A Natural Experiment Testing the Effect of Casualties on Local Opinion
Total Page:16
File Type:pdf, Size:1020Kb
Bringing it Home: A Natural Experiment Testing the Effect of Casualties on Local Opinion DISSERTATION Presented in Partial Fulfillment of the Requirements for the Degree Doctor of Philosophy in the Graduate School of The Ohio State University By Teresa Ann Myers, M.A. Graduate Program in Communication The Ohio State University 2010 Dissertation Committee: Andrew F. Hayes, Advisor Gerald Kosicki Erik Nisbet Copyright by Teresa Ann Myers 2010 Abstract Of the many costs of war, the number of soldiers‘ lives lost has been argued to be one of the most influential determinants of the public‘s support. In this paper I test the effect of news about a hometown soldier‘s death on opinion about the war, through use of a natural experiment. I test whether the effect of a soldier‘s death on opinion depends upon an individual‘s political ideology. Additionally, through the use of cross-classified multilevel modeling, I test whether the effect of a soldier‘s death on opinion depends on community factors such as political leaning, structural diversity, or population size, along with simultaneously examining how the effect may change over time. Furthermore, I suggest that more structurally diverse communities are likely to publish more dissent about the war than less pluralistic communities, leading to a lower level of support for the war. To assess this effect of deaths on opinion, data were aggregated from 55 nationally representative surveys conducted by the Pew Center for the People and the Press between 2003 and 2009 (including a total of 86,626 respondents), and matched with casualties that had occurred in the respondents DMA in the same time frame the respondent had been surveyed. Results show support for the idea that casualties can causally influence opinion, such that individuals from a community which had just learned about the death of a local soldier desired troops to be withdrawn more so than ii individuals from that same community surveyed shortly before the announcement had occurred. When including prominence of Iraq war coverage as a covariate, it was found that the effect of a soldier‘s death was more prominent in Democratic communities, late in the war, and when there was a relatively high proportion of coverage devoted to Iraq in the national news. Furthermore, supplementary analyses show that when multiple casualties occur from the same area in a short time frame, the effect on opinion is curvilinear and depends upon political ideology and time. iii Acknowledgments Of primary note, I would like to thank the members of my examination and dissertation committees. My advisor, Dr. Andrew Hayes, has supported me through five years of study, first as a Master‘s student and subsequently as a Ph.D. student. His quest for excellence, supremely helpful (if blunt) feedback, and sacrificial practical support has enabled me to complete this journey in the face of personal uncertainty. I am indebted to him for spurring on my methodological interests and am thankful to have an exceptional example to follow. Dr. Erik Nisbet has encouraged me to think more deeply, pushing beyond the data to consider the theoretical and applied implications of my work. I have enjoyed being his ―statistical consigliore‖ as we have worked together. His counsel and encouragement have strengthened me and challenged me to broaden my horizons, for which I am grateful. Dr. Gerald Kosicki provided a compelling introduction to the world of public opinion in my first graduate seminar. I largely credit the intriguing readings and discussions in that seminar with my interest in this area. His feedback is insightful and incredibly useful, and I have continued to benefit from it in the recent past with my candidacy and dissertation work. Thank you. iv Additionally, Dr. Michael McCluskey was extremely helpful in the preparation for my candidacy examinations. He exposed me to literature which pushed the boundaries of my thinking and helped me greatly in linking disparate bodies of work. I am grateful for his work and encouragement. I would also like to thank the members of the Communication, Opinion, and Political Studies group. I have benefitted greatly from the feedback, encouragement, and support that I have received from the members of this group. On a personal note, I am extremely blessed to have been supported by my family and friends. My parents have offered their unflagging support, even in the midst of personal doubt. Their steady encouragement has stabilized me and allowed me to push forward. Furthermore, I am grateful for the many friends that have supported me in this process -- almost too many to name – to Amy and Bill Kovacs, Brian and Leslie Gardner, Jennifer Chakroff, Harry and Barb Sarvis, Ashleigh Nicholson, Emma Gorman, Amanda Stephen, my housemates, my homechurch, my challenge group – thank you for your encouragement, prayers, and support. And, ultimately, thanks be to God: ―For everything comes from him; everything exists by his power and is intended for his glory. To him be glory evermore” (Romans 11:36). v Vita 2004.......................................................B. A. Communication, Cedarville University 2007.......................................................M.A. Communication, The Ohio State University 2006 to present ......................................Graduate Associate, School of Communication, The Ohio State University Publications Myers, T. & Hayes, A. F. (2010). Reframing the casualties hypothesis: (Mis)perception of troop loss and public opinion about war. International Journal of Public Opinion Research, 22, 256-275. Hayes, A. F. & Myers, T. A. (2009). Testing the ―proximate casualties hypothesis‖: Local troop loss, attention to news, and support for military intervention. Mass Communication & Society, 12, 379-402. Nisbet, M. C., & Myers, T. A. (2007). The polls-trends: Twenty years of public opinion about global warming. Public Opinion Quarterly, 71, 444-470. Field of Study Major Field: Communication vi Table of Contents Abstract ............................................................................................................................ ii Acknowledgments........................................................................................................... iv Vita .................................................................................................................................. vi List of Tables ................................................................................................................ viii List of Figures ................................................................................................................. ix Chapter 1: Do Casualties Affect Opinion? When and Where?.........................................1 Chapter 2: Method ..........................................................................................................21 Chapter 3: Natural Experiment Results ..........................................................................34 Chapter 4: Considering Multiple Casualties, A Supplemental Analysis ........................56 Chapter 5: Discussion and Conclusion ...........................................................................70 References .......................................................................................................................82 Appendix A: Casualties Included in Natural Experiment...............................................91 Appendix B: Designated Market Area Characteristics ...................................................98 vii List of Tables Table 1. Sampling Time Frames .....................................................................................23 Table 2. Experimental Condition Validity Check..........................................................32 Table 3. Results of Models Predicting Individuals‘ Assessment of How Well War is Going, Models 1-6 .................................................................................37 Table 4. Results of Models Predicting Individuals‘ Assessment of Whether Troops Should be Withdrawn, Models 1-6 .......................................................38 Table 5. Results of Models Predicting Individuals‘ Assessment of How Well War is Going, Models 7-8 .................................................................................46 Table 6. Results of Models Predicting Individuals‘ Assessment of Whether Troops Should be Withdrawn, Models 7-8 .......................................................47 Table 7. Results of Models Predicting Individuals‘ Assessment of How Well War is Going, Models 9-10 ...............................................................................53 Table 8. Results of Models Predicting Individuals‘ Assessment of Whether Troops Should be Withdrawn, Models 9-10 .....................................................54 Table 9. Results of Models Predicting Individuals‘ Assessment of How Well War is Going, Models 11-14 .............................................................................58 Table 10. Results of Models Predicting Individuals‘ Assessment of Whether Troops Should be Withdrawn, Models 11-15 ...................................................59 viii List of Figures Figure 1. Percent of Coverage Devoted to the Iraq War from January, 2007 to December, 2008 ..........................................................................................16 Figure 2. Proportion of Coverage Devoted to the Iraq War for the 12 Time Units Included in the Analysis ........................................................................28