Yarine Fawaz (CEMFI and Universidad Autonoma de Madrid)

Abstract: This study investigates the potential of response times (RT) as a predictor of cognitive decline, leveraging previously untapped data from the Survey of Health, Ageing, and Retirement in Europe (SHARE). Recorded automatically during Computer-Assisted Personal Interviews (CAPI), RTs offer a unique opportunity to study cognitive and decision-making processes without added respondent burden. We first validate RTs as meaningful predictors of cognitive decline, building on Sanders et al. (2025). Slower RTs – recorded automatically during standard cognitive modules – are strongly associated with lower baseline cognitive scores and greater subsequent decline, even after controlling for age, gender, and baseline cognition. RTs also predict deterioration across a range of health outcomes, including frailty, mental health, and mortality, underscoring their potential as early indicators of physiological aging. We then examine the predictive content of RTs for financial outcomes. Slower RTs are associated with subsequent wealth losses, above and beyond what is captured by standard cognitive measures and interviewer fixed effects. Replicating the findings of Mazzonna and Peracchi (2024) in a European context, we show that individuals who experience cognitive decline without being aware of it are particularly vulnerable to wealth decumulation. Importantly, RTs are strongly associated with being unaware of one’s own cognitive deterioration. This suggests that RTs can help identify individuals at risk of poor financial decision-making due to undiagnosed cognitive impairment. Finally, we document heterogeneity by gender and country, with stronger associations observed among male financial respondents, and in countries with higher room left for individual decision-making in wealth management. Taken together, our findings demonstrate that response times – routinely collected but often ignored – contain rich information about both cognitive and financial vulnerability at older ages.