A better understanding of what drives long term care (LTC) expenditures is important for all countries with aging populations. We employ unique new data sources to analyze the determinants of LTC spending in the Netherlands. First, we use two-part models, to analyze institutional LTC and homecare expenditures for the entire 55+ population, conditioning not only on age, sex, time-to-death (TTD), but also on cause-of-death and co-residence status. These have profound effects. Those living alone, as well as those who deceased from diabetes, mental illness, stroke, diseases of the respiratory or digestive system have higher LTC expenditures, while a neoplasm death resulted in lower expenditures. Secondly, we examine homecare expenditures among a sample of non-institutionalized individuals conditioning, additionally, on morbidity and disability. Finally, we reconsider the roles of age and TTD, when controlling for the most important determinants of LTC use - morbidity, disability and co-residence - andillustrate their relevance for forecasting LTC expenditures. Our analysis reveals that TTD is not a predictor of homecare expenditures when disability is controlled for, while age and co-residence are. We therefore conclude that it is time to drop time-to-death from LTC expenditure models as it merely acts as a proxy for disability status.
