**Principal Investigator(s):** **Elizabeth Simas** University of Houston Email: [ensimas@uh.edu][1] Home page: [https://www.uh.edu/class/political-science/faculty-and-staff/professors/simas/][2] **Sample size**: 2796 **Field period**: 07/14/2017-03/16/2018 **Abstract** Work on candidate positioning suggests that ambiguous statements help candidates by broadening their appeal and providing leeway to take different positions once in office. While experimental research has explored the electoral effects posited by the first, empirical testing of the second is lacking. As such, we offer a survey experiment that better tests whether ambiguity can shield a legislator from the backlash that comes from eventually taking a position with which voters disagree. We also test whether the effects of rhetoric vary across issue type. Specifically, we examine whether voters react differently when evaluating matters related to principles or core beliefs vs. matters that are more pragmatic (see Tavits 2007). **Hypotheses** H1: While ambiguity may be preferable to an outright flip-flop in many cases, the possibility of still violating voters’ expectations creates situations where it ultimately offers a candidate no improvement over simply expressing the intended position. H2: The effects of ambiguity will be contingent on issue type. On principled issues, voters are more likely to value the consistency of a politician who maintains a clearly-stated position; on more pragmatic issues, voters' greater tolerance for flexibilty should minimize if not wholly eliminate the differences between the various rhetorical strategies. **Experimental Manipulations** Subjects were exposed to information about a hypothetical candidate from their own party. Random assignment determined whether the subject was assigned to the transgender rights (i.e., principled) domain or the business incentives (i.e., pragmatic domain). Within each, subjects were shown either a clear or ambiguous statement from the candidate. Then later, subjects were told that the candidate either voted for or against a bill on that issue. **Key Dependent Variables** Subjects were asked to use a 0-10 scale to rate the candidate on likability, open-mindedness, integrity, and the extent to which he represents his constituents well. Subjects were also asked to indicate how they believed that the candidate would vote on a bill increasing regulations on the sale of guns. **Summary of Findings** Pooling across issues, we find that the candidate is always worst off when voting opposite a clearly stated position. But when comparing ambiguity to maintaining the same, clear position, there is only one significant difference (p<.05), as clarity garnered the candidate higher ratings (6.22 vs. 5.77) on representing his constituents well among those who agreed with the final vote. When separating the two domains, we find that our treatment effects are almost exclusively in the transgender rights domain. When considering statements and votes about a bathroom bill, subjects gave the worst ratings when the candidate changed his position. But in 6 of the 8 comparisons, we see no statistically significant differences between maintaining a clear vs. an ambiguous position regardless of the subject’s agreement with the final vote. The two cases where the boosts are significant (p<.05) do occur show that maintaining a clear position vs. taking an ambiguous position raises the candidates mean likability rating from 7.13 to 7.55 and increases perceptions of his representation from 6.44 to 7.16. In contrast, we observe almost no differences between the various rhetorical strategies in the business incentives domain. **References** Elizabeth N. Simas, Kerri Milita, and John Barry Ryan. Accepted. "Ambiguous Rhetoric and Legislative Accountability." *Journal of Politics.* [1]: mailto:ensimas@uh.edu [2]: https://www.uh.edu/class/political-science/faculty-and-staff/professors/simas/