Voting behavior and political campaigns
MetadataShow full item record
This dissertation tries to answer whether electoral campaigning can affect voting behavior. I estimate the effects of the two marketing skills, canvassing and network marketing, on voters by using two datasets from the United Kingdom and Taiwan, and a multinomial probit model, which scholars usually apply to consumers' behavior. However, because there are various social interactions among voters, a direct application of a multinomial probit model on voting behavior will cause a biased estimation. Hence, in the first part, I examine different kinds of social interactions among voters, and propose a method to deal with these effects. By this method, I find that 5.23 percent of the voters in England did not vote for their most preferred candidates because of their expectations on other voters' behavior. In the second part, after properly controlling the social interactions, I estimate the effects of different types of canvassing on the voters in England. I found that doorstep canvassing done by challengers can affect voters, while that done by incumbents has no significant effect on voting behavior. The difference between incumbents and challengers suggests that canvassing affects voters by offering them new information, so incumbents' canvassing has no effect because voters have known incumbents well. Moreover, telephone canvassing has no significant effect on voting regardless of who does this. This implies that only the information channeled by doorstep canvassing is crucial in household voting decisions. In the third part, I examine the effect of a special marketing skill used by political machines in Taiwan: network marketing. I find that candidates from machines have a great capacity for mobilizing voters' ward or village heads, which are typically opinion leaders in voters' communities. I also find that the solicitation of voters' ward or village heads can affect voters' behavior. Again, the solicitation for a challenger is more powerful than that for an incumbent. It implies that the solicitation is also an information channel. Yet, the solicitation for an incumbent can still affect voters, which is not like the case of canvassing in England.
Thesis (Ph.D.)--Boston UniversityPLEASE NOTE: Boston University Libraries did not receive an Authorization To Manage form for this thesis or dissertation. It is therefore not openly accessible, though it may be available by request. If you are the author or principal advisor of this work and would like to request open access for it, please contact us at email@example.com. Thank you.