The Tax Office Was Told It Isn't Collecting Hard Enough. 1.3 Million Small Businesses Will Find Out What That Means.

Abstract illustration of a wide field of small identical shapes with a measurement line being drawn across them

The Tax Office Was Told It Isn't Collecting Hard Enough. 1.3 Million Small Businesses Will Find Out What That Means.

The headline was $35.9 billion. The number that should worry an ordinary business is $26,800 — because that's the average, and averages are what targets get set against.

Back at the end of June, the Australian National Audit Office tabled a performance audit of how the ATO manages small business collectable debt. It found the ATO's management only partly effective, made eight recommendations, and the ATO agreed to all of them.

That was ten weeks ago. The trade press covered it in early July and moved on, which is the normal life cycle of an audit report and is exactly why it is worth returning to now. Nothing has visibly changed for a small business since June. That gap — between a regulator accepting a recommendation and the recommendation reaching a taxpayer — is the only useful preparation window anyone gets, and it is open right now.

The number that isn't the headline

Small business collectable tax debt was $35.9 billion in 2024-25 — roughly two-thirds of the $54.2 billion total, and $19.4 billion higher than in 2018-19. That is a doubling in six years, and most of the commentary has treated it as a story about the aggregate.

It isn't. Divide it out. More than 1.3 million small businesses carry collectable debt, at an average of about $26,800 each.

Twenty-six thousand dollars is not a distressed number. It is one quarter's BAS that got deferred during a bad month and never quite got caught up, or a superannuation guarantee charge from a payroll error, or an income tax bill that arrived when the cash didn't. It is the kind of balance a business owner describes as "we're on top of it." Which means the population the ANAO is describing is not a tail of failing businesses. It is a very large body of ordinary ones.

1.3 million+
Small businesses carrying collectable tax debt in 2024-25, at an average of roughly $26,800 each.
$35.9bn
Small business share of the $54.2bn total collectable debt — up $19.4bn on 2018-19.

Why a target changes behaviour more than a policy does

The ANAO's central operational criticism is worth stating precisely, because it is subtler than "collect more."

The ATO has a target for the overall collectable debt ratio. It does not have a target for the volume of small business collectable debt, and the audit found that without clearly defined targets or performance measures attached to its individual debt controls, the ATO cannot define what success looks like for any of them. The recommendation is to fix the measurement.

Anyone who has run a finance function knows what happens next, because we all do the same thing. A control with no target is a process. A control with a target is a number somebody is accountable for at the end of a reporting period. The activity does not change because someone wrote a new policy; it changes because a graph now has a line on it and the line has a name against it.

So the practical prediction is not "the ATO will get tougher." It is narrower and more useful: measured, consistent application of the escalation steps that are already published and already available, applied to a population that has grown used to the earlier steps not being followed through.

The escalation path is not a secret

None of the ATO's firmer-action tools are new, and the sequence is published. Reminders and contact. A warning that firmer action is being considered. Then, depending on the debt and the response: garnishee notices to a bank or a debtor, director penalty notices, disclosure of the debt to credit reporting bureaus, and at the far end, wind-up action.

Two features of that list matter more than the rest to anyone reading this from a small or medium business.

Director penalty notices are personal, and they are not confined to unpaid PAYG withholding. The regime reaches PAYG withholding, superannuation guarantee charge and GST. A director who thinks of an ATO balance as a company liability is describing a subset of it.

Credit reporting disclosure has a defined gate, and it sits a long way above the average. The ATO may report a business tax debt to a credit bureau only where the business has an ABN, is not an excluded entity, has at least $100,000 overdue by more than 90 days, and is not engaging with the ATO to manage it. A formal intent-to-disclose notice comes first. Registered charities, deductible gift recipients and government entities are excluded outright — which takes a good part of this blog's readership out of this particular mechanism, though not out of the director penalty regime, which carries no dollar threshold at all. Where disclosure does happen, the debt becomes visible to lenders, insurers and trade credit providers, and gets read as a signal about management rather than about one quarter's cash. The listing is removed once the debt is paid in full or the business is effectively engaged. The credit decisions taken while it was visible are not.

The thing most businesses can't actually tell you

Ask a small business what it owes the ATO and you will usually get one number, taken from the integrated client account balance. That number is nearly useless for managing the risk, because the escalation exposure attached to a dollar of GST is different from a dollar of superannuation guarantee charge, which is different again from a dollar of company income tax. Same balance, entirely different consequence for a director.

What is actually needed is an ageing of the ATO position by debt type and by period, sitting alongside the general interest charge accruing on each. Almost nobody has this, for a boring reason: the ATO account transaction history exports as a long, flat, chronological list of postings that is genuinely tedious to read, and reconstructing the composition of a balance from it by hand takes an afternoon that nobody schedules.

Where AI earns its keep here

This is a narrow, unglamorous and genuinely good use case: taking a messy chronological transaction export and normalising it into a structured ageing — debt type, period, original amount, payments applied, interest accrued, current balance.

It is good precisely because it is mechanical. There is no judgement in it. Every figure has a source line you can point at, the totals have to reconcile back to the account balance, and if they don't, you know immediately that something is wrong. That reconciliation is not optional and it is not a formality — it is the control that makes the whole exercise trustworthy, and it takes a minute.

What comes out the other side is the thing a director should be looking at once a quarter: how much of this is personally exposed under the director penalty regime, how much is accruing interest at a rate that makes a bank facility cheaper, and how old the oldest component is. Those three answers change what you do. The single balance figure doesn't.

What AI should not be doing is deciding whether to enter a payment arrangement, judging whether a remission request has merit, or interpreting a director's position under the penalty regime. Those are advice, they carry personal liability, and they need a registered practitioner.

On the data: an ATO account transaction history is about as sensitive as business data gets, and it usually carries identifiers alongside it. Before it goes into an AI tool, confirm whether the vendor trains its models on customer inputs — the safest posture is a tool where your data isn't retained for training. And de-identify first: strip names, tax file numbers and participant identifiers before upload. A no-training contract governs what the vendor does with your data; APP 11 still governs whether you should have sent it in that form.

Four things to do this quarter

Split the balance. Ageing by debt type and period, reconciled to the account. If you cannot produce this in under an hour, that is the finding.

Quantify the director-exposed portion separately. PAYG withholding, superannuation guarantee charge and GST. Put that number in front of the directors specifically, not folded into a total.

Price the general interest charge against your actual alternatives. Businesses routinely carry an ATO balance at a rate well above what a facility would cost them, because the ATO does not send a monthly statement that makes the comparison obvious. Make the comparison yourself.

If you are going to engage, engage early. The published escalation path gives considerably more room to a taxpayer who makes contact before firmer action starts than to one who responds after. That asymmetry is the whole design, and it is the cheapest lever available.

Can you tell your board how much of your ATO balance is personally exposed?

Most businesses can quote the balance and cannot break it down — which means the directors are carrying a number nobody has sized. PFL provides senior-level outsourced finance, management reporting, and AI automation for Australian NFP, NDIS, and SME organisations, including getting the tax position onto one page that a board can act on.

Talk to PFL →
This post is general commentary based on publicly available information and does not constitute legal or tax advice. Director penalty exposure, remission and payment arrangement outcomes depend entirely on individual circumstances. Always seek independent professional advice before acting.
Timothy, CPA is Managing Director of Professional Financelink (PFL), providing senior-level outsourced finance, management reporting, and AI automation for Australian NFP, NDIS, and SME organisations. 20+ years in finance leadership across NFP, NDIS and SME.

Comments

Popular posts from this blog

Google Gemma 4 Just Launched — And It Might Solve Finance's Biggest AI Privacy Problem

Claude vs Gemini for Australian Finance: An Honest Comparison After 12 Months of Using Both

Why NFP Boards Are Finally Talking About AI — And What the Finance Team Should Do Before They Ask