A substantial body of international research has shown that household expenditure
on food and non-durables signicantly decreases at the time of retirement a finding
that is inconsistent with the standard life-cycle model of consumption if retirement is
an anticipated event. This fall in expenditure has become known as the `retirement-
consumption puzzle.' We analyze rich Australian panel data to assess the Australian
evidence on the puzzle. We nd strong evidence of a fall in expenditures on groceries,
food consumed at home and outside meals with retirement. The observed decline
in expenditure is explained by a subset of households experiencing an unanticipated
wealth shock, such as a major health event or long-term job loss, at the time of
retirement. This nding is corroborated by an analysis of alternative measures of
household well-being, including indicators of nancial hardship, and self-reported
nancial and life satisfaction. For the majority of households retirement is anticipated
and there is no decline in economic welfare at retirement. However, for an important
minority, retirement is `involuntary' and these households experience a marked decline
across all indicators of economic well-being.