Two more years of work moves three levers at once

This looks at working longer before drawing from the practical end — what holds up once conditions stop being ideal.

Three effects, not one

Working longer adds further contributions, which increases the balance directly for as long as those contributions keep arriving each month. It removes years of withdrawals, and in most cases that is the larger of the two effects by a considerable margin. It shortens the remaining period the portfolio must fund, which raises the sustainable withdrawal rate on what is left.

These three act together, which is why the effect of even a short delay is larger than most people expect. The arithmetic is straightforward and does not depend on any assumption about future returns at all.

The interaction with guaranteed income

Deferring a state or workplace pension increases the amount in many systems, on terms set by legislation or scheme rules. Those terms differ substantially between countries and are periodically revised, so only the current rules are relevant. The increase is effectively the purchase of additional guaranteed income using the payments you gave up.

On an ordinary week, whether that represents good value depends on health, on other income and on how long the income is likely to be received. This is a genuine decision with a jurisdictional answer, and it deserves local regulated advice rather than a general rule.

What it costs

The cost is years of life not spent in retirement, and those years cannot be recovered at any later point. Health and energy are typically highest in the early retirement years, and delaying spends part of that period at work. People in physically demanding or high-stress work may not have the option available to them at all.

The useful part is this: framing the decision purely as arithmetic ignores the part that most people actually care about most. The honest version presents both sides rather than treating a delay as self-evidently sensible.

Partial retirement as a middle option

Reducing hours captures a useful part of the effect while returning a good deal of the time you were trading away. Even modest continuing earnings cut the withdrawal rate during the most demanding early years of drawdown.

Some pension arrangements allow phased access alongside part-time work, and others do not permit it at all. Employer flexibility varies, and the option is considerably easier to negotiate before leaving than afterwards.

For many households this is the arrangement that resolves the trade-off better than either extreme does.

Deciding it in advance

A plan naming the conditions under which you would work longer is more useful than a plan naming a fixed date. Those conditions might be a portfolio value, a market condition or simply whether the work is still tolerable.

Deciding them in advance avoids the version where a bad market forces the decision under pressure. It also makes the option visible as a form of flexibility rather than as evidence that the plan failed. Flexibility named in advance is worth considerably more than flexibility discovered during a crisis.

What not to conclude

None of this says working longer is the right answer, and for many people it is neither possible nor desirable. It says only that the effect is larger than intuition suggests, which is worth knowing when weighing up the trade-off honestly.

On an ordinary week, it also says a small delay can substitute for a large amount of additional saving at an earlier stage. The reverse holds equally: stopping earlier requires a considerably larger portfolio than most people assume. Both directions of that arithmetic are worth seeing clearly before the date is finally set.

The takeaway

A short delay does three things at once. That is why the arithmetic surprises people in both directions.

The version you keep doing is the version that works.