Date of Award


Document Type

Open Access Dissertation

Degree Name

Doctor of Philosophy (PhD)



First Advisor

Ellen A. Bruce

Second Advisor

Frank W. Porell

Third Advisor

Christian E. Weller


This paper examined patterns of decumulation and the role that health events and marital disruption play in forming those patterns. Study data were drawn from six biennial waves of the HRS (1998 - 2008), and merged RAND HRS data files for the period 1998 - 2008. The a priori expectation was that there will be variation in drawdown strategies households employ.

Findings suggest that patterns of dissaving are heterogeneous. The five most prevalent patterns were discussed. Households predominantly transitioned between oversaving and overspending. Households are expected to have a goal of on target spending therefore the observed cycle's dissaving will influence the next cycle's draw down rate in an attempt to maintain a sustainable drawdown rate. Markov model results suggest that households do recalibrate their depletion rate as a function of their last depletion rate. This study hypothesized that the onset of a health condition or a spouse's admission to a nursing home would be associated with an excessive decumulation of assets. These hypotheses were unsupported by the research. Marital transitions as predictors of decumulation were only partially borne out by the results. Divorce was also expected to increase the likelihood of overspender however this relationship was not significant. Loss of spouse was associated with an increased likelihood of overspending.

One of the major contributions of this study is the identification of patterns of dissaving in retirement. Various life course, demographic and decumulation factor variables were determinants of these patterns. Overall results suggest that elders have a difficult time managing to an on target drawdown. This study concludes with a national decumulation policy directive outline.