’Mark My Words’: Measuring Personality of Global Leaders

• In Event: Measurement and Method

Abstract

This essay introduces an alternative method of measuring personality traits of global leaders from their speeches using recent computational advances. It enables scholars to make wide-ranging cross-national comparisons that have not been possible hitherto, adding value to the extant works, which are mostly US-centric, rely on expert ratings, and are designed for specific tasks only. Scholars firmly believe that socially relevant and salient personal characteristics are encoded in the natural language. Based on this fundamental idea, this work uses a machine learning algorithm to estimate personality scores on the framework of cross-culturally validated ‘Big Five’ that is described in terms of five dispositional traits–openness to experience, conscientiousness, extraversion, agreeableness, and emotional stability. The availability of personality estimates of the world leaders from the speeches they delivered at the annual jamboree of the United Nations General Assembly from 1946 to 2022 opens several research avenues to test the substantive implication of leader personality on her policy decisions. I also validate my findings with spontaneous forms of speeches that are believed to be more representative of a speaker’s personality, giving confidence to the scholars in using widely available public statements to perform the tasks at hand. In an application to political behavior, I demonstrate that my findings provide evidence for the established narratives about the personality of populist leaders.

Author

• Mahesh Acharya, The University of Mississippi