Nothing changes on a payslip until 6 April 2029. That is the first thing to say about the salary sacrifice cap, because it is routinely discussed as though it has already happened.
What changes then is narrow. Only the first £2,000 a year sacrificed into a workplace pension keeps its National Insurance exemption. Anything above that attracts National Insurance from both the employer and the employee.
How the Arrangement Works Today
Salary sacrifice, or salary exchange, has a simple mechanic. An employee gives up part of their gross salary and the employer increases its pension contribution by the same amount.
The effect, as Blake Morgan sets out, is that an employee contribution becomes an employer contribution. Because the salary is lower, both sides pay less National Insurance on it. There is currently no ceiling, provided the post-sacrifice salary stays above the minimum wage.
That is the exemption being trimmed. Employees will still be able to contribute as much as they like after 2029; the contributions simply stop being NI-free above £2,000. Employers will report the total sacrificed through existing payroll software.
The enabling law, the National Insurance Contributions (Employer Pensions Contributions) Act 2026, became law on 29 April 2026. Most of the operational detail is still to arrive in regulations, which is a material caveat for anyone modelling this now.
£468m of National Insurance, and 2.8 Million Smaller Pensions
The costings are where the measure becomes interesting rather than merely technical.
The Treasury is expected to collect around £468 million a year in additional National Insurance, with the wider package raising roughly £4.8 billion once extra income tax receipts are counted. Around 4.3 million workers contribute £2,000 or less through salary sacrifice and are untouched.
Set against that, HMRC expects more than 2.8 million people to reduce their pension contributions as a result.
Those two numbers belong in the same sentence more often than they appear in one. A measure raising £468m a year in NI is also forecast to make 2.8 million people save less for retirement. Both are official expectations, not campaign claims, and the second is a behavioural forecast baked into the costing rather than an unintended side effect somebody spotted later.
About £84 Each
For an individual caught by the cap the direct cost is modest: roughly £84 more a year on average.
£84 is not a sum that changes a household budget. It is, however, the kind of number that changes a default. Salary sacrifice is typically presented to staff as free efficiency, and the framing does the persuading. Once a portion of it carries National Insurance, the pitch becomes conditional, and conditional pitches convert less well.
That is the most plausible route from a small per-person cost to a 2.8 million-person behavioural response. Nobody rationally abandons pension saving over £84. Plenty of people quietly stop increasing it.
The two headline figures also do not divide into each other, and that is informative rather than contradictory. At £84 a head, £468m would imply well over five million payers. The gap exists because the £468m counts employer National Insurance as well as employee, so a large share of the yield never appears on anyone’s payslip at all.
The Case the Chancellor Made
The measure was announced in the Budget of 26 November 2025 by the then chancellor, Rachel Reeves.
Her argument was about drift. Salary sacrifice was intended to be a small part of the pensions system, she said, but its cost was set to triple, from around £2.8bn in 2016-17 to a projected £8bn by 2030, with “the greatest benefit going to highest earners”. She called the cap “a pragmatic step”.
The distributional defence sits in the Budget documentation, which said the cap would “shield” 74% of basic rate taxpayers using salary sacrifice. That is a real protection and it is also the reason the yield is comparatively small: a relief mostly left intact does not raise much.
What Employers Should Be Doing Before 2029
Three things are worth starting well before the deadline, none of them urgent this year.
The first is knowing the number. Employers rarely have a clean view of how many staff sacrifice above £2,000, because the figure only matters after the cap exists. That population is the entire exposure, on both cost and communication.
The second is the employer NI line itself. The charge falls on both sides, so a workforce with generous sacrifice arrangements carries a real cost increase, and there is a standing concern that some of it is eventually reflected in salaries rather than absorbed.
The third is the benefits pitch. A scheme sold as costless needs re-explaining before the first payslip changes, not after. That work lands on the same teams already managing rising employment costs, in a market where public sector pay growth is running at nearly twice the private rate.
The Employer Side Is the Quiet Half
n
Most coverage of the cap describes what an employee loses. The charge falls on both parties, and the employer half is the one that shows up in a budget line rather than a payslip.
n
An employer running a generous scheme has, until now, had a genuine incentive to encourage sacrifice: every pound moved from salary to pension reduced its own National Insurance bill as well as the employee’s. After April 2029 that incentive stops at £2,000 per head. Above the cap the employer pays National Insurance on contributions it is making on the employee’s behalf, which inverts the logic of the arrangement for exactly the staff who use it most.
n
There is a standing concern, raised repeatedly since the Budget, that some of this cost is eventually reflected in salaries. That is not a prediction anyone can evidence yet, but it is the mechanism to watch, because it is how an employer-side charge becomes an employee-side one without any policy changing.
n
The reporting duty is comparatively light. Employers report the total sacrificed through existing payroll software, so the compliance burden is a software update rather than a new return. The harder task is the one no system does for you: identifying which employees sit above £2,000, what their contribution behaviour is likely to be once the exemption thins, and whether the scheme as currently sold still describes what staff will actually get.
n
The Three-Year Gap Is the Point
A 2029 start date on a measure legislated in 2026 is unusually long, and the length is doing work.
It gives employers time to renegotiate benefit structures, gives payroll providers time to build the reporting, and gives the regulations time to arrive. It also gives three years in which the behavioural forecast can be tested against what savers actually do, and three years of Budgets in which a £2,000 threshold could be adjusted, frozen or superseded.
For now the planning assumption is straightforward. The relief survives for most people, disappears above £2,000 for a substantial minority, and the government’s own expectation is that a meaningful number will respond by putting less into their pension. The revenue is small, the population affected is not, and the policy will be judged on which of those two numbers turns out to matter more. It sits alongside a run of targeted business tax changes, including the 20% business rates cut coming for 32,000 pubs and clubs, in a period where the direction of travel is narrower reliefs applied more selectively.
About The Author
More Stories
UK House Prices Rose 2% and London Fell for a Tenth Month
Trusted Payments Holds 10% in Escrow Until You Sign Off
UK Manufacturing Hired at a Two-Year High While Output Slowed