Ambiguous Correlation

Many decisions are made in environments where outcomes are determined by the realization of multiple random events. A decision maker may be uncertain how these events are related. We identify and experimentally substantiate behavior that intuitively re‡ects a lack of con…dence in their joint distribution. Our …ndings suggest a dimension of ambiguity which is di¤erent from that in the classical distinction between risk and “Knightian uncertainty.” Boston University, [email protected] and University of British Columbia, [email protected]. We are grateful for discussions with Kyoungwon Seo and Chew Soo Hong, for detailed comments from Aurelien Baillon and especially Peter Wakker, for suggestions by two referees and an editor, and for comments from audiences in several conferences and workshops. We are grateful also for the …nancial support of the National Science Foundation award SES-1216339 (Epstein) and the Social Science and Humanities Research Council of Canada grant 435-2012-0577 (Halevy). Anujit Chakraborty, David Freeman, Terri Kneeland and students from UBC’s ECON490 classes of 2013, 2014 and 2015 (Experimental Economics section) provided excellent research assistance. Earlier versions of this paper were titled “No Two Experiments Are Identical.” First version: February 2014.