Key takeaways
- England's average Band D bill is £2,392 in 2026-27, up £111 or 4.9% on the year before. A 4.99% rise in 2027-28 would add £119.36 to it.
- At 4.99%, the yearly increase runs from £79.57 at Band A to £238.72 at Band H, because the statutory proportions put Band H at exactly three times Band A.
- Nothing is settled for 2027-28. The 2026-27 principles report says it "relates to 2026-27 only", and each year's limit needs its own approval by the House of Commons.
- Six councils, Wandsworth and Westminster among them, are proposed to have no referendum limit at all in 2027-28 and 2028-29.
- 274 of the 384 authorities under referendum principles took the maximum in 2026-27, or 71% of them, so the ceiling is the central case rather than the outside one.
What a 4.99% council tax increase in 2027 costs at every band
You want one figure: what your bill does next April, and by how much. Here's the short answer on the council tax increase for 2027 in England. Most councils can lift their Band D charge by up to 4.99% without holding a referendum, and the government has said it intends to keep that limit in place through 2027-28. Applied to the England average Band D bill of £2,392 for 2026-27, 4.99% comes to £119.36 a year, or £9.95 a month.
Your own band scales off that one number. Section 5 of the Local Government Finance Act 1992 fixes the proportions between bands as "6: 7: 8: 9: 11: 13: 15: 18", where 6 is Band A and 9 is Band D. So Band A pays 6/9 of the Band D charge and Band H pays 18/9, which is three times Band A. Those ratios don't move when a council raises its rate. A percentage increase therefore lands on all eight bands at once, and the cash it costs you is simply your band's share of it.
The table runs three rates against the 2026-27 England average. None of them is a confirmed 2027-28 figure. The 4.99% column is the one to read first, because that's the ceiling most councils are expected to have, and 274 of them took their ceiling in 2026-27.
| Band | Share of Band D | 2026-27 average bill | +2.99% a year | +4.99% a year | +4.99% a month | +8.99% a year |
|---|---|---|---|---|---|---|
| Band A | 6/9 | £1,594.67 | £47.68 | £79.57 | £6.63 | £143.36 |
| Band B | 7/9 | £1,860.44 | £55.63 | £92.84 | £7.74 | £167.25 |
| Band C | 8/9 | £2,126.22 | £63.57 | £106.10 | £8.84 | £191.15 |
| Band D | 9/9 | £2,392.00 | £71.52 | £119.36 | £9.95 | £215.04 |
| Band E | 11/9 | £2,923.56 | £87.41 | £145.89 | £12.16 | £262.83 |
| Band F | 13/9 | £3,455.11 | £103.31 | £172.41 | £14.37 | £310.61 |
| Band G | 15/9 | £3,986.67 | £119.20 | £198.93 | £16.58 | £358.40 |
| Band H | 18/9 | £4,784.00 | £143.04 | £238.72 | £19.89 | £430.08 |
The chart plots the middle column. It's the same shape as the band ladder itself, which is the point: at a single percentage rate, the cash increase is three times larger at Band H than at Band A, and the gap between any two bands widens every year the rate is applied.
Why the cap is 4.99% and not 5%
The limit isn't drafted as a cap. It's drafted as a definition of excess, which is why the figure councils announce is such an odd one. The 2026-27 principles report puts it like this: a council's basic amount of council tax is excessive "if the authority's relevant basic amount of council tax for 2026-27 is 5% (comprising 2% for expenditure on adult social care, and 3% for other expenditure), or more than 5%, greater than its relevant basic amount of council tax for 2025-26".
Because 5% itself counts as excessive, the highest a council can go is the largest number below it. MHCLG's own statistics treat "increases of 4.985% or above" as a maximum increase, and 4.99% is simply the largest two-decimal figure that stays under the line. There's no magic in the second decimal place. It's just the near side of a threshold.
The 5% comes in two parts. Three points are for general spending, and two are ring-fenced to adult social care, a flexibility social care authorities have had since 2016-17. MHCLG puts the Band D amount attributable to that precept at £35.14 in 2026-27, against £22.08 in 2022-23. It also warns that those figures understate the cumulative effect, because every increase becomes the following year's baseline.
The 2027-28 limit isn't set, and three documents decide it
This is the part the headlines skip. The 2026-27 settlement is the first multi-year settlement in a decade, running from 2026-27 to 2028-29, and the provisional version, which opened for consultation on 17 December 2025, said the government would "maintain core referendum principles as they were in 2025-26 over the multi-year Settlement". It then listed them: "a core referendum threshold of up to 3%", plus "an additional adult social care precept of up to 2% per year without a referendum". That's where the expectation of a 4.99% council tax increase in 2027 comes from, and it is a stated intention rather than a decision.
The decision itself is annual. Principles "must be set out in a report which must be laid before and approved by the House of Commons", and the 2026-27 report is blunt about its own reach: "Whilst the 2026-27 local government finance settlement covers the period 2026-27 to 2028-29, this report relates to 2026-27 only." The draft for 2026-27 was published on 17 December 2025, to be "laid before the House of Commons for approval at the time of the final settlement", and the final report was ordered to be printed on 9 February 2026. The 2027-28 version follows roughly that calendar.
The statute also has a consequence worth knowing. If the report for a year isn't approved in time, "no principles have effect for that year and accordingly no authority's relevant basic amount of council tax is capable of being excessive for that year". The cap isn't a permanent feature of the tax. It's an annual vote.
Six councils are proposed to have no referendum limit at all
The same provisional settlement contains something the 4.99% framing hides completely. To "enable areas to rebalance disparities in their council tax levels should they wish to, the government proposes not setting referendum principles for 6 authorities in 2027-28 and 2028-29". The criteria are that an authority is "upper tier", "would have 95% of their income protected", and has "council tax levels lower than average".
The six are named: "City of London, Hammersmith and Fulham, Kensington and Chelsea, Wandsworth, Westminster and Windsor and Maidenhead". The document says their "Band D taxpayers in these councils are paying between £450 and £1280 less than the average in England", covering 500,000 households, and that the change releases "£250 million more funding for public services in places with higher need".
For a household in one of those six, the table above sets no upper bound. Wandsworth's average area Band D is £1,028 in 2026-27, the lowest in England, and 4.99% of that is only £51.30. Whether any of the six raise bills by more than that, and by how much, isn't something any published document answers yet. It's a council budget decision taken in February.
Three councils were already allowed 9% in 2026-27
The 8.99% column is not a worst case invented for the table. It's the rate that applied in 2026-27 to three authorities. The principles report sets a limit of "9% (comprising 2% for expenditure on adult social care, and 7% for other expenditure)" for North Somerset Council, Shropshire Council and Worcestershire County Council. Trafford, Warrington and Windsor and Maidenhead were on 7.5%, and Bournemouth, Christchurch and Poole on 6.75%.
These are not random. MHCLG's notes record that "a corresponding approach has been applied for authorities in receipt of exceptional financial support". If your council sits in one of those bespoke categories, the relevant ceiling is the one written into its own paragraph of the principles report, not the 4.99% that covers everyone else. At Band A, 8.99% is £143.36 a year rather than £79.57.
Your bill is not the England average, and the spread is about £1,737
The £2,392 in the table is an average of averages, and it conceals an enormous range. MHCLG reports that "average area Band D council tax ranges from £1,028 (Wandsworth) to £2,765 (Dorset)", with "206 (70%) of authorities" between £2,300 and £2,600. The gap between the two extremes is about £1,737 a year on the same band of house.
That matters for the arithmetic because a percentage increase amplifies a difference it didn't create. A 4.99% rise is £137.97 at Dorset's Band D and £51.30 at Wandsworth's. Same rate, £86.67 apart, before anyone argues about who deserves which. The proposal to lift the limit off the six lowest-charging councils is the government's answer to exactly that chart, and it's why the distribution matters more than the average.
Two thirds of English homes sit below Band D
Band D is a reference band, not a typical home. It's the charge on "a Band D dwelling occupied as a main residence by at least two adults", in the value range £68,001 to £88,000 at 1991 prices. The Valuation Office Agency counted 25,944,800 dwellings on the England list at 31 March 2026, and the distribution leans heavily downward. Band A alone holds 6,123,220 of them, or 23.6%, and "the most frequent council tax band was band A".
Bands A, B and C together account for 65.1% of English homes. Band D itself is 15.7%. Everything from E to H is 19.2%, and Band H, the top band, is 155,210 homes or 0.6% of the stock. So for roughly two households in three, the real 2027 number is the Band A to C row, not the headline. MHCLG publishes the alternative measure for this reason: "the average council tax per dwelling will be £1,868 in 2026-27". If you're tallying what a property costs to keep rather than what a band charges, the true cost of owning a home puts council tax alongside the other running lines.
The cap applies to part of your bill, not all of it
Here's the strongest objection to reading the 4.99% ceiling as a ceiling on your bill. The referendum test applies to the "relevant basic amount of council tax (i.e. the average Band D figure having excluded local precepts)". Parish and town council precepts sit outside it, and the provisional settlement is explicit that there's "no council tax referendum principle for town and parish councils".
That exclusion is not academic. The average Band D parish precept rose to £100 in 2026-27, "an increase of £8, or 8.2% from 2025-26", which is well above the 4.99% that constrained the authorities above it. If you live in a parished area and your parish council has plans, your total bill can rise by more than any of the three columns in the table. Police and fire precepts have their own separate limits too, stated in pounds rather than percentages.
Council tax bands in 2027 still rest on 1 April 1991 values
None of this touches which band you're in. "Council tax bands are based on the value of the dwelling as at 1 April 1991", and anything built since gets a notional 1991 value. The ladder has eight rungs with a top band starting at £320,000 of 1991 value, so every home above that pays the same 18/9 whatever it's worth now.
No revaluation has been announced. The government's consultation on council tax ruled the wider question out of scope in terms: "This consultation is focused on the administration of council tax and will not seek views on fundamental reform of the council tax system." What is coming instead is a second tier. The high value council tax surcharge is "a new charge on owners of residential property in England worth £2 million or more in 2026, taking effect in April 2028", and the Valuation Office identifies those homes through "a targeted valuation exercise" rather than by moving anyone's band. A tier above the ladder is not the same thing as rebuilding it.
What this arithmetic cannot tell you
Every figure in the table is built on one input, the England average Band D of £2,392, and your council's own Band D may be nothing like it. Treat the table as the shape of the increase rather than a quote for your address. The authority-level figures are published, and your own bill is the only one that settles the question.
Two further gaps. The table assumes no discounts, exemptions or premiums, because Band D is defined on at least two adults in the property. A single-occupancy discount, a local council tax support award or a second homes premium all move the cash without touching the percentage. And every council tax increase for 2027 in the table is a scenario. The 2.99% column is the core threshold on its own, which is what binds an authority that leaves the 2% adult social care precept unused. The 8.99% column is the rate three councils were given for 2026-27. Neither has been proposed for 2027-28. The second home council tax premium works the other way round, doubling the band charge outright rather than moving the percentage.
The longer context is worth one line. The England average Band D was £1,444 in 2012-13 and £2,392 in 2026-27, a rise of about 66% over fourteen years. Four of the last five years ran at 4.9% or above. That's a run of outcomes, not a rule, and it's the kind of series that gets broken by a policy change rather than by arithmetic.
What would change the conclusion
If the 2027-28 principles report departs from the multi-year statement, the 4.99% column stops being the central case. The intention was published in December 2025 and the binding instrument is a separate report, so watch the provisional settlement in December and the final report in February.
If the six-authority exemption survives into that report, half a million households in London and Berkshire have no ceiling at all, and the arithmetic above becomes a floor rather than a cap for them. If it's dropped, those councils stay on 4.99% like everyone else.
If a revaluation were ever announced, the band ratios stop being the whole story, because the question becomes which rung you land on rather than what the rung costs. Nothing in that consultation points this way, and the surcharge due in April 2028 is built to sit above the bands rather than to disturb them.
The date to put in the diary is the provisional settlement, which opened for consultation on 17 December in 2025. That document names the core threshold and the adult social care precept for the year ahead. Until it does, every 2027-28 figure anyone quotes you, including the ones above, is a multiplication rather than a measurement.