Using Austria as a case study, this paper demonstrates how capturing life course heterogeneity improves the accuracy and policy relevance of socio-demographic projections, and how considering this population heterogeneity impacts labour force dynamics and economic dependency ratios. We introduce and apply the microsimulation model microDEMS, focusing on education, migration background, health and labour market participation. Using administrative data, the model ensures longitudinal consistency of labour market careers, including insurance periods, and considers pension rules and reforms. Despite its level of detail, microDEMS is consistent with official demographic projections. To assess sensitivity, we create alternative scenarios that illustrate how different factors affect future labour force dynamics. The main result of our simulation analysis is the quantification of substantial mitigating effects of improvements in education and already adopted changes in pension legislation, which together reduce the impact of ageing on the economic dependency ratio by 55%.