By Professor Constantine Boussalis, Dr Mirya Holman, and Dr Travis Coan, published July 2026.
Abstract:
Political elites demonstrate that they belong in power through a wide set of actions, including communicating that they can express emotions like happiness. In this paper, we argue that happiness is a central tool of political communication and examine how and when politicians present as happy through facial images. We argue that women in power will express happiness more frequently as it is a dominance display that is acceptable for both women and leaders. We evaluate over 460,000 Facebook posts by Members of Congress (MOCs), extracting images and then faces from these posts, and identifying images of MOCs out of those faces. We then use a deep learning-based face analysis framework to detect happiness in facial emotions and combine this with affiliative language from automated text analysis. Women MOCs express 15% more happiness than men across all images but are not more positive in text. We use Facebook’s unique emotional reactions to show that, despite women’s higher level of happiness, this is not rewarded with more engagement by the audience. Through these evaluations, we provide evidence that gender shapes the political communication choices of political elites and the importance of images and happiness in communication.