Nonseparable Preferences in Public Opinion

公众舆论中不可分割的偏好

• 批准号: 9617679
• 负责人: Dean Lacy
• 金额: --
• 依托单位: Ohio State University Research Foundation -DO NOT USE
• 依托单位国家: 美国
• 项目类别: Standard Grant
• 财政年份: 1997
• 资助国家: 美国
• 起止时间: 1997-05-01 至 1999-04-30
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=9617679&HistoricalAwards=false
• 关键词: Nonseparable; Preferences; Public; Opinion

9617679 Lacy A substantial body of research on mass political behavior assumes that individuals have "separable preferences." This assumption requires that a person's preference on one issue does not depend on preference on other issues. The assumption is so prevalent that public opinion surveys typically ask a respondent her opinion on one issue after another without links across issues. What if a persons wants the federal government to prohibit abortions, but only if the government also increases spending on sexual education, public day care, and aid to families with dependent children? Or what if a person wants a tax cut only if spending cuts come from the defense budget, not from social welfare? If a person's opinion on one issue depends on the outcome of or the options available on another issue, her preferences are "nonseparable." This project investigates the extent and impact of nonseparable preferences in public opinion and voting behavior. Small-group experiments assist the principal investigator in formulating new survey questions to determine the impact nonseparable preferences have on survey responses. The investigator uses new survey questions in a random sample telephone survey of 400 residents of Columbus, Ohio. Data from the survey enables social scientists to judge whether nonseparable preferences explain some common empirical problems in public opinion, including question-order effects, over-time instability in preferences, and the apparently minimal impact of issues on voter choice in elections. With new survey instruments available to capture nonseparable preferences, political scientists are better equipped to investigate voting behavior and public opinion, and policymakers have a clearer view of what people actually want on important public policy issues. ***

9617679花边大量关于大众政治行为的研究假设个人有“可分离的偏好”。这个假设要求一个人对一个问题的偏好不依赖于对其他问题的偏好。这种假设是如此普遍，以至于民意调查通常会问受访者对一个又一个问题的看法，而不涉及问题之间的联系。如果一个人希望联邦政府禁止堕胎，但前提是政府同时增加在性教育、公共日托和对有独立儿童的家庭的援助方面的支出，该怎么办？或者，如果一个人只希望削减国防预算，而不是社会福利支出，那该怎么办？如果一个人对一个问题的看法取决于另一个问题的结果或可用选项，那么她的偏好是“不可分离的”。这个项目调查了民意和投票行为中不可分离偏好的程度和影响。小组实验协助主要研究者制定新的调查问题，以确定不可分偏好对调查反应的影响。调查人员对俄亥俄州哥伦布市的400名居民进行了随机抽样电话调查，使用了新的调查问题。来自调查的数据使社会科学家能够判断不可分离的偏好是否解释了民意中一些常见的经验问题，包括问题顺序效应、偏好的长期不稳定性，以及选举中问题对选民选择的明显最小影响。有了新的调查工具来捕捉不可分离的偏好，政治学家就能更好地调查投票行为和公众舆论，政策制定者也能更清楚地了解人们在重要的公共政策问题上真正想要什么。＊＊＊