For the 50th time.

Thirteen months later, on 12 May 2026, the government legislated exactly the changes he had ruled out. The 50% capital gains tax discount will be reduced to 30% (inflation-indexed) from 1 July 2027. Negative gearing will be tightened to new builds only. The bill passed the House on 4 June.

So how does a fifty-times ruling-out become a policy? Here is the sentence Albanese offered on ABC Radio National the week the Budget dropped, verbatim:

Government is about making the right decisions for the right reasons for the times that you are in… if we do change our position on any policy, we will explain why.

Two things about this sentence are worth sitting with.

First, it is a defence that could be deployed by any government for any reversal. It has no content specific to negative gearing or capital gains tax. It doesn't explain what changed in the “times” between the 50th ruling-out and the Budget speech. There is no reference to a new economic condition, a fresh piece of Treasury modelling, a shifted revenue need. There is only the abstract structure of a defence: things changed, so we changed.

Second, and more precisely: it treats the ruling-out itself as if it were merely one input into a broader decision-making process. The problem is that a ruling-out is not an input. A ruling-out is a promise. It is the specific speech-act of foreclosing a future decision so that voters can rely on it when deciding how to vote. To describe reversing a promise as “making the right decisions for the times that you are in” is to describe promise-making itself as optional. If a fifty-times ruling-out survives contact with an executive calculation, no ruling-out means anything, at any count.

This is not an argument that the underlying policy is wrong. Reasonable economists have argued for years that the CGT discount and negative gearing settings should change. It may well be that the government's 2026 position is materially better economics than its 2025 position. That is a separate question. The question here is what a ruling-out is worth when the person doing the ruling out uses the language of irreversibility (“How hard is it?”), attaches a specific number to it (fifty), and does so on camera during a campaign.

The Ledger's view is that the answer is: nothing. A ruling-out is worth nothing. Which is fine as an operating principle — voters can adjust their expectations accordingly — but only if that operating principle is stated openly. What is not fine is running the ruling-out for the votes it produces at the election, then reaching for a defence that would apply equally well if no ruling-out had ever been given.

The receipts, dated and cited, live in the tracker. If the argument above is wrong, the argument is wrong on the tone. The facts are not in dispute.