“…A further potential explanation comes from comparing the results for the 36–55 age group, which are significant when we do not fully account for time‐invariant unobserved heterogeneity (PSM) and non‐significant when we do (FE, FEIS). This discrepancy suggests that the lack of significance in the FE models may be due to time‐invariant factors relevant for electoral participation, such as social origins (Jeannet, 2022; Plutzer, 2002), personality (Emmenegger et al., 2017b), and birth cohort (Alwin & Krosnick, 1991; Grasso et al., 2019; Neundorf & Niemi, 2014). Empirically, related panel data studies find that unemployment experiences do not affect social participation after age 54 (US, Brand & Burgard, 2008), and political participation after ages 35 (Germany, Emmenegger et al., 2017) and 39 (Sweden, Österman & Brännlund, 2023), albeit with a smaller magnitude for the latter paper.…”