%0 Report %D 2008 %T Wealth Change and Active Saving at Older Ages %A Michael D Hurd %A Susann Rohwedder %K Consumption and Savings %K Event History/Life Cycle %K Net Worth and Assets %X According to the simple lifecycle model single persons are predicted to decumulate assets at advanced age, when mortality risk is high to reduce the risk of dying with substantial wealth. Empirically it has been difficult to show this prediction in micro data. In this paper we discuss the most common limitations in existing data. We provide empirical evidence of dissaving at older ages by single persons using the unusually rich data from the Health and Retirement Study. We present lifecycle patterns of dissaving based on three very different kinds of data: those that are derived from wealth change, those derived from measures of active saving defined as disposable income minus consumption and those that are derived from model simulation. Based on wealth change we find evidence of dissaving for singles and limited evidence for couples: couples preserve wealth longer to provide for the surviving spouse. However, rates of active saving imply much smaller wealth decumulation for singles and no decumulation at all for couples. Decumulation based on model simulation lies between the two. We suspect that the discrepancies are partly due to the treatment of taxes and partly due to consumption being under measured. %I RAND and Netspar %G eng %4 life Cycle/Asset accumulation/Dissaving/wealth %$ 62660