Auto-enrolment solved who saves. It hasn't settled how much, or what happens to old pots.
Participation among eligible employees rose from 47% to 90% in little over a decade — a genuine policy success. Three separate questions remain open: whether the minimum contribution is enough, what replaces defined benefit as it winds down, and what happens to the pension pots people leave behind when they change jobs.
Automatic enrolment is one of the clearer policy successes covered anywhere on this site: participation among eligible employees rose from 47% in 2012 to 90% by 2025, with opt-out rates staying low. But three separate questions sit underneath that success story: whether the minimum 8% contribution is actually enough (it is not really 8% of anyone’s salary, and industry consensus says even a true 8% likely would not be enough anyway), what happens as defined benefit provision keeps shrinking to a shadow of its former membership, and what happens to the pension pots people accumulate and then leave behind every time they change jobs.
Read this first — what else changes the meaning of the headline
- Auto-enrolment participation, eligible employees
- 90%
2025 — up from 47% in 2012, just before automatic enrolment began. Full trend chart on the funding page.
DWP, Workplace pension participation and savings trends of employees, 2009 to 2025
- Effective contribution rate on total salary, at the £50k band ceiling
- ~7%
Not the 8% headline. The minimum applies to "band earnings" only, so the rate on someone's WHOLE salary is always lower — and lower still for anyone earning close to the £10,000 eligibility threshold.
Pensions Policy Institute, "Automatic Enrolment Contributions Briefing Paper"
- DB active members
- 662k
2025 — down from 1.98m in 2012. Only 7% of all DB members are still actively building up benefits.
The Pensions Regulator, "Occupational defined benefit landscape in the UK 2025"
- Small pension pots (£1,000 or less)
- 13m
Growing by around 1m a year — mostly created when someone changes job and their old workplace pot is left behind, unconsolidated.
DWP, "£1,000 retirement savings boost from plans to bring together small pension pots"
Two separate issues hide inside the phrase "8% minimum contribution" — what it is actually 8% of, and whether 8% is even the right target in the first place.
It is not 8% of your salary
The 8% minimum (3% of it from the employer) applies to “band earnings” — the slice of pay between £6,240 and £50,270 — not total salary. The 8% minimum applies to "band earnings" (the slice between £6,240 and £50,270), not total salary — so the EFFECTIVE contribution rate on someone's whole pay packet is always less than 8%, and varies by income. PPI's own modelling shows it starts around 3% of total salary at the £10,000 eligibility threshold, rises to a peak of around 7% at the £50,270 upper band limit, then falls again above that as a shrinking share of total pay counts as "band earnings" at all.
The 2017 statutory AE review recommended removing the £6,240 lower earnings limit, so contributions would apply from the first pound of income. The Pensions (Extension of Automatic Enrolment) Act 2023 gives government the power to do this, but as of this page's last check it had not been implemented — so the 8% minimum is still not actually 8% of salary for anyone.
Even a true 8% may not be enough
Industry consensus, reflected across PPI's and others' modelling, is that 8% contributions are unlikely on their own to deliver a retirement income that maintains someone's working-life standard of living. 12% has been the most commonly suggested alternative, though PPI's own paper states this as the range under discussion rather than a single settled figure this site independently re-derives.
Status quo
8% total, at least 3% from the employer.
Increase the minimum to around 12%
Would likely require increases from both employer and employee.
Optional additional contribution, with an opt-down
Auto-enrol at a higher rate but let people opt down to the current 8% before opting out entirely, rather than an all-or-nothing increase.
Tiered contributions by income
Lower minimums for lower earners (already relying more on the State Pension), higher minimums for higher earners — mirrors a pattern already used in some DB schemes, not currently used in DC/AE.
None of these is government policy — they are the options an independent policy institute sets out as under discussion, presented here without a recommendation.
The funded-pensions page covers DB, DC and LGPS asset totals in depth. The membership side tells a starker version of the same story: defined benefit provision is not just shrinking in relative size — for almost everyone still in a DB scheme, it is no longer something they are actively building up.
The DC side of this shift — master trust growth, default-fund investment mix, and how much of it is invested in the UK — is covered on who holds Britain’s retirement savings and where the money is invested, not repeated here.
Auto-enrolment's own success has created a side effect: every job change that triggers a new enrolment can leave a small, forgotten pot behind at the old employer's scheme.
The government's response is in the Pension Schemes Bill 2025: a A "multiple default consolidator" model — several authorised consolidator schemes, not one single national scheme — with member choice built in (a saver can choose their own consolidator or opt out of automatic consolidation entirely). Target go-live: 2030.
What this page cannot tell you
Real limits on what the adequacy and small-pots evidence can establish:
What contribution rate would actually be "enough"
How many of the 13 million small pots are genuinely "lost" versus simply small and inactive
What it means
- What the data directly shows
- Auto-enrolment participation rose sharply and has stayed high; the minimum contribution rate is lower in practice than its 8% headline suggests; DB active membership has fallen by roughly two-thirds since 2012; and small pension pots are accumulating faster than they are being consolidated.
- What can reasonably be inferred
- Getting people INTO pension saving and getting them enough saved for an adequate retirement are different achievements — auto-enrolment has substantially solved the first without yet resolving the second.
- What is disputed
- Whether contribution rates should rise, and if so how quickly and who should bear the cost (employers, employees, or both), is genuinely contested policy territory with real trade-offs on either side, not a settled question.
- A political judgement, not a finding
- No government has yet committed to raising the minimum contribution rate — doing so, and how, remains a live political choice.
- What the evidence cannot establish
- The actual retirement income any individual saver will end up with — that depends on their own contribution history, investment returns, and career, none of which a system-wide statistic can predict for a specific person.
Part of the pensions vertical (overview). The full auto-enrolment participation-rate chart is on the funding page. Related: who holds Britain’s retirement savings and where the money is invested.