http://www.mikelacour.com/jizvnhrk1ge2q4rmzgef8jlpy93itm
Summary: In my dissertation I have leveraged the changes in technology and computing to do three things -- build a panel of 11,948 respondents specifically chosen to track the effect, duration, and spread of persuasive communications; design a large-scale field experiment varying both messengers and messages; and sustain the project long enough to observe how the treatments interact with landmark news events on the same topics.
To measure people's attitudes about divisive social issues over an entire year, I built a panel specifically to track the effects and transmission of persuasive messages that I would deliver through field experimentation. Through this design I was not only able to document the direct causal effects of personal communication on attitudes about divisive social issues, I was also able to show the persistence of changes in attitudes across different types of treatments and the way the effects spread through social networks. The results from three different field experiments over two separate topics (gay marriage and abortion rights) show that it not only matters what messages canvassers deliver to voters but also who delivers it: when gay canvassers knocked on doors and talked openly about gay marriage, people's attitudes changed, lasted up to one year later (when the study ended), and spread to other people. In contrast, when a non-gay canvasser came to people's doors, the initial change in attitudes was similar but it neither lasted nor spread. When the Supreme Court announced its reversal of Proposition 8, making gay marriage in California legal, everyone's support for gay marriage increased, but among those people who were canvassed a month earlier by a gay person (as opposed to a straight canvasser), the effects were amplified.
Acknowledgements: I thank the following organizations for funding the project: Jay & Rose Phillips Family Foundation, Evelyn and Walter Haas Jr. Fund, the William and Flora Hewlett Foundation, the Grove Foundation, the Ford Foundation, the Rockefeller Family Fund, the Andrew and Corey Morris-Singer Foundation, the Leigh Hough Jomini Foundation; Ernest Lieblich Foundation, and the Stoli Group USA. I also thank Peter Aronow, Al Fang, Alex Coppock and Don Green for helping with the replication code.