In 2014, US researcher David Blanchett coined the term 'retirement spending smile'1, a concept that has since become embedded in mainstream retirement incomes thinking around the world.
The spending smile derives its name from the 'U-shape' curve that emerges when retiree spending is plotted against age – the simplistic explanation being that spending peaks in the early years (when it is focused on travel and home renovations and living the 'best life'), trends down in the middle years, and then ticks back up in later life as health costs rise.
More recently however – and especially relevant in Australia – experts who agree with the decline part of the smile are starting to question the evidence around the uptick. One of those experts is Blanchett himself.
Blanchett's seminal 2014 analysis was based on US data, where retirees must plan for a notoriously expensive health care system. Fresh research2 by Blanchett himself (“How Spending Evolves in Retirement: A Smile, a Smirk, or Something Else?”) has raised the possibility that the later life uptick may be less pronounced in countries with state-funded health and aged care systems such as Australia, although he stops short of testing this directly.
The smile, in other words, may look more like a smirk in these countries, and portfolios and plans built around the smile pattern may - by overlooking the psychological drivers of spending behaviours – be making retirees more frugal and conservative than their resources warrant.
What Australian data shows
Australian evidence from a range of sources supports this downward spending trajectory.
Milliman3 estimated that the median retired couple's expenditure falls by more than one-third (36.7%) as they move from their peak spending years in early retirement (65 to 69 years of age) and into older age (85 years and beyond), with the decline accelerating sharply once retirees pass 80.
The Grattan Institute's analysis⁴ also found no evidence of a late-life uptick. Drawing on the ABS household expenditure and bank transaction data covering more than 300,000 Australian retirees, Grattan finds spending slows from around age 70 and falls rapidly after 80, driven mainly by lower spending on transport, recreation, food and furnishings as retirees pay off their mortgages, drop work-related costs and become less physically able to travel and eat out.
Perhaps the more important question, though, is not whether retirees spend less as they age, but whether they are spending less than they safely could.
Are retirees underspending or overspending?
This question is one of the most contested in retirement income discussions, as illustrated by the contrasting conclusions reached by two recent influential studies.
Grattan Institute's 2025 Simpler Super research5 estimates around half of account-based pension holders draw only the legislated minimum, while the Super Members Council's Busting a Myth release6 cites newer data showing most retirees now withdraw above it, particularly those with smaller balances, where 81 per cent exceed it.
Both, though, measure withdrawals against a legislative floor, not against what a portfolio could sustainably support, and as a result, neither answer the main question this article asks.
Federal government data7 from 2015 found less than half of all pensioners drew down on their assets at all, with more than 40 per cent net savers. Grattan's analysis8 of the same cohort over time tells a similar story: the wealthiest fifth of retirees saw their super balance grow by 58 per cent in real terms between their early 60s and late 70s, while their overall net financial wealth, including savings held outside super, rose by 12 per cent. Middle-income retirees saw their total financial wealth rise by around 10 per cent as well, even though their super balances fell. Grattan excluded the value of the family home when measuring this wealth growth.
The role of confidence
Part of the explanation for the underspending phenomenon may lie in how retirees mentally sort their own money.
Blanchett and Finke's 2025 research9, tracking how US retirees actually fund their spending, found that around 85% of available lifetime income – including pensions, annuities, and Social Security retirement payments – gets spent each year, compared to only about half of wages and capital income. Spending from savings is lower again, with withdrawal rates for 65-year-old couples averaging just 2%, only around half the commonly cited 4% rule. Put another way, retirees readily spend money that arrives as income and hold back on capital they have to actively decide to draw down, treating a balance in an account as something to protect, and a regular payment as something to use.
Blanchett's most recent (2026) analysis10 quantifies this caution. He compared actual retiree spending against a “funded ratio”, a measure of what each retiree could safely afford to spend given their assets and expected income. A ratio above 1.0 means a retiree already has enough to sustain their current spending indefinitely, without cutting back. Yet retirees at the 1–1.49 funding ratio still cut real spending by 3.1% a year, and even those with a ratio of 1.5–1.99 cut back by 1.2% a year. Only once assets reached double what was actually needed (a ratio of 2 and over) did spending see any growth, and even then, by just 1.1% a year in real terms. In other words, retirees with no financial need for caution seemingly keep behaving cautiously anyway.
This is the same pattern we identified as Chapter One hesitancy in our Two-Chapter Retirement research11 – clients who have the money, but not the confidence to spend, condemning them to a sub-optimal retirement lifestyle in many cases.
From insight to action
Clearly there is a perception gap to be closed, between what a retiree can spend and what they believe they can spend. This is not a job for willpower; it’s a job for the adviser – to instil more confidence through the right guidance, and through products that solve the right problems.
Blanchett and Finke's finding, that retirees spend income far more readily than capital, offers us a clue. If a balance in an account gets protected while a regular payment gets used, then converting more of a client’s retirement savings into a guaranteed income stream should create more permission, and more confidence, to spend.
The obstacle has traditionally been the belief that securing guaranteed income means giving up flexibility and access. Traditional annuities may offer certainty, but that certainty typically comes at the cost of the liquidity and control that irreversibility-averse clients are reluctant to surrender. Account-based pensions on the other hand offer flexibility but no certainty, leaving clients to effectively self-insure against longevity risk by spending more cautiously than required.
Income layering strategies tackle this conundrum head on, by treating guaranteed income as just one layer within a broader strategy, effectively allowing clients to ‘diversify’ the amount of commitment they are required to give. Products like Allianz Guaranteed Income for Life (AGILE) are built for exactly this scenario, allowing clients to calibrate how much of their income they choose to guarantee while retaining flexible access to capital if circumstances change. Rather than forcing a choice between certainty and control, this approach secures the retirement clients can't yet see (Chapter Two), so they don't need to second-guess the one they can (Chapter One).
For further reading, download our Two-Chapter Retirement paper to discover:
- why even financially well-prepared clients default to minimum drawdowns, delay key decisions and leave money on the table
- how the fear of irreversible decisions – not longevity or market risk – is the greatest emerging threat to retirement outcomes;
- what a new framework for retirement planning means for how you structure conversations, sequence decisions and build lasting client confidence.