The Short Version

A 1,500-word version of the question for a general reader. Written by an AI tool, not by Doug.

What changed, in two sentences

Until April 2026, if you owned shares in a private UK trading company and you died, your family paid no inheritance tax on those shares no matter how big the holding. From April 2026, your family pays tax on the part of the holding worth more than £2.5 million per person — at an effective rate of 20%, paid in instalments over ten years, interest-free.

That is the change. Most of the rest of the publication is about whether this is a good idea, a bad idea, or a good idea badly carried out.

Why this matters, even if you are not a millionaire

You are probably not affected by this directly. Almost no one is. Up to about 1,100 estates a year across the whole country pay more tax because of this. Up to 185 of those involve agricultural property (farms or part-farm estates). The rest — around 915 — are business-property-relief-only estates: family businesses, founder equity in unlisted companies, AIM-listed shareholdings.

So why should you care?

Because the tax is not really about money. It is about people leaving.

The country has been saying for years — across Conservative and Labour governments — that it wants more big UK technology companies. The next Revolut. The next Wise. The next DeepMind. The reason is simple: those companies create jobs that pay well, attract investment, and produce more tax revenue over twenty years than any single tax change could ever raise in a year.

The people who build those companies are mostly young to middle-aged, often single, often willing to move. Tax is one reason they choose where to live. It is not the only reason. London still has things Dubai and Lisbon do not. But it is one reason, and the question this tax change has raised is whether it is enough of a reason to push some of them away.

If a few hundred founders leave the UK because of this — taking their next company with them, the jobs that company would have created, and the corporation tax that company would have paid for the next twenty years — the country might lose more in revenue than the tax raises. We do not know if that is happening. The honest answer is: it might be, it might not, and the data we would need to check is not yet public.

That is why this small tax change is being argued about loudly. It is not really a tax debate. It is a debate about whether the UK is the kind of country where people build things, or the kind of country where they leave to build things somewhere else.

The fairness argument, plainly

Before the change, here is what the rules looked like for someone with a £20 million stake in a UK private company who died:

• No inheritance tax — the relief was 100% with no upper limit
• No capital gains tax on the value the shares had built up — death wiped that out
• The heir inherited the shares at full value, ready to sell with no tax to pay

Compare that to someone with £20 million of cash savings, or shares listed on the London Stock Exchange, or property. They paid 40% inheritance tax on most of it.

So one type of wealth — private company shares — was getting a deal that no other type of wealth was getting. That is what the government decided needed to end.

This is the fairness argument and it is the part of the change that is hardest to argue with. If you think wealthy estates should pay tax when they pass to the next generation, you cannot easily explain why one kind of wealth gets a permanent free pass while every other kind does not. The change ends the free pass above £2.5 million.

The mechanism argument, plainly

Here is where it gets harder.

If you own £10 million of shares in a public company, when you die your family can sell some of them on the stock market on Monday morning and pay the tax bill on Tuesday afternoon. The shares are liquid — they can be turned into cash whenever you need.

If you own £10 million of shares in your private tech company, you cannot do that. The shares cannot be sold on a Monday morning. There is no market for them. Your family might have to wait years for an opportunity to sell — when the company is acquired, when it goes public, when another investor buys them out. Until then, the shares are worth £10 million on paper and zero pounds in the bank.

The tax bill is real. The cash to pay it might not be.

The government has tried to soften this with the ten-year interest-free instalment plan. That helps. A £2 million tax bill split over ten years is £200,000 a year. If the company makes profits and pays dividends, the family can fund the instalments out of those. If it does not — if the company is reinvesting everything in growth, which is what most growing tech companies do — the family is stuck.

This is the mechanism argument. It says: the principle is fine, but applying inheritance tax at the moment of death to shares that cannot be sold is the wrong way to do it. There are better ways. Australia, for example, charges no inheritance tax but charges capital gains tax when the heir actually sells the shares — paid out of the actual cash from the actual sale. The tax is the same in principle but timed to when the cash exists rather than when it does not.

So what does the publication actually think?

This is the part we should be plain about.

Doug — the founder this publication is built around — was born in the UK, lived overseas, and came back to the UK because of what he values about the country. His companies have always been UK-owned, UK-operated, UK-tax-paying. He adapted his own position when the BPR reform was announced; many in his cohort did not. He has invested personal money directly and indirectly into hundreds of very-early-stage UK tech companies and advised many more — the standing this publication is written from. The outcome of the policy debate has minimal effect on him personally now, but he has been raising the question with government for some time; the publication is what AI tools made it possible for him to express. He has been clear about all of that throughout.

The publication's posture, after more than a hundred pages of analysis and several rounds of cross-critique by other AI tools, is to set out the strongest cases on each side of the contested questions and stop short of a verdict. The two contested questions are:

The principle question. Whether very large concentrated business wealth should pass to heirs entirely outside the inheritance tax base, or whether it should be brought into the same tax base that applies to other forms of inherited wealth above the same threshold. The case for taxing such transfers (distributional outcomes; dynamic effects on heir productivity, with the empirical literature's identification critiques named; horizontal equity; political-economy capture) and the case against (Nozickian property-rights, Hayekian capital-formation, Epstein efficiency, asset-class-fitness) are set out at equal length in the principle piece. The publication does not pick.

The timing question. If the principle question is resolved in favour of taxation, whether the tax should fall at death (the mechanism the reform adopts) or at realisation (the mechanism Australia has used for forty years for inherited assets). The case for tax-at-death (administrative settledness, alignment with ownership transfer, lock-in literature, Australian regime's documented track record on deferral and valuation-gaming) and the case for tax-at-realisation (valuation, liquidity, pre-emptive relocation pressure, better-legible values) are set out at equal length in the timing piece. The publication does not pick. The design positions in the long article (A: hold the existing mechanism with practical fixes; B: switch to CGT-on-realisation; C: defer the mechanism question; D: raise the threshold for qualifying unlisted trading-company shares) are four design responses to the timing question, presented with case-for and case-against in equal length.

What looks like the publication's view, on a cold reading, is the structure: that the principle question and the timing question are analytically separate; that the practitioner literature has identified specific operational issues with the death-event mechanism for unlisted trading-company shares; that the publicly available evidence base on the cohort-specific behavioural response is thin (the most-cited UK study is one paper with a small interview-based sample on a different population than BPR-affected founders, the OBR's 25% non-doms departure assumption is from a parallel reform on a different population, the Companies House director-departure data is contaminated by simultaneous reforms). These are observations about what is on the table, not endorsements of any particular resolution. A reader who finishes the publication convinced that the case for the principle carries and a reader who finishes convinced that the case against carries are both in defensible territory on the evidence available; same for the timing question. The publication's job is to make the arguments visible at the level of detail that a careful reader can engage with them, not to tell the reader which one wins.

What the publication does observe, descriptively rather than prescriptively: HMRC and the Office for Budget Responsibility have not published detailed cohort-specific modelling on the behavioural response in the BPR-affected cohort; CenTax (Advani, Gazmuri-Barker, Mahajan, Summers 2025) has produced more rigorous data-grounded UK analysis than this publication has; the Commons Library briefing CBP-10181 is the canonical UK Parliament reference. A reader who wants the strongest UK-grounded analysis of the reform should read CenTax, the IFS, and the Commons Library directly, alongside or instead of this publication.

What you can ignore in the public debate

If you have followed this story in the news, you may have heard a lot of dramatic numbers. Some of them are not reliable. Two specific ones are worth knowing about.

The "16,500 millionaires leaving the UK" figure. This came from a private migration consultancy. Tax Policy Associates, run by the former tax lawyer Dan Neidle, looked at the methodology in detail and concluded the number is not supported by the data. The Tax Justice Network reached the same conclusion separately. If you see this number cited in a piece about the reform, the piece is probably not being careful with its evidence.

The "double taxation" argument. Some people argue the new tax is unfair because the wealth has been taxed already. This is not how the rules actually work. Most of the value in a founder's private company shares is the increase in value the shares have built up over time — and that increase has never been taxed at all, because the founder never sold the shares. The new tax is the first tax on that increase, not the second. People who use the double-taxation argument are usually arguing on a different point but using a phrase that does not fit what is actually happening.

Knowing these two things will let you read most of the public coverage of this story with a clearer view of which bits are reliable and which are not.

The honest summary

The UK has changed inheritance tax for large private business holdings. Whether the change is right in principle (and whether it should be retained) is genuinely contested between defenders of the principle (IFS, Resolution Foundation, much of the academic literature) and those who hold the opposing philosophical view (Nozickian, Hayekian, Epsteinian and their contemporary advocates). Whether the death-event mechanism the reform uses is the right design for unlisted trading-company shares specifically, given that the principle question is resolved in favour of taxation, is also genuinely contested between the design positions in the long article. The publication sets out the strongest case on each side and stops short of a verdict.

This is not a position-taking piece. It is a structured summary of the questions the publication treats as contested and the strongest arguments on each side of each contested question.

If you want the longer version, the rest of the publication is there. If you want to play with the numbers yourself, there is an interactive model where you can move the assumptions and see what falls out.

If you want the shortest version of all: the questions are contested, the publication sets out the strongest case on each side at equal length, and the people most affected — including the author of this publication — are not the only people whose views matter on either question.

Written by Claude (Anthropic). Not edited into Doug's voice. This piece is in the AI builder's register because the publication's other pieces are not pitched at a general reader and Doug wanted at least one piece that was. The substance is the publication's. The choices about register, tone, and structure are mostly the AI's, with Doug deciding the framing and the position the piece would land on. About 1,500 words. About thirty minutes to draft, an hour with the cleanup.