Unit 4: Endogenous Candidates C&D (2004)

Question 1

One idea in the endogenous candidates model is to allow the type of candidate to differ, therefore implemented policies depend on the C elected. Give an example of this. What are the two views around this type of model?

Answer 1

Evidences shows that more women elected -> more women friendly policies introduced.

View 1: Candidates CAN fully commit to policy before being elected, tf identity is irrelevant
View 2: Candidates CAN’T fully commit to policy before being elected

Question 2

Explain the framework of Chattopadhyay and Duflo (2004)?

Answer 2

Same as O+S (1996) except citizen candidates can no longer commit to a specific policy, and there are different fixed costs of entry between candidates

Question 3

Part of the C+D (2004) model we look at?

Answer 3

When citizens cannot commit to a specific policy THEREFORE they implement their preferred policy!

Question 4

C+D (2004) What do we assume about candidates?

Answer 4

They KNOW each other’s preferences, and use voting (strategically) to influence the implemented policy!

Question 5

C+D (2004) What country and area type does the study use?

Answer 5

India, village level

Question 6

C+D (2004) How do costs vary between candidates?

Answer 6

Women have a higher fixed cost of running than men

Question 7

C+D (2004) Policy space assumptions? (2)

Answer 7

1) W+Ms preferences can overlap

2) All preferences can be ordered on 1-D policy space

Question 8

C+D (2004) Explain how the policy preferences are set-up in the model?

Answer 8

• A policy is chosen in the [0,1] interval
• Each citizen ideal policy: ω(i)
• Men and Women are assumed to have different preferences: W∈[0,W], M∈[M,1] (M can be less than W tf overlap allowed)

Question 9

C+D (2004) Each citizen chooses whether to enter/not. What happens if they enter? What are the relative payoffs or entering and not entering?

Answer 9

If they enter they face a fixed cost: 𝛿, where 𝛿(W) greater than 𝛿(M).
Payoffs from implemented policy x(j):
Non-candidate citizen i payoff: -|x(j)-ω(i)|
Candidate citizen i payoff: -|x(j)-ω(i)| - 𝛿(i)

Question 10

What is the fixed cost viewed as in the paper?

Answer 10

A reputation cost

Question 11

What is the final policy implemented?

Answer 11

It is a MIX between the ideal policy of the elected candidate AND the elite’s policy option, μ’

Question 12

2 reasons the elected candidate can’t implement exactly what their policy-stance is?

Answer 12

1) they may be vulnerable to elite’s wishes/pressures

2) they may be controlled by the village assembly

Question 13

Assumption about the preferences of the elite?

Answer 13

They are more PRO-MALE than the MVs preferences:

μ’>m

Question 14

Equation for the policy implemented by the winner?

Answer 14

x(j) = αω(j) + (1-α)μ’

Question 15

What happens if no one enters?

Answer 15

μ’ is implemented

Question 16

2 main results of the paper?

Answer 16

1) The implemented policy can be relatively far from the MV

2) There exists a range of parameters whereby there is no woman candidate equilibrium

Question 17

What is a solution to the no woman candidate equilibrium region?

Answer 17

Reservation policy; ie. seats being held for women!

Question 18

2 key questions the paper asks?

Answer 18

1) When will a woman run, even without a reservation policy?
2) When will the reservation policy improve:
a) welfare of the female MV
b) Welfare of overall MV

Question 19

What are the conditions for:
a) No women enter when unopposed?
b) No women enter when against a man?
What does this show?

Answer 19

a) 𝛿(W)-0.5𝛿(M) > μ’-m

b) 𝛿(W) > m-(1-α)μ’
(don’t quite get this)
Entry costs are too high for women; only women with ‘extreme views’ would have motivation to run, but then a man (with less extreme views and lower costs) would enter and win for certain tf no motivation for women to enter!

Question 20

Will the reservation policy help?

Answer 20

Not necessarily; still range of parameters whereby no women run

Question 21

What happens if 𝛿(W)>αμ’?

Answer 21

No women will run even with the reservation policy since either the cost of entry is too high, or the power of the elected candidate is too weak relative to the power of the elite

Question 22

What happens if 𝛿(W) is too high?

Answer 22

Reservation will replace democracy - with lobby! (why????????????) slide 43

Question 23

When will the reservation system increase the utility of both the MV and the female MV?

Answer 23

𝛿(W) is less than/equal to μ’-(1-α)μ’ and P1 is respected(?), AND if the most pro-woman outcome implemented by a man is to the right of the most pro-man policy implemented by a woman (ie. they overlap) (slide 44)

Question 24

How was the model tested?

Answer 24

Preferences were measured by number of M and F complaints to the council on different issues
Implemented policy measured by different types of policies being implemented

Question 25

Hypothesis of the model testing?

Answer 25

Policy outcomes should be different in areas with and without reservations for women!

Question 26

Main finding of the paper?

Answer 26

Elected women under reservation policy implemented more women-directed policies

Question 27

2 general conclusions from the paper?

Answer 27

1) Identity of a candidate MATTERS

2) Reservation policy can be efficient!