Australians surviving to 65 have conditional life expectancy of roughly 85 (men) or 88 (women) — meaningfully above the at-birth average of 81. Life expectancy is a distribution: about 25% of women at 65 reach 95 or beyond. Planning to average life expectancy means half of retirees outlive their plan. For healthy retirees, planning to age 90-95 is correct. For couples, the relevant horizon is the longer-lived partner's death.
For Australian retirees and pre-retirees, life expectancy is one of the most important inputs into retirement planning. It determines how long retirement income needs to last; how aggressively or conservatively the portfolio should be invested; how much spending is sustainable; whether longevity-protected products like annuities are warranted. The standard public statistic — average life expectancy at birth, currently around 81 years for Australian men and 85 for Australian women — is widely cited but is not the right input for retirement planning. Better longevity literacy supports better planning, and the framing of life expectancy as a distribution rather than a point produces materially different conclusions than the simple average.
The first issue with using average life expectancy at birth is that it's an average for births, not for retirees. Many people who will reach retirement have already survived past the ages where mortality reduces the birth-cohort average. Conditional life expectancy at age 65 — the average remaining lifespan for someone who has already survived to 65 — is meaningfully higher. For Australian men aged 65, it's approximately 85 years on average; for women, approximately 88. The conditional figure is the relevant baseline, not the at-birth figure.
The second issue is that life expectancy is a distribution, not a point. Half of people live longer than the median by definition. About 25% of women at age 65 will live to 95 or beyond. About 10-15% of men at 65 will live to 95+. For retirement planning, the right tail of the distribution is where planning failure is most damaging — and ignoring it produces inadequate planning for the substantial number of retirees who actually live to those ages.
The asymmetry is real. A retiree planning to age 85 who dies at 80 has unspent savings — an estate result, but not a retirement planning failure (the retiree has not run out of money during life). A retiree planning to age 85 who lives to 95 faces 10 years of inadequate retirement income — running out of savings, reducing lifestyle, becoming dependent on Age Pension or family. The two outcomes are not symmetric in cost. Planning conservatively for longevity (assuming a longer life) produces a small reduction in spending capacity in expectation but eliminates the catastrophic right-tail outcome. Planning aggressively (assuming a shorter life) produces slightly higher spending but exposes the retiree to severe right-tail risk.
For most healthy retirees, planning to age 90-95 rather than age 85 captures a meaningful portion of the right tail. Planning to age 85 underestimates the horizon for substantial numbers of retirees by definition.
For couples, the situation is even more nuanced. The relevant planning horizon is when the longer-lived partner dies — typically substantially later than either partner's individual life expectancy. Approximate joint life expectancy figures for an Australian couple both aged 65: at least one partner alive at age 90 has approximately 50% probability; at least one partner alive at age 95 has approximately 20-25% probability; at least one partner alive at age 100 has approximately 5% probability. For couples, planning the household's retirement income to support both partners through to about age 90-92 (with the right-tail consideration that one partner may live substantially longer) is appropriate. The death of one partner produces a significant change in retirement income picture — the Age Pension transitions from couple rate to single rate; super death benefits transfer to the survivor; living costs reduce somewhat but not proportionally; the surviving partner often experiences a real reduction in living standard. Planning for the household's joint trajectory (and the eventual transition to single survivor) supports better outcomes than planning to either partner's individual horizon.
Several factors substantially affect individual longevity beyond the population average. Health at age 65 — smokers and those with chronic conditions have shorter expected lifespans; those with very good health and family longevity have longer. Socioeconomic status — higher-income, better-educated retirees typically live longer than population average, sometimes by several years. Lifestyle — physical activity, diet, social connection, and other lifestyle factors substantially affect longevity. Family history — strong family longevity (parents living to 90+) suggests longer expected lifespan than population average. Specific conditions — heart disease, diabetes, cancer history all affect individual longevity expectations.
For retirement planning, individual circumstances should adjust the population-average expectations. A 65-year-old with very good health, no chronic conditions, family history of longevity, and high socioeconomic status may reasonably plan for age 95 or beyond. A 65-year-old with multiple health concerns may plan for a shorter horizon. The adjustment is meaningful — sometimes 5-10 years either way — and supports more accurate retirement planning than blanket use of population averages.
Several practical implications follow from better longevity literacy. Plan to a higher age than the average — use age 90-95 rather than age 85 as the baseline planning horizon for healthy retirees. Consider longevity protection products — annuities, deferred lifetime annuities, group self-annuities address right-tail risk that account-based pensions don't. Maintain growth allocation longer — long planning horizon supports continued growth-asset exposure rather than aggressive derisking. Coordinate with the rising equity glide path concept — for retirees with longer horizons, gradually rebuilding equity through retirement may be appropriate. Plan for the surviving partner — the longer-lived partner's needs are part of joint planning. Adjust for individual circumstances — population averages need adjustment for health, family history, lifestyle.
For most retirees, planning to a longer horizon than seems intuitively reasonable produces better outcomes than planning to the average. The cost of overestimating is modest (slightly less spending in retirement, larger eventual estate); the cost of underestimating is catastrophic (running out of money in late retirement, when returning to work or substantially reducing spending is rarely realistic).
A few common pitfalls. Planning to average life expectancy at birth substantially underestimates the planning horizon for typical retirees. Using point estimates rather than distribution misses the right-tail risk. Ignoring joint life expectancy for couples plans to the wrong horizon. Not adjusting for individual circumstances misses material differences in expected lifespan. Underestimating right-tail risk — living to 95+ has meaningful probability and matters substantially for planning. Treating "I don't expect to live that long" as planning input — personal expectation may understate actual probability, and planning should be robust to a range of outcomes rather than relying on subjective forecast.
For retirees and pre-retirees, this is one of those structural framings that changes the planning conversation. Once the longevity literacy is in place — conditional expectancy at 65 rather than at-birth average; distribution rather than point; couple's joint horizon rather than individual; adjusted for personal circumstances — the rest of the retirement plan typically looks meaningfully different from one built on the simpler "plan to age 85" assumption.
Key takeaways
- Life expectancy figures at birth underestimate the planning horizon for retirees. Conditional life expectancy at age 65 is approximately 85 for Australian men and 88 for Australian women — materially higher than the commonly cited at-birth averages — because many people who die before retirement pull the at-birth average down.
- Life expectancy is a distribution, not a point. About 25% of women at age 65 will live to 95 or beyond; roughly 10-15% of men at 65 will reach 95+. Planning to the median or average means the plan will run short for a substantial proportion of the population who actually live longer.
- The asymmetry matters: dying earlier than planned leaves unspent savings (an estate result). Living longer than planned means 10+ years of inadequate income at the stage of life when returning to work is rarely possible. Planning conservatively for longevity eliminates the catastrophic right-tail outcome at a modest cost in expected spending.
- For couples, the relevant planning horizon is when the longer-lived partner dies — not either individual's life expectancy. For a couple both aged 65, there is approximately 50% probability that at least one partner is alive at 90, and 20-25% probability at 95.
- Individual circumstances adjust the population average materially. Better health, higher socioeconomic status, positive family history of longevity, and active lifestyle can each extend expected lifespan by several years above the population average. A healthy 65-year-old with family longevity may reasonably plan to age 95 or beyond.
Frequently asked questions
What is conditional life expectancy at 65 and why is it the right planning number?
Conditional life expectancy at 65 is the average additional lifespan for someone who has already survived to age 65. For Australian men at 65 it is approximately 85 years; for women, approximately 88 years. This is substantially higher than the at-birth figures (around 81 for men, 85 for women) because those averages include people who die before reaching retirement and pull the average down. For retirement planning, the conditional figure at 65 is the correct baseline — the at-birth figure substantially underestimates the typical planning horizon.
Why is longevity risk more dangerous than dying earlier than expected?
The two outcomes are not symmetric in cost. If you die earlier than your plan assumed, the result is unspent savings — an estate outcome, not a retirement planning failure. If you live longer than planned, the result is inadequate retirement income during the years when returning to work is rarely realistic. Planning conservatively for longevity — using a longer planning horizon such as age 90-95 instead of 85 — eliminates the catastrophic right-tail outcome at the modest cost of slightly lower expected spending throughout retirement.
How should couples approach longevity planning differently from singles?
For couples, the relevant planning horizon is not either partner's individual life expectancy — it's the expected lifespan of the longer-lived partner. For an Australian couple both aged 65, there is approximately 50% probability that at least one partner is alive at 90, and 20-25% probability that at least one is alive at 95. The death of the first partner changes the retirement income picture materially: the Age Pension transitions from couple rate to single rate, living costs reduce somewhat but not proportionally, and the surviving partner often experiences a real reduction in living standard. Planning for the household's joint trajectory — and the eventual single-survivor transition — produces better outcomes than planning to either individual's horizon.
How do individual health and lifestyle factors adjust the standard life expectancy estimate?
The population-average figures are starting points only. Smokers and those with multiple chronic conditions typically have shorter expected lifespans than the average; those with very good health at 65, no chronic conditions, active lifestyle, and strong family history of longevity (parents living to 90+) have longer. Higher socioeconomic status and education are also associated with longer lifespan, sometimes by several years. The practical adjustment is meaningful — sometimes 5-10 years either way — and supports more accurate planning than applying the same figure regardless of individual circumstances.