Desktop version

Home arrow Business & Finance

  • Increase font
  • Decrease font


<<   CONTENTS   >>

Related studies

Electoral promises as a commitment device

Few previous studies consider platforms as a partial commitment device by introducing the cost of betrayal as in Chapter 2. Austen-Smith and Banks (1989) consider a two-period game based on a retrospective voting model in which the probability of winning in the next election decreases if office-motivated candidates betray the platform. Grossman and Helpman (2005, 2008) develop a legislative model in which office-motivated parties announce platforms before an election and the victorious legislators who are policy- motivated decide policy. If legislators betray the party platform, the party punishes them. Hummel (2010) supposes that there exist costs when policy announcements are different between primary and general elections. On the contrary, my model is based on the prospective voting and two-candidate competition models without primaries and assumes that policy-motivated candidates decide on both a platform and a policy.

Note that Austen-Smith and Banks (1989) consider only a decrease in the probability of winning as the cost of betrayal, while Grossman and Helpman (2005,2008) consider only a party’s discipline as the cost of betrayal. However, the cost of betrayal in this book also includes many types of costs such as a decrease in approval ratings and a negotiation cost with Congress; therefore, I include these in the current term as the cost of betrayal.

Electoral promises as a signal

Banks (1990) and Callander and Wilkie (2007) consider campaign platforms as signals about the candidate’s policy preference by introducing the cost of betrayal as in Chapter 3. However, there are two important differences between my models and theirs. First, in their work, candidates automatically implement their own ideal policies after winning an election. However, if there is a cost of betrayal, a rational candidate would wish to adjust the policy to be implemented to reduce that cost after the election. Second, Banks (1990) and Callander and Wilkie (2007) consider that candidates care about policy only when they win - their utility is set to zero when they lose regardless of the policy implemented by their opponent. However, policy-motivated candidates should care about policy when they lose.

I relax these assumptions and make more reasonable ones by examining rational choices about a policy to be implemented and candidates who care about policy regardless of the election results. These two differences are critical to obtaining my results. For example, Chapter 2 shows that some candidates have an incentive to run even though they know with certainty that they will lose an election because it induces the opponent to approach the median policy, and thus the loser’s ideal policy. It is impossible to obtain this implication by assuming the above two assumptions. First, if candidates implement their own ideal policies automatically, it is impossible to induce the opponent to approach the median policy. Thus, for the loser, there is no way to change the opponent’s policy. Second, if a candidate does not care about policy when he/she loses, the loser does not have an incentive to change the opponent’s policy because his/her utility is zero in all cases. Thus, these two differences provide a way and an incentive to induce the opponent to approach the median policy.

Banks (1990) and Callander and Wilkie (2007) show that a moderate type may defeat an extreme type when asymmetric information about each candidate’s ideal policies exists. On the contrary, the results in Chapter 3 contradict those in the aforementioned studies. That is, an extreme type has a higher probability of winning than a moderate type. It is also impossible to gain this opposite result by assuming the above two assumptions. First, if candidates implement their own ideal policies automatically, an extreme candidate will lose against a moderate type when he/she reveals his/her type to voters. Thus, a separating extreme type cannot obtain a higher probability of winning in a semi-separating equilibrium. Second, if a candidate does not care about policy when he/ she loses, an extreme type will not have as strong an incentive to prevent his/her opponent from winning. Therefore, under the assumptions of Banks (1990) and Callander and Wilkie (2007), an extreme type does not have either a method or an incentive to win against a moderate type.

In Huang (2010), candidates strategically choose both a platform and an implemented policy but do not care about policy when they lose. Huang (2010) also supposes sufficiently large benefits from holding office and shows that candidates cluster around or at the median policy. On the contrary, my models in Chapters 2 and 3 suppose that candidates care about policy even after losing and does not consider the benefits from holding office as large. Other research also considers that an electoral promise can work as a signal of the functioning of the economy (Schulz, 1996), the candidate’s degree of honesty (Kartik and McAfee, 2007), and political motivation (Callander, 2008).

 
<<   CONTENTS   >>

Related topics