Research

Dog Whistling

Abstract

(Link to draft.) News and media commentators frequently accuse political candidates of using "dog whistles": words and phrases that have secondary, usually nefarious meanings. Whistling targets attuned subsets of a population, attempting to convey information the remainder of the audience might dislike (or, at least, judge negatively during a campaign). Voter beliefs and knowledge are manipulated through the use of such coded speech, directly impacting the outcome of a political contest. I model candidates attempting to shift voter choices during a campaign through dog whistling. Results indicate that coded communication shifts voter support in a candidate's favor, but does so when its use is limited to select candidate types. Further, while the media may argue the content of the whistle as the important story, I find the presence of a whistle, independent of whether it reached all intended targets, is enough to increase a candidate's vote share.

Author:

Ian Palmer Cook, University of Pittsburgh

Dissertation, Formal Theory, Voting   :   May 23, 2013 9:02 PM  :  read more »


About the Author

Ian P. Cook is a Ph.D. student at the University of Pittsburgh, studying American Politics. He has also been a researcher with the RAND Corporation. No material on this site implies endorsement by either institution.