Reacting to: Scotland's tax hike may have backfired as receipt falls (City A.M.) →
Most of the coverage of this story is about Scotland, the Laffer curve, and whether a politician got their sums wrong. That is the least interesting part of it. What actually happened here is that a government moved one tax rate by three points, held everything else constant, and got to watch — in real HMRC outturn data — what people with control over their own income do next. Very little of UK tax policy ever gets tested that cleanly.
And the answer is not the one the headlines imply. Almost nobody appears to have moved house. What they did instead was change the form their income arrived in — a dividend rather than a salary, a pension contribution rather than a bonus. That is not a Scottish loophole. It is the ordinary machinery of an owner-managed business, and my honest view is that it is the single most under-used lever sitting in front of directors in Belfast, Bolton and Bexley too. Scotland has just given everyone an unusually clear reading of how much it is worth.
What the data actually shows
The analysis is by Dan Neidle's research body Tax Policy Associates, reported by City A.M. on 26 July. Scotland's top rate of income tax went from 47p to 48p in April 2024, against 45p in the rest of the UK, and a new "advanced rate" of 45% appeared on income between £75,000 and £125,140 where the rest of the UK charges 40%. On 9 July, HMRC published the first outturn covering a full year of the 48p rate, which turned the projections into something checkable.
The numbers, as reported:
- The static estimate — rate multiplied by the people in the band, before anyone reacts — was that the 48p rate would raise £53m.
- The Scottish Fiscal Commission's own costing expected 85% of that to disappear in taxpayer responses, leaving £8m.
- Neidle's reading of the outturn is that it is instead costing around £22m, which he calls a conservative estimate that could reach £30m.
- For scale, Scotland's total income tax take is £18.6bn. Even on the worst reading this is a rounding error — roughly one part in eight hundred.
The supporting detail is the part I find persuasive. Scotland's share of UK PAYE income tax in 2024-25 was 8.43% — essentially unchanged against the previous eight years. Its share of self-assessment income tax fell to 4.90%, which Neidle describes as a statistically significant drop. PAYE holding steady while self-assessment falls is not the signature of people leaving. It is the signature of people who file a tax return — landlords, investors, company owner-managers — managing the number down.
Why the lever exists at all
Here is the structural fact that makes the whole thing work, and it is worth knowing even if you never set foot in Scotland.
Holyrood can set income tax rates and bands on non-savings, non-dividend income only — wages, pensions, self-employment profits, rental profits. Everything else stays with Westminster. GOV.UK puts it in one line on its Income Tax in Scotland page: you pay the same tax as the rest of the UK on dividends and savings interest. Capital gains tax, the personal allowance, the dividend allowance and every relief are reserved as well.
So a Scottish company owner who takes a dividend instead of a bonus has stepped outside the devolved tax base without moving, without restructuring, and without doing anything a tax inspector would blink at. A Scottish director making a pension contribution has done the same. Both are, in Neidle's phrase, cheap, legal and immediate.
Putting real numbers on it
These are illustrations using published 2026/27 rates, not a substitute for looking at your own accounts.
First, the employee case, to size the gap. Take a salary of £100,000 with a standard £12,570 personal allowance. In Scotland the bill runs through six bands — 19%, 20%, 21%, 42% to £75,000 and then 45% — and comes to £30,732. The same salary in England or Northern Ireland is taxed at 20% then 40% and comes to £27,432. A gap of just over £3,300, and note that most of it is the advanced rate, not the famous 48p. National Insurance is identical on both sides of the border because it was never devolved.
Now the owner-director case, which is where behaviour changes. Take £30,000 of company profit and a director already taxed at the top rate, and ask what actually lands in their pocket.
- Route A — a bonus through payroll. £30,000 of company cost supports a gross bonus of £26,087 once employer's National Insurance at 15% is covered. For a Scottish director, 48% income tax plus 2% employee NI takes exactly half: £13,043 net. For a director in England, 45% plus 2% leaves £13,826 net.
- Route B — a dividend. £30,000 of profit bears corporation tax at 25%, leaving £22,500 to distribute. Dividend tax at the additional rate of 39.35% leaves £13,646 net — the same figure on both sides of the border, because dividends are not devolved.
Read those two lines together and the behaviour explains itself. In England the bonus wins by about £180. In Scotland the dividend wins by about £603. The three-point difference in one rate flips the answer, on the same profit, in the same company, for the same person. Multiply £603 by every top-rate owner-manager in Scotland who has an accountant, and you have Neidle's £22m without a single removal van.
Two things sharpen it further. If the company pays corporation tax at 19% rather than 25% — profits at or under £50,000 — the dividend route nets closer to £14,738, and the gap widens again. And the pension route dwarfs both: a £30,000 employer contribution is normally deductible for corporation tax, so £30,000 lands in the pension for a net company cost of £22,500. Against a bonus, the arithmetic is stark — every £1 of cash bonus reaching a Scottish top-rate director costs the company £2.30 of profit. The annual allowance is £60,000, tapering where threshold income exceeds £200,000 and adjusted income exceeds £260,000.
What this means in practice — and it depends who you are
If you are an owner-director anywhere in the UK. The lesson is not "take dividends". It is that the salary-versus-dividend split is a live decision with a four-figure answer, and most owners set it once at incorporation and never revisit it. City A.M. notes that Andy Burnham and John Healey will face pressure to raise taxes on higher earners to fund spending later this year, and that Burnham has rowed back on an earlier pledge to take the top rate to 50%. Nobody outside the Treasury knows what lands in the autumn. What you can control is whether your extraction is modelled or inherited. Our post on how to pay yourself from a limited company walks through the mechanics, and the salary and dividend calculator will give you your own version of the numbers above in about two minutes.
If you are a Scottish taxpayer on PAYE with no company. You have the fewest levers, which is the uncomfortable finding underneath all of this. Your £3,300 gap on £100,000 is not avoidable by reclassifying anything. Pension contributions get relief at your Scottish marginal rate, which makes them worth more to you than to your English equivalent — the one place where the higher rates work in your favour.
If you are a landlord. Rental profits are non-savings, non-dividend income, so Scottish rates apply to them in full. That interacts badly with the Section 24 restriction on mortgage interest relief, which inflates taxable profit before the rate is applied. Scottish landlords feel that combination harder than anyone.
Two things to do this week
1. Model your extraction for the current year before you draw anything else. Not last year's split repeated. Run the actual profit, your actual other income and the £500 dividend allowance through the salary and dividend calculator, and put the answer in writing. If your accountant has not sent you a remuneration plan for 2026/27, that is a reasonable thing to ask for.
2. Decide your pension contribution now rather than in March. An employer contribution has to be paid by the company's year end to land in that period's corporation tax computation, and the £60,000 allowance does not roll on indefinitely — unused allowance carries forward three years and then falls away. Booking it in July rather than discovering it in a January panic is worth real money.
What is still uncertain, and when you will know
Neidle is careful about his own finding, and so am I. The £22m rests on the assumption that Scottish incomes would otherwise have grown as they did in the rest of the UK; single-year tax data carries noise, and non-tax effects can swing the number of high earners in a small population. The Scottish Government's response, quoted by City A.M., is that taxpayer numbers and liabilities grew strongly in 2024-25 and that the count of top-rate taxpayers grew faster in Scotland than in the rest of the UK. That is not a rebuttal of the per-taxpayer point, but it is a real fact that sits alongside it.
Two dates will move this on. The next annual HMRC Scottish income tax outturn will show whether 2024-25 was a step change or a wobble — one year does not settle a question this contested. And the autumn Budget will tell you whether any of this shaped the Treasury's thinking on higher-rate taxation UK-wide. Until then, the honest position is that the direction of travel is well evidenced and the exact figure is not.
None of which changes the practical point. Whatever the Chancellor does in the autumn, the value of getting your salary, dividend and pension mix right went up, not down. That is ordinary tax planning work, it is part of what we do for clients on our packages rather than a bolt-on, and it is considerably easier to do in July than in the week before a year end.