A strong intergenerational divide in voting behaviour has become a frequently reported stylized fact in post-referendum analyses, raising the question of whether myopia of elderly voters is a cause for concern in ageing societies. However, identifying the origins of a generation gap is empirically challenging because age and birth cohort are collinear at any given point in time. Building on a previous study, we summarize how this problem can be overcome. We discuss an unconstrained rank regression approach that allows estimating the causal ageing effect on political attitudes conditional on arbitrary cohort effects in a flexible manner. As they age, voters of all cohorts become more politically conservative and less supportive of reform projects aimed at protecting the environment, benefitting the working population, and redistributing wealth from the rich to the disadvantaged. As population ageing progresses, forward-thinking reform projects will have declining chances of winning majorities in direct votes.