Skip to main content
main content, press tab to continue
Article | Pensions Briefing

UK Spring Budget 2023: The Money Purchase Annual Allowance

By David Robbins and Dave Roberts | March 9, 2023

How much people who dip into pensions can save in future was a loose end left by ‘pension freedom’. There are no easy answers.
Pension Board and Trustee Consulting|Pensions Corporate Consulting|Pensions Risk Solutions|Pensions Technology|Retirement

As pensions policy announcements go, the big reveal of “pension freedom” in March 2014 provided unrivalled theatre: George Osborne told the House of Commons that “no one will have to buy an annuity,” insurers’ share prices tumbled, and a wrongfooted Leader of the Opposition made no mention of the Budget’s defining announcement in his response.

It did, though, leave a few loose threads. One was how much, if anything, over-55s who utilised the new freedoms would be allowed to contribute to defined contribution (DC) pensions thereafter. Nine years later, this is still a live issue on which Mr Osborne’s successor-but-five is being lobbied1. Unfortunately, there are no great options: this is a “choose your poison” policy choice where the best the Chancellor can do is balance competing risks.

For three years before the full fat version of pension freedom was announced (ie, since 2011), people who demonstrated that they had a secure annual income of at least £20,000 had been allowed to access remaining defined contribution pot balances as they wished. Entering an arrangement that permitted this meant forswearing the opportunity to pay into a DC pot in future, at least without paying a punitive Annual Allowance charge. While no policy was spelled out, the 2014 arithmetic assumed that this approach would be extended when pension flexibility went mass market. But with the rollout of automatic enrolment by then well under way, this would have required employers to nudge many of their older workers into incurring a tax penalty.

At the other extreme, giving over-55s free rein to divert a chunk of their pay into a pension and take it straight out again would make one quarter of that money tax-free (provided the member did not breach the Lifetime Allowance) even though nothing would have been put aside for retirement. And there could be National Insurance savings on top if employers offered pension contributions in lieu of pay. Rules on “recycling” tax-free lump sums, which seek to prevent them being used to finance a significant increase in contributions, would have limited the fiscal red ink in this scenario, but these require HMRC to judge the taxpayer’s motives and how much might normally be contributed2.

In July 2014, the Government found a middle ground: a £10,000 Money Purchase Annual Allowance, with no facility to carry forward unused allowances, would apply once someone had made a taxable withdrawal (except where cashing out a small pot: for some people, this will be a reason not to consolidate their DC pensions). As well as £10,000 being a lot more than zero, this allowed people accessing their savings early to escape the restriction provided that they only took their tax-free lump sum and left the remainder invested.

The short-term cost of loosening the rules in this way was more than cancelled out by the expected acceleration in tax revenues from the simultaneous decision to continue to permit transfers out of funded defined benefit schemes3, but the policy still had an air of impermanence: the minister responsible told MPs: “We will, of course, continue to look at the matter closely to ensure that the system is not exploited at a significant cost to the Exchequer”4.

If that sounded ominous, so it proved: in 2017, the MPAA was cut to £4,000. The consultation paper on this change did not say that the system was being exploited; nor did the projected revenue – £76 million a year, rolled forward to today5 – suggest this. But the Government may have wanted to close the stable door before the horse bolted – or, at least to house a less valuable horse in the stable with the dodgy door.

When cutting the MPAA to £4,000, HM Treasury initially suggested that this should not cause problems for automatic enrolment because 8% of ‘qualifying earnings’, the statutory minimum level of default contributions, is always less than £4,0006. Currently, qualifying earnings are earnings between £6,240 and £50,270, so 8% of qualifying earnings is worth a maximum of £3,522.40.

However, many large employers automatically enrol people at more generous contribution levels, frequently using alternative scheme quality tests where pensionable earnings must at least equal basic pay. In a widely-used alternative test, the minimum combined employer and employee contribution rate is 9%, so someone whose basic salary is £44,445 or more would by default have more than £4,000 a year going into their DC pot. HM Treasury dismissed this objection, noting that someone who exceeded the MPAA by £5 would incur an Annual Allowance charges of just £27. (This of course means that collection costs for HMRC and associated hassle for taxpayers would be high in proportion to charges levied. That will be true for some people wherever the limit is set, but the number in this position should be higher with a low MPAA.)

This issue might become more acute now that the Government is supporting the Pensions (Extension of Automatic Enrolment) (No. 2) Bill and suggesting that this will pave the way for the bottom of the qualifying earnings band to be abolished – albeit without offering a timetable. If and when this happens, 8% of qualifying earnings will be worth more than £4,000 for people with earnings above £50,000 (though the excess would be a maximum of just £21.60, if the upper qualifying earnings threshold remains where it is).

Some other things that have changed since 2017 might also point to the right answer today being higher than it was then, though without neutralising concerns about the scope for abuse. For example:

  • The increase in the price level means that, had the MPAA been indexed, it would be set to reach around £4,900 in April 2023. (However, while recent inflation has been much higher than forecast, some erosion of the MPAA’s real value would have been part of the 2017 plan.)
  • Some over-55s will have tapped into their pension savings sooner than planned, to tide them over while unable to work during the pandemic.
  • The Government is eager to encourage early retirees back to work, with the Chancellor recently telling them “‘Britain needs you’ and we will look at the conditions necessary to make work worth your while”8.

A higher MPAA would improve the terms on which some over-55s can trade leisure for money. But the Government may need some persuading that it would move the dial for people who have embraced a retired lifestyle and who can afford not to go back (and that, if it did, the positive effect on the public finances would outweigh the cost of letting people who continued to work after accessing their pensions contribute more). It might also worry that some people may be more tempted to stop work if they could more easily repair any dent that a negative market shock might make to their pot balance. And if the aim were to use the pension tax system to encourage work at older ages, making it harder to access savings in the first place might be a more obvious approach.

If the Chancellor is persuaded to ease constraints on people currently or prospectively affected by the MPAA, the £4,000 level is an easy lever to pull, but other changes are imaginable.

For example, the Office of Tax Simplification suggested in 2019 that contributions to pension schemes used to meet employers’ duties under automatic enrolment legislation need not count towards the MPAA limit9. This would be a significant loosening, as it would mean allowing people to contribute £4,000 on top of automatic enrolment contributions, which could themselves exceed £4,000 in some cases.

A variant of this model would be to set the MPAA at the higher of £4,000 (or some other amount) or the ‘normal’ contribution to an employer’s scheme. But this could mean that two individuals in otherwise identical circumstances would face different limits depending on whether their employer chose the top or bottom of a matching scale to be the default contribution; employers are also free to establish different schemes, or different contribution scales, for different groups of employees, and some of these will be far more generous than required to comply with minimum quality standards under automatic enrolment.

Another Office of Tax Simplification idea was the Government “could remove tax relief from amounts recontributed after a withdrawal, but not otherwise limit pension saving”. This was inspired by the ISA regime, where limits apply to amounts saved each year, net of withdrawals. But if a DC pension equivalent only applied on an in-year basis, people who had withdrawn significant sums in the past (and whose ability to contribute in future is affected by what these withdrawals did to their bank balance or debt levels) would be able to contribute up to £40,000 a year – or more, if speculation about an imminent boost in the standard Annual Allowance proves accurate. Applying it to running totals has attractions in principle but would require records of withdrawals and contributions made since 2015 and may irk people who had made withdrawals in the past in the expectation that they could still contribute something.

Whereas the original £10,000 limit had been estimated to affect only 2% of pension savers over 5510, AJ Bell recently reported having been told by the Government that 25% of over-55s saving in workplace DC schemes contribute more than £4,00011. This indicates how many people might have to reduce their contributions if they accessed their pensions flexibly. But it doesn’t shed light on how many of the people subject to the MPAA would otherwise be contributing more than £4,000. (HMRC statistics indicate that more than two million individuals had taken taxable flexible withdrawals by Q2 202212. But many of these will be retired.)

Nor is there much data on how rigorously the policy is being enforced. Statistics on people paying Annual Allowance charges do not split out charges in respect of MPAA breaches, and in 2019 HMRC declined to say how many people had been fined for failing to tell the scheme into which they were paying that they had triggered the MPAA13.

Without these numbers, it is hard to reach firm conclusions about what effect the MPAA has had. But the tensions between competing policy objectives mean that any change made in the Budget may amount to replacing one sticking plaster with another. The 2023 Budget may not be the final act of the drama started in 2014.


1. For example, a letter sent to HM Treasury by the Lang Cat, and signed by a number of pension providers and industry bodies, calls for an immediate rise in the Money Purchase Annual Allowance to £10,000 and for the Government to discuss possible improvements to the MPAA.
2. These rules are summarised in HMRC’s Pensions Tax Manual 133810.
3. Autumn Statement 2014 policy costings, page 46.
4. David Gauke, Hansard: House of Commons Debates, 29 October 2014, column 328.
5. OBR policy measures database – see rows 1424 and 1425 of the “tax measures” worksheet.
6. Reducing the money purchase annual allowance: consultation, HM Treasury, November 2016, paragraph 3.12.
7. Reducing the money purchase annual allowance: consultation response, HM Treasury, March 2017, paragraph 2.11.
8. Jeremy Hunt’s speech at Bloomberg, 27 January 2023.
9. Taxation and Life Events, Office of Tax Simplification, October 2019.
10. Freedom and choice in pensions: government response to the consultation, HM Treasury, July 2014, page 6.
11. MPAA risks rise as quarter of pension savers over 55 contribute over limit, Professional Adviser, 15 February 2023.
12. HMRC private pension statistics, table 9.
13. HMRC’s response to a Freedom of Information request from Royal London, reported in HMRC unsure on number of pension tax rulebreakers, Pensions Age, 18 July 2019.


Director, Retirement
email Email

email Email

Contact us