In this paper we test the 'red herring' hypothesis for expenditures on long-term care. The main contribution of this paper is that we assess the 'red herring' hypothesis using an aggregated measure that allows us to control for entering the final period of life on the individual level. In addition we implement a model that allows for age specific time-to-death (TTD) effects on Long Term Care. We also account for the problem that mortality, and therefore TTD, are themselves influenced by care expenditure. For our analysis we use administrative data from the Swedish statistical office. In contrast to many previous empirical studies, we are able to use the entire population for estimation instead of a sample. Our identification strategy is based on fixed effects estimation and the instrumental variable approach to achieve exogenous variation in TTD. Our results indicate that although time-to-death is a relevant indicator for long term care, age itself seems to be much more important for the projection of long-term care expenditure.