There is a joke that the Laffer curve is a napkin drawing that convinced a generation of politicians to cut taxes for the rich. The napkin part is true. The rest is a misunderstanding of what the curve actually says.
The Laffer curve is not a theory about rich people. It is arithmetic. Tax revenue is the tax rate multiplied by the tax base. If you raise the rate and the base stays the same, revenue rises. If the base shrinks, revenue can fall. At some rate - nobody knows exactly where - the second effect overtakes the first, and a higher tax collects less money. Every tax has such a point. The only question is where it sits.
That question is usually impossible to answer in practice, because you cannot run the same country twice, once with the higher rate and once without. But Scotland has just done the closest thing to a controlled experiment that tax policy ever gets. And the early data suggests the curve is real - and that Scotland may have parked its top rate on the wrong side of it.
The experiment
In 2017, Scotland gained the power to set its own income tax rates. Since then it has raised its top rate three times: from 45p to 46p, then to 47p, and in April 2024 to 48p. The rest of the United Kingdom kept its top rate at 45p throughout.
This is the experimental setup. Same currency. Same labor market. Same legal system. No border, no customs, no paperwork. A Scottish high earner can move to England by changing an address - or, in many cases, without moving at all. The only difference between Scotland and the rest of the UK is the tax rate itself, three points higher.
Dan Neidle of Tax Policy Associates called it “about as close to a controlled experiment as tax policy gets.” He was not being charitable. He was being accurate.
The revenue projections for the 48p rate were unusual from the start. Multiply the new rate by the number of people earning above the threshold, and the rise should have collected £53 million. But the Scottish Fiscal Commission, the official forecaster, adjusted for how taxpayers respond to higher rates - avoidance, reduced effort, migration - and its estimate collapsed to £8 million. Eighty-five percent of the static revenue disappeared before the rate even took effect.
That should have been the headline. The official forecaster of the Scottish Government was saying, in effect, that the 48p rate would raise almost nothing. Instead, the rate went ahead anyway.
What the data shows
In July 2026, HMRC published the first full-year data covering the 48p rate. The comparison with the rest of the UK is stark.
The number of Scottish taxpayers earning above £125,140 grew 17 percent - but that is mostly the frozen threshold pulling people into the band, not new wealth. The income declared by each of those taxpayers tells the real story: it fell 7.6 percent in Scotland, while rising 5.4 percent in the rest of the UK. Total income above the threshold grew 8 percent in Scotland against 16 percent in the rest of the UK.
Then there is self-assessment income - the tax that people declare themselves rather than having it deducted from a salary. Scotland’s share of UK self-assessment income fell. The drop was statistically significant, and it was the only major Scottish tax statistic to fall in absolute terms. PAYE, the tax deducted from salaries, held steady. The money that vanished was exactly the income people control themselves.
Neidle’s estimate: the 48p rate cost Scotland about £22 million in its first year - more than the £8 million it was forecast to raise, and possibly more than the zero the government might privately have hoped for. The range is £15 to £30 million.
The mechanism is the lesson
This is the part that matters, because it tells you how the curve works in practice. The response was not millionaires fleeing to Dubai. Migration is real, but it is slow and second-order. What happened instead was cheaper and faster: people who control their own income simply moved it.
A Scottish company owner pays herself in dividends instead of salary. Dividends are reserved to Westminster, outside Scotland’s tax bands. A Scottish employee makes a bigger pension contribution, reducing declared income. Both are legal, immediate, and invisible to anyone reading a headline.
This is the structural trap of Scotland’s tax powers. Holyrood can set rates only on employment and self-employment income. Everything else - dividends, savings income, capital gains, the personal allowance, every relief - stays with Westminster. So the base of Scotland’s income tax is precisely the portion of income that is easiest to move on paper. Tax that base harder, and it does not fight back by moving house. It just changes form.
There is a second layer most people never see. The headline 48p rate is not what high earners actually face. The personal allowance tapers away between £100,000 and £125,140, and the interaction of the taper with the new 45 percent advanced rate produces a marginal rate of 69.5 percent in Scotland - against 62 percent in England. The person earning £110,000 in Scotland is not paying 48 percent on their next pound. They are paying closer to 70. That is not a rounding error in the design; it is the design.
The denial
Here is where the story stops being arithmetic and becomes politics.
Back in 2021, the Scottish Fiscal Commission acknowledged that the first rise, from 45p to 46p, had “raised limited additional revenues, and might even have resulted in a small loss of receipts.” The Commission said, in public, that the policy might not be raising money. The government raised the rate again. And again.
When the new data came out this year, the Scottish Government’s response, as reported by the Sunday Times, did not engage with the numbers at all. That is the tell. When a policy is a symbol, the data is an inconvenience. The 48p rate was never chiefly about revenue - £22 million is a rounding error against Scotland’s £18.6 billion income tax take. It was about demonstrating that Scotland taxes its highest earners more heavily than England. That is a political message, and the message survives the arithmetic.
Bastiat’s framework fits this perfectly. The visible effect is the rate: 48p, the highest in the UK, a gesture of taxing the rich. The invisible effects are the income that changed form, the self-assessment base that shrank, the £22 million that never arrived. Politicians are rewarded for the visible. The invisible arrives later, diffuse, and attributable to nobody. If you want the full argument, it is in The Seen and the Unseen.
The caveats, honestly stated
If you have read this far, you deserve the other side of the argument. The case is strong, but it is not proven, and anyone who tells you otherwise is selling something.
First, the data is a nine-point annual series. Statistical tests on so few observations are fragile. It could still be noise, though the two independent comparisons - the income ratio against the rest of the UK, and the absolute fall in self-assessment - point the same way.
Second, the North Sea. Scotland has a highly paid, shrinking offshore oil and gas sector, and that is mostly a Scotland-specific effect. But the income that went missing is far larger than the sector could plausibly explain, and the last genuine oil price crash, in 2020, did not produce this pattern.
Third, the counterfactual. The rest of the UK is not a perfect Scotland-without-the-48p-rate. London’s finance and professional services incomes dragged the UK average up, and it is possible Scottish high earners would have grown more slowly anyway. If so, the £22 million loss is an overestimate.
Fourth, the precedent cuts both ways. When the UK itself ran a 50p additional rate from 2010 to 2013, the revenue effect was so close to zero that nobody could agree even on its sign. That is consistent with the Laffer curve - and it is also a reminder that “close to the top of the curve” does not always mean “over the top.”
The honest summary: the data is consistent with the hypothesis that Scotland’s top rate is on the wrong side of the Laffer curve. It does not prove it. Next year’s data will tell us more, and the granular data that would settle it - bunching around the £125,140 threshold, pension and dividend shifts, address changes - exists inside HMRC but has not been published. The Institute for Fiscal Studies has speculated, more cautiously, that Scotland’s top-rate rises “may have reduced revenues.” The mainstream of the profession is moving the same direction, one hedge at a time.
The lesson
None of this requires believing in voodoo economics. It requires believing in arithmetic, and in the fact that people respond to incentives. Tax a thing and you get less of it. Tax income that can change form at the stroke of a pen, and you get less of that income. The rest is detail.
The deeper lesson is about what taxes are for. If the goal of the 48p rate was revenue, it failed - it appears to have lost money. If the goal was a symbol, it succeeded, and the arithmetic is irrelevant. The danger is when the second goal is sold as the first: a gesture dressed as a policy, whose costs are paid invisibly and later, by people who did not vote for it.
That is why the Laffer curve matters even when the sums are small. It is the mechanism by which a popular tax quietly stops working. Scotland is a small country and £22 million is a small number. But the mechanism scales. Every government that raises a rate past the revenue-maximizing point is running the same experiment, with the same invisible cost, on a larger stage. We just got to watch this one happen in real time.
If you want to understand why a government would keep raising a rate its own forecaster said would raise almost nothing, the answer is not in the spreadsheet. It is in the incentives - the same ones documented in Incentives Matter: The Economics of Obvious. Politicians are rewarded for the gesture. The bill arrives later, and somebody else pays it. That is not a bug in the system. It is the system.