | π1 ). Non-permissive norms (Transgressors opposed): (φ(T ) = 0, ρ = σA ) Proposition: Non permissive norms are consistent at π1 provided α c · (1 − β · δ) ≤ 1 − σA I.e., same condition as in the stationary model Beliefs about the state do not affect the existence of this political norm I.e, this political norm is consistent no matter what citizens believe about the state Permissive norms (Transgressors unopposed): (φ(T ) = δ, ρ = 1 − σD )