A tax break bought by promising never to change is grandfathering, not reform

The price of the discount is a promise never to decide anything

A discretionary trust exists for one reason: the person running it gets to choose, each year, who receives the income. The government is now offering to tax those trusts more lightly on condition that they never make that choice again.

That is the deal reported today. A new minimum tax of 30 per cent is coming for these structures. Holders of discretionary trusts can be exempt from it if they agree to hold their distributions constant. Whether you call that a workaround or a concession, the mechanism is the same: relief is purchased with a permanent undertaking not to change who gets what.

Start with why the minimum rate was proposed at all. The complaint about discretionary trusts is that a business owner facing the top marginal rate can direct income to a spouse with little other income, an adult child at university, a retired parent — each taxed at a lower rate. The household ends up with the same money and a smaller tax bill than a wage earner who cannot split anything. The minimum rate is a blunt answer to that: set a floor and the arithmetic stops working.

Freezing the split does not undo the split

Here is the problem with the exemption. The revenue cost of income splitting comes from the split itself, not from its variability. A trust that spread income four ways in the baseline year and never touches it again is not shifting income opportunistically — it does not need to. It already did.

So the exemption spares exactly the arrangements that were built well before the rule arrived, and denies relief to anyone who has to build one after. Two businesses of identical size and identical family structure end up on different tax rates because of when they organised themselves. Tax systems that sort taxpayers by vintage rather than by conduct lose legitimacy quickly, and they are very hard to unwind, because the people holding the concession now have something to defend.

Constant in dollars, or constant in shares?

The entire policy turns on a word the announcement does not define. Constant in dollar terms and constant in proportional terms are opposite rules.

If the baseline is a schedule of dollar amounts, then every extra dollar the business earns has to land somewhere outside the frozen schedule — plausibly with the highest-rate beneficiary. The concession then quietly erodes as the business grows, and grows most punitive for the enterprises doing best. If the baseline is a set of percentages, the shelter scales without limit: the more successful the business becomes, the more income is permanently held below the top rate. One version is a growth penalty. The other is an uncapped entitlement. Neither is neutral, and choosing between them is not a drafting detail to be settled in explanatory material after the vote.

Families are not constant

Distributions change because lives change. A beneficiary dies. A couple separates. A child finishes study and starts earning, so directing income to them no longer works and may now be worse than doing nothing. A new partner buys into the business. Each of these forces a change in who receives what.

If the price of any of those events is losing the exemption and jumping to the 30 per cent floor, the concession is a trap for the unlucky and a gift to the static. And if there is a list of permitted adjustments — as there must be, because the alternative is grotesque — then every item on that list is a doorway, and the advisory industry will spend the next decade mapping it. The measure will be back before Parliament within five years either way.

The best case for the deal

The strongest argument for the constancy condition is administrative, and it is not weak. Revenue authorities cannot audit motive. They can audit arithmetic. A rule that says “your distribution pattern must not move” is cheap to verify from returns already lodged, needs no judgment about intent, and eliminates the behaviour that most offends a progressive rate scale: the annual reshuffle, where income is aimed each June at whichever family member is currently cheapest. On that view the target was never the small business with a settled family arrangement. It was the structure rebuilt every year-end. And a rule that bites only when you change something is politically survivable, which matters for a measure that would otherwise be repealed by the next government.

But administrability is a reason to prefer one enforcement method over another. It is not a reason to convert an avoidance opportunity into a property right. The reshuffle is the symptom; the split is the cost. Buying peace with the auditors by handing existing holders a permanent, unreviewable position is the most expensive way to solve a paperwork problem.

What would settle it

One document. The baseline year, the unit of measurement, whether the schedule is indexed, whether the exemption ever expires, and the exhaustive list of events that permit adjustment. Every serious question about this policy is answered on that page, and none of it has been published.