The retirement savings gap between men and women is larger than the pay gap in most countries where both are measured. This is not surprising once the inputs are laid out, and none of the inputs is about financial competence.
Why the gap is larger than the earnings gap
Four factors multiply rather than add.
Lower average earnings produce lower contributions where contributions are a percentage of pay.
Career interruption produces years of zero or reduced contribution, and those years are frequently in the early-to-middle period when compounding has the most time to work.
Part-time work reduces contributions proportionally and, in some systems, drops workers below auto-enrolment thresholds entirely — meaning zero rather than reduced.
Longer life expectancy means the accumulated total must fund more years.
A smaller pot, funded for fewer years, spread over a longer retirement. Each factor is modest; their product is not.
The threshold problem
This one is worth isolating because it is a design flaw rather than an inevitability.
Auto-enrolment systems typically have an earnings threshold below which enrolment does not apply. The stated rationale is to avoid enrolling people for whom saving is inappropriate.
The effect is that part-time and low-paid workers — who are disproportionately women — are excluded from the single most effective savings mechanism ever implemented.
Worse, in systems where the threshold applies per job rather than per person, someone holding two part-time positions can earn above the threshold in total and be enrolled in neither.
This has been identified repeatedly in policy analysis and remains largely unaddressed in several jurisdictions.
The arithmetic of starting early
Compound growth is exponential in time, which means the value of a contribution depends far more on when it is made than on how large it is.
The standard illustration is worth doing. A sum invested at twenty-two and left for forty-five years experiences many more doublings than the same sum invested at forty and left for twenty-seven. The difference in final value is a multiple, not a margin.
The practical consequence is that early contributions are worth several times later ones of the same size, and that a small amount started at twenty-two beats a much larger amount started at thirty-five.
This is exactly the wrong shape for how information is distributed. Retirement planning content is aimed at people in their forties and fifties, who are the ones with money and interest. The people for whom the information is most valuable are twenty-two and are not reading it.
The interruption planning question
For anyone who anticipates a career interruption, several things follow that are rarely stated.
Contributions made before an interruption are worth more than contributions after, both because of compounding time and because income after interruption is frequently lower.
Where systems permit voluntary contributions during a period out of work, or credit periods of caring, these are frequently underused because they are poorly publicised.
In several jurisdictions, a partner can contribute to a non-earning spouse's pension, sometimes with tax relief. This is not widely known and is one of the more effective available responses to the interruption problem.
And in relationship breakdown, pensions are frequently the largest asset after property and are frequently overlooked in settlements. Analyses of divorce settlements have found pension sharing arrangements used far less often than the asset values would justify.
What actually gets people to start
Defaults, overwhelmingly. Auto-enrolment increased participation dramatically wherever it has been implemented, and the effect dwarfs anything achieved by education campaigns.
Auto-escalation — contributions rising automatically with pay increases — has similarly strong evidence behind it, because it avoids the moment of active decision that most people postpone indefinitely.
The behavioural finding underneath both is that the barrier is not disagreement with saving. It is the requirement to make a decision, which people defer regardless of how much they agree with the principle.
The four things to do before twenty-five
Enrol, and never opt out. The match, where it exists, is compensation being declined.
Contribute above the minimum if possible, since minimums are set for administrative convenience rather than adequacy.
Check the default fund. Default allocations are frequently conservative in a way that is inappropriate for a forty-year horizon, and the difference in outcome between a conservative and a growth allocation over that period is substantial.
Consolidate old accounts when changing jobs. Small forgotten pots are extremely common and are eroded by fees.
None of this is complicated. All of it is worth more done at twenty-two than anything done at forty-five, and almost nobody tells twenty-two-year-olds that the clock is the asset.