Research

Abstract:  This paper explores the effect of the stress from national leadership on leaders' facial aging, mortality and decision-making. I use a one-to-one matched sample of election winners and the opponents they defeated. I measure apparent age directly from photographs with two machine-learning models, and define facial aging as the gap between predicted apparent age and true age. Comparing winners and losers before and after elections in a difference-in-differences design, I find that winning office accelerates apparent aging, and that the effect emerges most strongly after leaders leave office: during tenure winners look about 1.1 years younger than matched losers, but about 3.1 years older once they step down. Two forces appear to combine. Office-related stress and its cumulative health burden drive the aging that surfaces after leaders leave, while the youthfulness observed during tenure largely reflects masking. Consistent with this, winners’ hair greying is suppressed in office and jumps sharply at stepdown, and removing hair from the age models roughly halves the in-office effect. A larger apparent-age gap is also associated with lower approval the following year.