“…Mortality risk models were estimated using data on individuals' survival over the period 1995-2018 and data on SSPs for their survival beliefs. These models control for individual characteristics such as gender, socioeconomic status, and health behavior, which can be correlated with PLE-knowledge and related to survival or survival belief (e.g., Cutler, Deaton, andLleras-Muney 2006, 2011;Delavande and Rohwedder 2011;Grevenbrock et al 2020;Hurd and McGarry 2002;Kalwij, Alessie, and Knoef 2013;. The associations between the covariates and annual mortality belief (see Table 1, right columns) are in line with previous findings, especially in terms of health characteristics, such as having bad health, chronic illnesses, smoking and drinking alcohol (e.g., Teppa 2012;Bissonnette, Hurd, and Michaud 2017).…”