Year: 2019
Type: dissertation
Abstract: CampaigningEach candidate has a fixed amount of campaign resources.A candidate's objective is to maximize her/his expected number of obtained votes.11 To this end, each candidate decides whether to focus on policy or ability in campaigning.In other words, each candidate determines the fraction of campaigns s/he spends on policy (e.g., the fraction of time devoted to campaigning policy in speeches) 𝐶 𝑘 and the fraction of campaigns s/he spends on ability 1 -𝐶 𝑘 .In particular, each candidate chooses 𝐶 𝑘 ∈ {𝐶 𝐻 , 𝐶 𝐿 }, where 0 < 𝐶 𝐿 < 𝐶 𝐻 < 1. 12 𝐶 𝑘 = 𝐶 𝐻 (𝐶 𝐿 ) represents that candidate 𝑘 focuses on policy (ability).Campaigns on policy persuade a naïve voter that the voter's ideal policy is the candidate's policy.The challenger's (incumbent's) campaigns on ability persuade a naïve voter that the incumbent's ability is low (high).Notice that the challenger cannot persuade voters that her/his ability is high because the ability has not been realized.The fraction of voters a candidate persuades is given by 𝑝 𝑘 (𝐶 𝑘 , 𝐶 -𝑘 ) ∈ (0, 1) and 𝑛 𝑘 (1 -𝐶 𝑘 , 1 -𝐶 𝑘 ) ∈ (0, 1).Candidate 𝑘 persuades 𝑝 𝑘 (𝐶 𝑘 , 𝐶 -𝑘 ) × 100 percent of naïve voters that candidate 𝑘 's policy is good.Further, candidate 𝐴 (𝐵) persuades 𝑛 𝐴 (1 -𝐶 𝐴 , 1 -𝐶 𝐵 ) × 100 (𝑛 𝐵 (1 -𝐶 𝐵 , 1 -𝐶 𝐴 ) × 100) percent of naïve voters that the incumbent's ability is high (low).Therefore, the challenger's campaigns on ability are negative campaigns on the incumbent's ability.We assume that 𝑝 𝑘 and 𝑛 𝑘 satisfy the following regular conditions. ASSUMPTION 1i.(Symmetry) 𝑝 𝑘 (𝑥, 𝑦) = 𝑝 -𝑘 (𝑥, 𝑦) = 𝑝(𝑥, 𝑦) and 𝑛 𝑘 (1 -𝑥, 1 -𝑦) = 𝑛 -𝑘 (1 -𝑥, 1 -𝑦) = 𝑛(1 -𝑥, 1 -𝑦) for 𝑘 ∈ {𝐴, 𝐵}, and 𝑥, 𝑦 ∈ {𝐶 𝐻 , 𝐶 𝐿 }.ii. (Full persuasion)For 𝑘 ∈ {𝐴, 𝐵} , 𝑝 𝑘 (𝐶 𝑘 , 𝐶 -𝑘 ) + 𝑝 -𝑘 (𝐶 -𝑘 , 𝐶 𝑘 ) = 1 and 𝑛 𝑘 (1 -and 𝑝 𝑘 (𝐶 𝑘 , 𝐶 𝐻 ) < 𝑝 𝑘 (𝐶 𝑘 , 𝐶 𝐿 ) and 𝑛 𝑘 (1 -𝐶 𝑘 , 1 -𝐶 𝐻 ) > 𝑝 𝑘 (1 -𝐶 𝑘 , 1 -𝐶 𝐿 ) for 𝐶 𝑘 ∈ {𝐶 𝐻 , 𝐶 𝐿 }.(i) The effect of campaigns on naïve voters is the same across candidates, (ii) all the naïve 11 The vote share affects post-election policy making though candidates win the election with half of all votes.Alternatively, maximizing the winning probability under aggregate uncertainty about voters' preferences would be qualitatively similar to maximizing the vote share in our framework.12 In reality, voters could not distinguish a small difference in 𝐶 𝑘 (i.e., they would observe only whether a candidate focuses on policy or ability).Thus, the binary choice is meaningful.
DOI: 10.15083/0002001793
Author: Kasamatsu, Satoshi
Institution: The University of Tokyo
Language: English
Subfield: Economics and Econometrics
Domain: Social Sciences
Open Access status: green
OpenAlex ID: W7207736077